{ "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": 4.2660064697265625, "learning_rate": 0.0, "loss": 0.9782, "mean_token_accuracy": 0.7547097206115723, "num_tokens": 38293.0, "step": 1 }, { "epoch": 0.00025442055718102023, "ewc_loss": 0.0, "ewc_loss_diag": 0.0, "ewc_loss_parallel": 0.0, "grad_norm": 4.474783420562744, "learning_rate": 4.23908435777872e-09, "loss": 1.0326, "mean_token_accuracy": 0.7414199113845825, "num_tokens": 79961.0, "step": 2 }, { "epoch": 0.0003816308357715303, "ewc_loss": 2.0339285811132868e-13, "ewc_loss_diag": 1.8648277366750676e-17, "ewc_loss_parallel": 2.0122792321330962e-16, "grad_norm": 4.321020126342773, "learning_rate": 8.47816871555744e-09, "loss": 0.9021, "mean_token_accuracy": 0.7778692841529846, "num_tokens": 118063.0, "step": 3 }, { "epoch": 0.0005088411143620405, "ewc_loss": 2.6147972675971687e-12, "ewc_loss_diag": 2.1250362580715887e-16, "ewc_loss_parallel": 2.5951463200613034e-15, "grad_norm": 4.4616594314575195, "learning_rate": 1.271725307333616e-08, "loss": 0.976, "mean_token_accuracy": 0.7540358901023865, "num_tokens": 149306.0, "step": 4 }, { "epoch": 0.0006360513929525506, "ewc_loss": 1.9667822925839573e-11, "ewc_loss_diag": 1.6861512186494565e-15, "ewc_loss_parallel": 1.9539925233402755e-14, "grad_norm": 4.2763671875, "learning_rate": 1.695633743111488e-08, "loss": 0.9986, "mean_token_accuracy": 0.7505897879600525, "num_tokens": 192477.0, "step": 5 }, { "epoch": 0.0007632616715430606, "ewc_loss": 1.191438059322536e-10, "ewc_loss_diag": 9.936496070395151e-15, "ewc_loss_parallel": 1.1812772982011666e-13, "grad_norm": 4.617750644683838, "learning_rate": 2.11954217888936e-08, "loss": 0.9607, "mean_token_accuracy": 0.762528657913208, "num_tokens": 226246.0, "step": 6 }, { "epoch": 0.0008904719501335708, "ewc_loss": 4.129105946049094e-10, "ewc_loss_diag": 4.618527782440651e-14, "ewc_loss_parallel": 4.0678571622265736e-13, "grad_norm": 4.454232692718506, "learning_rate": 2.543450614667232e-08, "loss": 1.0129, "mean_token_accuracy": 0.7554869651794434, "num_tokens": 263593.0, "step": 7 }, { "epoch": 0.001017682228724081, "ewc_loss": 9.604264050722122e-10, "ewc_loss_diag": 7.993605777301127e-14, "ewc_loss_parallel": 9.521272659185342e-13, "grad_norm": 4.262063503265381, "learning_rate": 2.967359050445104e-08, "loss": 0.9099, "mean_token_accuracy": 0.7726661562919617, "num_tokens": 298202.0, "step": 8 }, { "epoch": 0.001144892507314591, "ewc_loss": 1.4115357771515846e-09, "ewc_loss_diag": 1.1812772982011666e-13, "ewc_loss_parallel": 1.3997691894473974e-12, "grad_norm": 4.525773048400879, "learning_rate": 3.391267486222976e-08, "loss": 1.0362, "mean_token_accuracy": 0.7471110224723816, "num_tokens": 340163.0, "step": 9 }, { "epoch": 0.0012721027859051012, "ewc_loss": 5.73345459997654e-09, "ewc_loss_diag": 5.151434834260726e-13, "ewc_loss_parallel": 5.6843418860808015e-12, "grad_norm": 4.65696382522583, "learning_rate": 3.815175922000848e-08, "loss": 1.0422, "mean_token_accuracy": 0.7435706257820129, "num_tokens": 372703.0, "step": 10 }, { "epoch": 0.0013993130644956112, "ewc_loss": 8.440110832452774e-09, "ewc_loss_diag": 7.212008767965017e-13, "ewc_loss_parallel": 8.355982572538778e-12, "grad_norm": 4.196098804473877, "learning_rate": 4.23908435777872e-08, "loss": 0.9769, "mean_token_accuracy": 0.7517222166061401, "num_tokens": 414187.0, "step": 11 }, { "epoch": 0.0015265233430861213, "ewc_loss": 1.1175870895385742e-08, "ewc_loss_diag": 8.952838470577262e-13, "ewc_loss_parallel": 1.1084466677857563e-11, "grad_norm": 4.610984802246094, "learning_rate": 4.662992793556592e-08, "loss": 1.0084, "mean_token_accuracy": 0.7496888041496277, "num_tokens": 446133.0, "step": 12 }, { "epoch": 0.0016537336216766315, "ewc_loss": 3.4924596548080444e-08, "ewc_loss_diag": 3.566924533515703e-12, "ewc_loss_parallel": 3.456079866737127e-11, "grad_norm": 4.781329154968262, "learning_rate": 5.086901229334464e-08, "loss": 1.0517, "mean_token_accuracy": 0.7322452664375305, "num_tokens": 477130.0, "step": 13 }, { "epoch": 0.0017809439002671415, "ewc_loss": 5.145557224750519e-08, "ewc_loss_diag": 4.405364961712621e-12, "ewc_loss_parallel": 5.093170329928398e-11, "grad_norm": 4.263498783111572, "learning_rate": 5.510809665112336e-08, "loss": 1.0084, "mean_token_accuracy": 0.7583171725273132, "num_tokens": 521754.0, "step": 14 }, { "epoch": 0.0019081541788576518, "ewc_loss": 6.705522537231445e-08, "ewc_loss_diag": 5.7411853049416095e-12, "ewc_loss_parallel": 6.639311322942376e-11, "grad_norm": 4.412315368652344, "learning_rate": 5.934718100890208e-08, "loss": 0.9559, "mean_token_accuracy": 0.7667866945266724, "num_tokens": 560311.0, "step": 15 }, { "epoch": 0.002035364457448162, "ewc_loss": 8.475035429000854e-08, "ewc_loss_diag": 6.73594513500575e-12, "ewc_loss_parallel": 8.412825991399586e-11, "grad_norm": 5.255704402923584, "learning_rate": 6.358626536668081e-08, "loss": 1.0333, "mean_token_accuracy": 0.7409570813179016, "num_tokens": 595214.0, "step": 16 }, { "epoch": 0.002162574736038672, "ewc_loss": 1.6763806343078613e-07, "ewc_loss_diag": 1.9326762412674725e-11, "ewc_loss_parallel": 1.6552803572267294e-10, "grad_norm": 4.822842597961426, "learning_rate": 6.782534972445952e-08, "loss": 1.0437, "mean_token_accuracy": 0.7420077919960022, "num_tokens": 631427.0, "step": 17 }, { "epoch": 0.002289785014629182, "ewc_loss": 2.942979335784912e-07, "ewc_loss_diag": 2.887645678129047e-11, "ewc_loss_parallel": 2.9103830456733704e-10, "grad_norm": 4.860193729400635, "learning_rate": 7.206443408223825e-08, "loss": 1.094, "mean_token_accuracy": 0.7348534464836121, "num_tokens": 671155.0, "step": 18 }, { "epoch": 0.0024169952932196924, "ewc_loss": 3.948807716369629e-07, "ewc_loss_diag": 3.319655661471188e-11, "ewc_loss_parallel": 3.9108272176235914e-10, "grad_norm": 4.5098114013671875, "learning_rate": 7.630351844001696e-08, "loss": 0.9982, "mean_token_accuracy": 0.7531075477600098, "num_tokens": 704515.0, "step": 19 }, { "epoch": 0.0025442055718102024, "ewc_loss": 4.637986421585083e-07, "ewc_loss_diag": 3.888089850079268e-11, "ewc_loss_parallel": 4.602043190971017e-10, "grad_norm": 4.475988388061523, "learning_rate": 8.054260279779568e-08, "loss": 0.9738, "mean_token_accuracy": 0.7595764994621277, "num_tokens": 745414.0, "step": 20 }, { "epoch": 0.0026714158504007124, "ewc_loss": 5.513429641723633e-07, "ewc_loss_diag": 4.6838977141305804e-11, "ewc_loss_parallel": 5.493347998708487e-10, "grad_norm": 4.532570838928223, "learning_rate": 8.47816871555744e-08, "loss": 1.0251, "mean_token_accuracy": 0.742141842842102, "num_tokens": 779384.0, "step": 21 }, { "epoch": 0.0027986261289912225, "ewc_loss": 6.295740604400635e-07, "ewc_loss_diag": 5.275069270282984e-11, "ewc_loss_parallel": 6.257323548197746e-10, "grad_norm": 4.447057247161865, "learning_rate": 8.902077151335312e-08, "loss": 1.0131, "mean_token_accuracy": 0.7493085861206055, "num_tokens": 812994.0, "step": 22 }, { "epoch": 0.0029258364075817325, "ewc_loss": 8.977949619293213e-07, "ewc_loss_diag": 1.0868461686186492e-10, "ewc_loss_parallel": 8.87666828930378e-10, "grad_norm": 4.411680698394775, "learning_rate": 9.325985587113183e-08, "loss": 0.9535, "mean_token_accuracy": 0.7668114900588989, "num_tokens": 849351.0, "step": 23 }, { "epoch": 0.0030530466861722425, "ewc_loss": 1.773238182067871e-06, "ewc_loss_diag": 1.9190338207408786e-10, "ewc_loss_parallel": 1.7535057850182056e-09, "grad_norm": 4.682273864746094, "learning_rate": 9.749894022891056e-08, "loss": 0.9975, "mean_token_accuracy": 0.7480342388153076, "num_tokens": 880319.0, "step": 24 }, { "epoch": 0.003180256964762753, "ewc_loss": 2.3692846298217773e-06, "ewc_loss_diag": 2.3283064365386963e-10, "ewc_loss_parallel": 2.342858351767063e-09, "grad_norm": 4.603948593139648, "learning_rate": 1.0173802458668929e-07, "loss": 1.092, "mean_token_accuracy": 0.7259310483932495, "num_tokens": 921902.0, "step": 25 }, { "epoch": 0.003307467243353263, "ewc_loss": 2.950429916381836e-06, "ewc_loss_diag": 2.582964953035116e-10, "ewc_loss_parallel": 2.9249349609017372e-09, "grad_norm": 4.621813774108887, "learning_rate": 1.05977108944468e-07, "loss": 1.0273, "mean_token_accuracy": 0.7426936626434326, "num_tokens": 959531.0, "step": 26 }, { "epoch": 0.003434677521943773, "ewc_loss": 3.3080577850341797e-06, "ewc_loss_diag": 2.8194335754960775e-10, "ewc_loss_parallel": 3.2741809263825417e-09, "grad_norm": 4.123265743255615, "learning_rate": 1.1021619330224672e-07, "loss": 0.9438, "mean_token_accuracy": 0.7642487287521362, "num_tokens": 997280.0, "step": 27 }, { "epoch": 0.003561887800534283, "ewc_loss": 3.7848949432373047e-06, "ewc_loss_diag": 3.346940502524376e-10, "ewc_loss_parallel": 3.754394128918648e-09, "grad_norm": 4.668990135192871, "learning_rate": 1.1445527766002544e-07, "loss": 1.0821, "mean_token_accuracy": 0.7309986352920532, "num_tokens": 1034608.0, "step": 28 }, { "epoch": 0.003689098079124793, "ewc_loss": 4.351139068603516e-06, "ewc_loss_diag": 3.6561687011271715e-10, "ewc_loss_parallel": 4.307366907596588e-09, "grad_norm": 4.229946136474609, "learning_rate": 1.1869436201780416e-07, "loss": 0.964, "mean_token_accuracy": 0.7616429328918457, "num_tokens": 1075907.0, "step": 29 }, { "epoch": 0.0038163083577153036, "ewc_loss": 4.649162292480469e-06, "ewc_loss_diag": 3.9835867937654257e-10, "ewc_loss_parallel": 4.627509042620659e-09, "grad_norm": 6.6350860595703125, "learning_rate": 1.229334463755829e-07, "loss": 1.0139, "mean_token_accuracy": 0.7543025016784668, "num_tokens": 1114492.0, "step": 30 }, { "epoch": 0.003943518636305814, "ewc_loss": 5.0067901611328125e-06, "ewc_loss_diag": 4.31100488640368e-10, "ewc_loss_parallel": 4.976755008101463e-09, "grad_norm": 4.343976020812988, "learning_rate": 1.2717253073336161e-07, "loss": 0.9362, "mean_token_accuracy": 0.7728985548019409, "num_tokens": 1151596.0, "step": 31 }, { "epoch": 0.004070728914896324, "ewc_loss": 6.9141387939453125e-06, "ewc_loss_diag": 8.87666828930378e-10, "ewc_loss_parallel": 6.83940015733242e-09, "grad_norm": 4.852578639984131, "learning_rate": 1.3141161509114031e-07, "loss": 1.0231, "mean_token_accuracy": 0.751845121383667, "num_tokens": 1191475.0, "step": 32 }, { "epoch": 0.004197939193486834, "ewc_loss": 1.1801719665527344e-05, "ewc_loss_diag": 1.3533281162381172e-09, "ewc_loss_parallel": 1.1699739843606949e-08, "grad_norm": 4.299690246582031, "learning_rate": 1.3565069944891904e-07, "loss": 1.0033, "mean_token_accuracy": 0.7538245916366577, "num_tokens": 1230726.0, "step": 33 }, { "epoch": 0.004325149472077344, "ewc_loss": 1.5854835510253906e-05, "ewc_loss_diag": 1.6589183360338211e-09, "ewc_loss_parallel": 1.57160684466362e-08, "grad_norm": 4.195230007171631, "learning_rate": 1.3988978380669777e-07, "loss": 0.9669, "mean_token_accuracy": 0.7631555199623108, "num_tokens": 1273523.0, "step": 34 }, { "epoch": 0.004452359750667854, "ewc_loss": 1.895427703857422e-05, "ewc_loss_diag": 1.8408172763884068e-09, "ewc_loss_parallel": 1.8742866814136505e-08, "grad_norm": 4.262237071990967, "learning_rate": 1.441288681644765e-07, "loss": 0.9694, "mean_token_accuracy": 0.7629618644714355, "num_tokens": 1312013.0, "step": 35 }, { "epoch": 0.004579570029258364, "ewc_loss": 2.1219253540039062e-05, "ewc_loss_diag": 1.979060471057892e-09, "ewc_loss_parallel": 2.0954757928848267e-08, "grad_norm": 4.3858160972595215, "learning_rate": 1.4836795252225522e-07, "loss": 0.9968, "mean_token_accuracy": 0.7516878843307495, "num_tokens": 1350139.0, "step": 36 }, { "epoch": 0.004706780307848874, "ewc_loss": 2.372264862060547e-05, "ewc_loss_diag": 2.0954757928848267e-09, "ewc_loss_parallel": 2.3515895009040833e-08, "grad_norm": 4.260179042816162, "learning_rate": 1.5260703688003392e-07, "loss": 0.9673, "mean_token_accuracy": 0.7587134838104248, "num_tokens": 1387670.0, "step": 37 }, { "epoch": 0.004833990586439385, "ewc_loss": 2.5391578674316406e-05, "ewc_loss_diag": 2.240994945168495e-09, "ewc_loss_parallel": 2.514570951461792e-08, "grad_norm": 4.045468330383301, "learning_rate": 1.5684612123781264e-07, "loss": 0.9724, "mean_token_accuracy": 0.7590323686599731, "num_tokens": 1431658.0, "step": 38 }, { "epoch": 0.004961200865029895, "ewc_loss": 2.753734588623047e-05, "ewc_loss_diag": 2.5029294192790985e-09, "ewc_loss_parallel": 2.7241185307502747e-08, "grad_norm": 4.135424613952637, "learning_rate": 1.6108520559559137e-07, "loss": 0.9046, "mean_token_accuracy": 0.7776740789413452, "num_tokens": 1467751.0, "step": 39 }, { "epoch": 0.005088411143620405, "ewc_loss": 2.9206275939941406e-05, "ewc_loss_diag": 2.7939677238464355e-09, "ewc_loss_parallel": 2.898741513490677e-08, "grad_norm": 4.388813495635986, "learning_rate": 1.6532428995337007e-07, "loss": 1.0039, "mean_token_accuracy": 0.7464303374290466, "num_tokens": 1507078.0, "step": 40 }, { "epoch": 0.005215621422210915, "ewc_loss": 3.24249267578125e-05, "ewc_loss_diag": 2.954038791358471e-09, "ewc_loss_parallel": 3.213062882423401e-08, "grad_norm": 4.281703472137451, "learning_rate": 1.695633743111488e-07, "loss": 0.9903, "mean_token_accuracy": 0.7556573748588562, "num_tokens": 1544856.0, "step": 41 }, { "epoch": 0.005342831700801425, "ewc_loss": 3.4332275390625e-05, "ewc_loss_diag": 3.157765604555607e-09, "ewc_loss_parallel": 3.3993273973464966e-08, "grad_norm": 4.460720539093018, "learning_rate": 1.7380245866892752e-07, "loss": 0.9778, "mean_token_accuracy": 0.7604745626449585, "num_tokens": 1576776.0, "step": 42 }, { "epoch": 0.005470041979391935, "ewc_loss": 3.552436828613281e-05, "ewc_loss_diag": 3.346940502524376e-09, "ewc_loss_parallel": 3.5390257835388184e-08, "grad_norm": 4.14864444732666, "learning_rate": 1.7804154302670624e-07, "loss": 0.9686, "mean_token_accuracy": 0.7592487335205078, "num_tokens": 1613819.0, "step": 43 }, { "epoch": 0.005597252257982445, "ewc_loss": 3.5762786865234375e-05, "ewc_loss_diag": 3.812601789832115e-09, "ewc_loss_parallel": 3.5390257835388184e-08, "grad_norm": 4.512175559997559, "learning_rate": 1.8228062738448497e-07, "loss": 1.0279, "mean_token_accuracy": 0.7475287914276123, "num_tokens": 1654834.0, "step": 44 }, { "epoch": 0.005724462536572955, "ewc_loss": 5.078315734863281e-05, "ewc_loss_diag": 7.450580596923828e-09, "ewc_loss_parallel": 5.005858838558197e-08, "grad_norm": 4.131200790405273, "learning_rate": 1.8651971174226367e-07, "loss": 0.9967, "mean_token_accuracy": 0.7512789964675903, "num_tokens": 1691238.0, "step": 45 }, { "epoch": 0.005851672815163465, "ewc_loss": 7.677078247070312e-05, "ewc_loss_diag": 9.89530235528946e-09, "ewc_loss_parallel": 7.59027898311615e-08, "grad_norm": 3.5285205841064453, "learning_rate": 1.907587961000424e-07, "loss": 0.8845, "mean_token_accuracy": 0.7720202207565308, "num_tokens": 1734446.0, "step": 46 }, { "epoch": 0.005978883093753975, "ewc_loss": 9.584426879882812e-05, "ewc_loss_diag": 1.1699739843606949e-08, "ewc_loss_parallel": 9.499490261077881e-08, "grad_norm": 3.918632984161377, "learning_rate": 1.9499788045782112e-07, "loss": 0.9498, "mean_token_accuracy": 0.7586487531661987, "num_tokens": 1770266.0, "step": 47 }, { "epoch": 0.006106093372344485, "ewc_loss": 0.00011301040649414062, "ewc_loss_diag": 1.2980308383703232e-08, "ewc_loss_parallel": 1.1175870895385742e-07, "grad_norm": 4.012243270874023, "learning_rate": 1.9923696481559985e-07, "loss": 1.0078, "mean_token_accuracy": 0.7411816120147705, "num_tokens": 1807979.0, "step": 48 }, { "epoch": 0.006233303650934996, "ewc_loss": 0.00012874603271484375, "ewc_loss_diag": 1.3853423297405243e-08, "ewc_loss_parallel": 1.2759119272232056e-07, "grad_norm": 3.726867437362671, "learning_rate": 2.0347604917337857e-07, "loss": 0.8979, "mean_token_accuracy": 0.76960289478302, "num_tokens": 1851061.0, "step": 49 }, { "epoch": 0.006360513929525506, "ewc_loss": 0.000141143798828125, "ewc_loss_diag": 1.4842953532934189e-08, "ewc_loss_parallel": 1.3969838619232178e-07, "grad_norm": 3.4986367225646973, "learning_rate": 2.0771513353115727e-07, "loss": 0.854, "mean_token_accuracy": 0.7844707369804382, "num_tokens": 1887647.0, "step": 50 }, { "epoch": 0.006487724208116016, "ewc_loss": 0.000148773193359375, "ewc_loss_diag": 1.5599653124809265e-08, "ewc_loss_parallel": 1.471489667892456e-07, "grad_norm": 4.123988628387451, "learning_rate": 2.11954217888936e-07, "loss": 1.0201, "mean_token_accuracy": 0.743654727935791, "num_tokens": 1924843.0, "step": 51 }, { "epoch": 0.006614934486706526, "ewc_loss": 0.000152587890625, "ewc_loss_diag": 1.594889909029007e-08, "ewc_loss_parallel": 1.5087425708770752e-07, "grad_norm": 4.107985019683838, "learning_rate": 2.1619330224671472e-07, "loss": 0.9464, "mean_token_accuracy": 0.7627004384994507, "num_tokens": 1967116.0, "step": 52 }, { "epoch": 0.006742144765297036, "ewc_loss": 0.000156402587890625, "ewc_loss_diag": 1.6763806343078613e-08, "ewc_loss_parallel": 1.5459954738616943e-07, "grad_norm": 3.7296180725097656, "learning_rate": 2.2043238660449345e-07, "loss": 0.9282, "mean_token_accuracy": 0.7578206658363342, "num_tokens": 1999575.0, "step": 53 }, { "epoch": 0.006869355043887546, "ewc_loss": 0.00015735626220703125, "ewc_loss_diag": 1.8277205526828766e-08, "ewc_loss_parallel": 1.555308699607849e-07, "grad_norm": 4.108229637145996, "learning_rate": 2.2467147096227215e-07, "loss": 1.0157, "mean_token_accuracy": 0.7412534356117249, "num_tokens": 2039396.0, "step": 54 }, { "epoch": 0.006996565322478056, "ewc_loss": 0.00015926361083984375, "ewc_loss_diag": 2.0023435354232788e-08, "ewc_loss_parallel": 1.5739351511001587e-07, "grad_norm": 4.028424263000488, "learning_rate": 2.2891055532005087e-07, "loss": 1.0472, "mean_token_accuracy": 0.7314740419387817, "num_tokens": 2080258.0, "step": 55 }, { "epoch": 0.007123775601068566, "ewc_loss": 0.00016307830810546875, "ewc_loss_diag": 2.130400389432907e-08, "ewc_loss_parallel": 1.6111880540847778e-07, "grad_norm": 3.9688925743103027, "learning_rate": 2.331496396778296e-07, "loss": 1.0242, "mean_token_accuracy": 0.7413282990455627, "num_tokens": 2116552.0, "step": 56 }, { "epoch": 0.007250985879659076, "ewc_loss": 0.0001659393310546875, "ewc_loss_diag": 2.223532646894455e-08, "ewc_loss_parallel": 1.6391277313232422e-07, "grad_norm": 3.703815221786499, "learning_rate": 2.3738872403560833e-07, "loss": 0.9224, "mean_token_accuracy": 0.7654881477355957, "num_tokens": 2161097.0, "step": 57 }, { "epoch": 0.007378196158249586, "ewc_loss": 0.00016689300537109375, "ewc_loss_diag": 2.293381839990616e-08, "ewc_loss_parallel": 1.648440957069397e-07, "grad_norm": 3.545050621032715, "learning_rate": 2.4162780839338705e-07, "loss": 0.8788, "mean_token_accuracy": 0.7753727436065674, "num_tokens": 2201190.0, "step": 58 }, { "epoch": 0.007505406436840096, "ewc_loss": 0.00017261505126953125, "ewc_loss_diag": 2.3981556296348572e-08, "ewc_loss_parallel": 1.695007085800171e-07, "grad_norm": 4.0433855056762695, "learning_rate": 2.458668927511658e-07, "loss": 1.0164, "mean_token_accuracy": 0.7402371764183044, "num_tokens": 2233725.0, "step": 59 }, { "epoch": 0.007632616715430607, "ewc_loss": 0.0001735687255859375, "ewc_loss_diag": 2.5262124836444855e-08, "ewc_loss_parallel": 1.7043203115463257e-07, "grad_norm": 3.78566575050354, "learning_rate": 2.501059771089445e-07, "loss": 0.9633, "mean_token_accuracy": 0.7512965202331543, "num_tokens": 2274459.0, "step": 60 }, { "epoch": 0.007759826994021117, "ewc_loss": 0.00017642974853515625, "ewc_loss_diag": 2.7124769985675812e-08, "ewc_loss_parallel": 1.73225998878479e-07, "grad_norm": 3.6620283126831055, "learning_rate": 2.5434506146672323e-07, "loss": 0.9116, "mean_token_accuracy": 0.7657475471496582, "num_tokens": 2317530.0, "step": 61 }, { "epoch": 0.007887037272611627, "ewc_loss": 0.0002346038818359375, "ewc_loss_diag": 3.632158041000366e-08, "ewc_loss_parallel": 2.3096799850463867e-07, "grad_norm": 3.6515870094299316, "learning_rate": 2.5858414582450195e-07, "loss": 1.0012, "mean_token_accuracy": 0.7472654581069946, "num_tokens": 2355260.0, "step": 62 }, { "epoch": 0.008014247551202136, "ewc_loss": 0.0003376007080078125, "ewc_loss_diag": 5.098991096019745e-08, "ewc_loss_parallel": 3.3155083656311035e-07, "grad_norm": 3.1433398723602295, "learning_rate": 2.6282323018228063e-07, "loss": 0.8849, "mean_token_accuracy": 0.7709324359893799, "num_tokens": 2397135.0, "step": 63 }, { "epoch": 0.008141457829792647, "ewc_loss": 0.00042724609375, "ewc_loss_diag": 6.239861249923706e-08, "ewc_loss_parallel": 4.209578037261963e-07, "grad_norm": 3.134906530380249, "learning_rate": 2.6706231454005935e-07, "loss": 0.9236, "mean_token_accuracy": 0.7585223913192749, "num_tokens": 2440206.0, "step": 64 }, { "epoch": 0.008268668108383158, "ewc_loss": 0.000514984130859375, "ewc_loss_diag": 6.891787052154541e-08, "ewc_loss_parallel": 5.066394805908203e-07, "grad_norm": 3.418057441711426, "learning_rate": 2.713013988978381e-07, "loss": 0.9613, "mean_token_accuracy": 0.7411068677902222, "num_tokens": 2469565.0, "step": 65 }, { "epoch": 0.008395878386973667, "ewc_loss": 0.00057220458984375, "ewc_loss_diag": 7.636845111846924e-08, "ewc_loss_parallel": 5.662441253662109e-07, "grad_norm": 3.1878321170806885, "learning_rate": 2.755404832556168e-07, "loss": 0.8907, "mean_token_accuracy": 0.7624822854995728, "num_tokens": 2514314.0, "step": 66 }, { "epoch": 0.008523088665564178, "ewc_loss": 0.000598907470703125, "ewc_loss_diag": 8.288770914077759e-08, "ewc_loss_parallel": 5.923211574554443e-07, "grad_norm": 3.073631525039673, "learning_rate": 2.7977956761339553e-07, "loss": 0.934, "mean_token_accuracy": 0.7513231635093689, "num_tokens": 2558466.0, "step": 67 }, { "epoch": 0.008650298944154687, "ewc_loss": 0.00061798095703125, "ewc_loss_diag": 8.847564458847046e-08, "ewc_loss_parallel": 6.109476089477539e-07, "grad_norm": 3.0598981380462646, "learning_rate": 2.8401865197117426e-07, "loss": 0.8716, "mean_token_accuracy": 0.7630219459533691, "num_tokens": 2596843.0, "step": 68 }, { "epoch": 0.008777509222745198, "ewc_loss": 0.00063323974609375, "ewc_loss_diag": 9.313225746154785e-08, "ewc_loss_parallel": 6.258487701416016e-07, "grad_norm": 3.7304868698120117, "learning_rate": 2.88257736328953e-07, "loss": 0.8597, "mean_token_accuracy": 0.7663498520851135, "num_tokens": 2632933.0, "step": 69 }, { "epoch": 0.008904719501335707, "ewc_loss": 0.000637054443359375, "ewc_loss_diag": 9.872019290924072e-08, "ewc_loss_parallel": 6.258487701416016e-07, "grad_norm": 2.8619351387023926, "learning_rate": 2.924968206867317e-07, "loss": 0.8415, "mean_token_accuracy": 0.7802233099937439, "num_tokens": 2673526.0, "step": 70 }, { "epoch": 0.009031929779926218, "ewc_loss": 0.000621795654296875, "ewc_loss_diag": 1.0337680578231812e-07, "ewc_loss_parallel": 6.109476089477539e-07, "grad_norm": 3.0864057540893555, "learning_rate": 2.9673590504451043e-07, "loss": 0.8824, "mean_token_accuracy": 0.7649776935577393, "num_tokens": 2707770.0, "step": 71 }, { "epoch": 0.009159140058516728, "ewc_loss": 0.000614166259765625, "ewc_loss_diag": 1.0570511221885681e-07, "ewc_loss_parallel": 6.034970283508301e-07, "grad_norm": 2.8440194129943848, "learning_rate": 3.009749894022891e-07, "loss": 0.844, "mean_token_accuracy": 0.7763500213623047, "num_tokens": 2745236.0, "step": 72 }, { "epoch": 0.009286350337107238, "ewc_loss": 0.0005950927734375, "ewc_loss_diag": 1.103617250919342e-07, "ewc_loss_parallel": 5.848705768585205e-07, "grad_norm": 2.974625587463379, "learning_rate": 3.0521407376006783e-07, "loss": 0.9197, "mean_token_accuracy": 0.7539865970611572, "num_tokens": 2780291.0, "step": 73 }, { "epoch": 0.009413560615697748, "ewc_loss": 0.000583648681640625, "ewc_loss_diag": 1.1362135410308838e-07, "ewc_loss_parallel": 5.736947059631348e-07, "grad_norm": 3.010233163833618, "learning_rate": 3.0945315811784656e-07, "loss": 0.9706, "mean_token_accuracy": 0.7466590404510498, "num_tokens": 2814771.0, "step": 74 }, { "epoch": 0.009540770894288259, "ewc_loss": 0.00057220458984375, "ewc_loss_diag": 1.1920928955078125e-07, "ewc_loss_parallel": 5.62518835067749e-07, "grad_norm": 2.731905698776245, "learning_rate": 3.136922424756253e-07, "loss": 0.8447, "mean_token_accuracy": 0.7742918729782104, "num_tokens": 2861375.0, "step": 75 }, { "epoch": 0.00966798117287877, "ewc_loss": 0.00055694580078125, "ewc_loss_diag": 1.2479722499847412e-07, "ewc_loss_parallel": 5.438923835754395e-07, "grad_norm": 2.6157805919647217, "learning_rate": 3.17931326833404e-07, "loss": 0.8393, "mean_token_accuracy": 0.7706247568130493, "num_tokens": 2896139.0, "step": 76 }, { "epoch": 0.009795191451469279, "ewc_loss": 0.00054168701171875, "ewc_loss_diag": 1.2945383787155151e-07, "ewc_loss_parallel": 5.289912223815918e-07, "grad_norm": 2.5611186027526855, "learning_rate": 3.2217041119118274e-07, "loss": 0.7982, "mean_token_accuracy": 0.785552442073822, "num_tokens": 2940269.0, "step": 77 }, { "epoch": 0.00992240173005979, "ewc_loss": 0.000522613525390625, "ewc_loss_diag": 1.3317912817001343e-07, "ewc_loss_parallel": 5.103647708892822e-07, "grad_norm": 2.761705160140991, "learning_rate": 3.2640949554896146e-07, "loss": 0.8471, "mean_token_accuracy": 0.7705661654472351, "num_tokens": 2971115.0, "step": 78 }, { "epoch": 0.010049612008650299, "ewc_loss": 0.00051116943359375, "ewc_loss_diag": 1.387670636177063e-07, "ewc_loss_parallel": 4.954636096954346e-07, "grad_norm": 2.4041247367858887, "learning_rate": 3.3064857990674013e-07, "loss": 0.7409, "mean_token_accuracy": 0.7952703237533569, "num_tokens": 3003727.0, "step": 79 }, { "epoch": 0.01017682228724081, "ewc_loss": 0.0005035400390625, "ewc_loss_diag": 1.4435499906539917e-07, "ewc_loss_parallel": 4.880130290985107e-07, "grad_norm": 2.5245933532714844, "learning_rate": 3.3488766426451886e-07, "loss": 0.7795, "mean_token_accuracy": 0.788129448890686, "num_tokens": 3045557.0, "step": 80 }, { "epoch": 0.010304032565831319, "ewc_loss": 0.00048828125, "ewc_loss_diag": 1.471489667892456e-07, "ewc_loss_parallel": 4.7497451305389404e-07, "grad_norm": 3.010434150695801, "learning_rate": 3.391267486222976e-07, "loss": 0.8489, "mean_token_accuracy": 0.7654989957809448, "num_tokens": 3080478.0, "step": 81 }, { "epoch": 0.01043124284442183, "ewc_loss": 0.000499725341796875, "ewc_loss_diag": 1.555308699607849e-07, "ewc_loss_parallel": 4.842877388000488e-07, "grad_norm": 2.496990203857422, "learning_rate": 3.433658329800763e-07, "loss": 0.8734, "mean_token_accuracy": 0.7627073526382446, "num_tokens": 3119205.0, "step": 82 }, { "epoch": 0.010558453123012339, "ewc_loss": 0.000522613525390625, "ewc_loss_diag": 1.6391277313232422e-07, "ewc_loss_parallel": 5.066394805908203e-07, "grad_norm": 2.831355094909668, "learning_rate": 3.4760491733785504e-07, "loss": 0.895, "mean_token_accuracy": 0.7556256055831909, "num_tokens": 3151816.0, "step": 83 }, { "epoch": 0.01068566340160285, "ewc_loss": 0.0006256103515625, "ewc_loss_diag": 1.7415732145309448e-07, "ewc_loss_parallel": 6.07222318649292e-07, "grad_norm": 2.6519296169281006, "learning_rate": 3.5184400169563376e-07, "loss": 0.8642, "mean_token_accuracy": 0.7615691423416138, "num_tokens": 3189765.0, "step": 84 }, { "epoch": 0.010812873680193359, "ewc_loss": 0.000797271728515625, "ewc_loss_diag": 1.8905848264694214e-07, "ewc_loss_parallel": 7.7858567237854e-07, "grad_norm": 2.4210314750671387, "learning_rate": 3.560830860534125e-07, "loss": 0.8789, "mean_token_accuracy": 0.7594888210296631, "num_tokens": 3228988.0, "step": 85 }, { "epoch": 0.01094008395878387, "ewc_loss": 0.0010528564453125, "ewc_loss_diag": 2.0582228899002075e-07, "ewc_loss_parallel": 1.0281801223754883e-06, "grad_norm": 2.507937431335449, "learning_rate": 3.603221704111912e-07, "loss": 0.8873, "mean_token_accuracy": 0.7570308446884155, "num_tokens": 3270648.0, "step": 86 }, { "epoch": 0.01106729423737438, "ewc_loss": 0.00128936767578125, "ewc_loss_diag": 2.3096799850463867e-07, "ewc_loss_parallel": 1.2665987014770508e-06, "grad_norm": 2.1864023208618164, "learning_rate": 3.6456125476896994e-07, "loss": 0.8332, "mean_token_accuracy": 0.7717821598052979, "num_tokens": 3312134.0, "step": 87 }, { "epoch": 0.01119450451596489, "ewc_loss": 0.001495361328125, "ewc_loss_diag": 2.5331974029541016e-07, "ewc_loss_parallel": 1.475214958190918e-06, "grad_norm": 2.242835760116577, "learning_rate": 3.688003391267486e-07, "loss": 0.8399, "mean_token_accuracy": 0.7698149085044861, "num_tokens": 3351904.0, "step": 88 }, { "epoch": 0.0113217147945554, "ewc_loss": 0.0016632080078125, "ewc_loss_diag": 2.6635825634002686e-07, "ewc_loss_parallel": 1.6391277313232422e-06, "grad_norm": 2.0374162197113037, "learning_rate": 3.7303942348452734e-07, "loss": 0.7902, "mean_token_accuracy": 0.7812255620956421, "num_tokens": 3389610.0, "step": 89 }, { "epoch": 0.01144892507314591, "ewc_loss": 0.0017852783203125, "ewc_loss_diag": 2.905726432800293e-07, "ewc_loss_parallel": 1.7583370208740234e-06, "grad_norm": 2.002359390258789, "learning_rate": 3.7727850784230606e-07, "loss": 0.8632, "mean_token_accuracy": 0.755573034286499, "num_tokens": 3425779.0, "step": 90 }, { "epoch": 0.01157613535173642, "ewc_loss": 0.0018157958984375, "ewc_loss_diag": 3.129243850708008e-07, "ewc_loss_parallel": 1.7881393432617188e-06, "grad_norm": 1.984840750694275, "learning_rate": 3.815175922000848e-07, "loss": 0.8378, "mean_token_accuracy": 0.7703469395637512, "num_tokens": 3463938.0, "step": 91 }, { "epoch": 0.01170334563032693, "ewc_loss": 0.00188446044921875, "ewc_loss_diag": 3.390014171600342e-07, "ewc_loss_parallel": 1.8551945686340332e-06, "grad_norm": 1.8219252824783325, "learning_rate": 3.857566765578635e-07, "loss": 0.7979, "mean_token_accuracy": 0.7801293134689331, "num_tokens": 3500494.0, "step": 92 }, { "epoch": 0.01183055590891744, "ewc_loss": 0.00191497802734375, "ewc_loss_diag": 3.632158041000366e-07, "ewc_loss_parallel": 1.8775463104248047e-06, "grad_norm": 1.7617970705032349, "learning_rate": 3.8999576091564224e-07, "loss": 0.7588, "mean_token_accuracy": 0.7862566113471985, "num_tokens": 3540901.0, "step": 93 }, { "epoch": 0.01195776618750795, "ewc_loss": 0.001861572265625, "ewc_loss_diag": 3.8743019104003906e-07, "ewc_loss_parallel": 1.8253922462463379e-06, "grad_norm": 1.6386908292770386, "learning_rate": 3.9423484527342097e-07, "loss": 0.7863, "mean_token_accuracy": 0.7773581147193909, "num_tokens": 3579275.0, "step": 94 }, { "epoch": 0.012084976466098461, "ewc_loss": 0.0017852783203125, "ewc_loss_diag": 4.023313522338867e-07, "ewc_loss_parallel": 1.7508864402770996e-06, "grad_norm": 1.6090322732925415, "learning_rate": 3.984739296311997e-07, "loss": 0.7558, "mean_token_accuracy": 0.7832346558570862, "num_tokens": 3614755.0, "step": 95 }, { "epoch": 0.01221218674468897, "ewc_loss": 0.0017242431640625, "ewc_loss_diag": 4.2654573917388916e-07, "ewc_loss_parallel": 1.6763806343078613e-06, "grad_norm": 1.678248405456543, "learning_rate": 4.027130139889784e-07, "loss": 0.8098, "mean_token_accuracy": 0.7684808373451233, "num_tokens": 3655744.0, "step": 96 }, { "epoch": 0.012339397023279481, "ewc_loss": 0.00168609619140625, "ewc_loss_diag": 4.4889748096466064e-07, "ewc_loss_parallel": 1.6391277313232422e-06, "grad_norm": 2.1375699043273926, "learning_rate": 4.0695209834675715e-07, "loss": 0.7919, "mean_token_accuracy": 0.7798059582710266, "num_tokens": 3697257.0, "step": 97 }, { "epoch": 0.012466607301869992, "ewc_loss": 0.0017547607421875, "ewc_loss_diag": 4.675239324569702e-07, "ewc_loss_parallel": 1.7061829566955566e-06, "grad_norm": 1.5420875549316406, "learning_rate": 4.111911827045358e-07, "loss": 0.7441, "mean_token_accuracy": 0.7918045520782471, "num_tokens": 3735737.0, "step": 98 }, { "epoch": 0.012593817580460501, "ewc_loss": 0.00165557861328125, "ewc_loss_diag": 4.880130290985107e-07, "ewc_loss_parallel": 1.6093254089355469e-06, "grad_norm": 1.5383129119873047, "learning_rate": 4.1543026706231454e-07, "loss": 0.7576, "mean_token_accuracy": 0.7867245674133301, "num_tokens": 3774212.0, "step": 99 }, { "epoch": 0.012721027859051012, "ewc_loss": 0.001556396484375, "ewc_loss_diag": 4.954636096954346e-07, "ewc_loss_parallel": 1.5050172805786133e-06, "grad_norm": 1.5595537424087524, "learning_rate": 4.1966935142009327e-07, "loss": 0.7461, "mean_token_accuracy": 0.7874270677566528, "num_tokens": 3807640.0, "step": 100 }, { "epoch": 0.012848238137641521, "ewc_loss": 0.0014801025390625, "ewc_loss_diag": 5.103647708892822e-07, "ewc_loss_parallel": 1.4230608940124512e-06, "grad_norm": 1.5036333799362183, "learning_rate": 4.23908435777872e-07, "loss": 0.7758, "mean_token_accuracy": 0.7826613187789917, "num_tokens": 3848908.0, "step": 101 }, { "epoch": 0.012975448416232032, "ewc_loss": 0.0014495849609375, "ewc_loss_diag": 5.178153514862061e-07, "ewc_loss_parallel": 1.3932585716247559e-06, "grad_norm": 1.5997337102890015, "learning_rate": 4.281475201356507e-07, "loss": 0.7844, "mean_token_accuracy": 0.7757166624069214, "num_tokens": 3886341.0, "step": 102 }, { "epoch": 0.013102658694822541, "ewc_loss": 0.00139617919921875, "ewc_loss_diag": 5.252659320831299e-07, "ewc_loss_parallel": 1.341104507446289e-06, "grad_norm": 1.4045037031173706, "learning_rate": 4.3238660449342945e-07, "loss": 0.662, "mean_token_accuracy": 0.8071210980415344, "num_tokens": 3917111.0, "step": 103 }, { "epoch": 0.013229868973413052, "ewc_loss": 0.00131988525390625, "ewc_loss_diag": 5.289912223815918e-07, "ewc_loss_parallel": 1.2665987014770508e-06, "grad_norm": 1.378424882888794, "learning_rate": 4.3662568885120817e-07, "loss": 0.7446, "mean_token_accuracy": 0.7875615954399109, "num_tokens": 3959642.0, "step": 104 }, { "epoch": 0.013357079252003561, "ewc_loss": 0.0012664794921875, "ewc_loss_diag": 5.438923835754395e-07, "ewc_loss_parallel": 1.214444637298584e-06, "grad_norm": 1.4055241346359253, "learning_rate": 4.408647732089869e-07, "loss": 0.7686, "mean_token_accuracy": 0.7824387550354004, "num_tokens": 4001028.0, "step": 105 }, { "epoch": 0.013484289530594072, "ewc_loss": 0.001220703125, "ewc_loss_diag": 5.587935447692871e-07, "ewc_loss_parallel": 1.169741153717041e-06, "grad_norm": 1.420809030532837, "learning_rate": 4.451038575667656e-07, "loss": 0.7787, "mean_token_accuracy": 0.7752779722213745, "num_tokens": 4045249.0, "step": 106 }, { "epoch": 0.013611499809184581, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 5.736947059631348e-07, "ewc_loss_parallel": 1.1473894119262695e-06, "grad_norm": 1.3764938116073608, "learning_rate": 4.493429419245443e-07, "loss": 0.7807, "mean_token_accuracy": 0.7788271903991699, "num_tokens": 4082369.0, "step": 107 }, { "epoch": 0.013738710087775092, "ewc_loss": 0.001190185546875, "ewc_loss_diag": 5.811452865600586e-07, "ewc_loss_parallel": 1.1324882507324219e-06, "grad_norm": 1.387166976928711, "learning_rate": 4.53582026282323e-07, "loss": 0.763, "mean_token_accuracy": 0.7792860269546509, "num_tokens": 4117448.0, "step": 108 }, { "epoch": 0.013865920366365603, "ewc_loss": 0.0011749267578125, "ewc_loss_diag": 5.997717380523682e-07, "ewc_loss_parallel": 1.1175870895385742e-06, "grad_norm": 1.2837514877319336, "learning_rate": 4.5782111064010175e-07, "loss": 0.7016, "mean_token_accuracy": 0.8009606599807739, "num_tokens": 4159132.0, "step": 109 }, { "epoch": 0.013993130644956112, "ewc_loss": 0.0011749267578125, "ewc_loss_diag": 6.183981895446777e-07, "ewc_loss_parallel": 1.1175870895385742e-06, "grad_norm": 1.2861028909683228, "learning_rate": 4.620601949978805e-07, "loss": 0.7347, "mean_token_accuracy": 0.7852941751480103, "num_tokens": 4199800.0, "step": 110 }, { "epoch": 0.014120340923546623, "ewc_loss": 0.00118255615234375, "ewc_loss_diag": 6.332993507385254e-07, "ewc_loss_parallel": 1.125037670135498e-06, "grad_norm": 1.423706293106079, "learning_rate": 4.662992793556592e-07, "loss": 0.7396, "mean_token_accuracy": 0.7921050786972046, "num_tokens": 4244513.0, "step": 111 }, { "epoch": 0.014247551202137132, "ewc_loss": 0.0012054443359375, "ewc_loss_diag": 6.51925802230835e-07, "ewc_loss_parallel": 1.1399388313293457e-06, "grad_norm": 1.284399390220642, "learning_rate": 4.705383637134379e-07, "loss": 0.6836, "mean_token_accuracy": 0.8015503287315369, "num_tokens": 4282858.0, "step": 112 }, { "epoch": 0.014374761480727643, "ewc_loss": 0.0012359619140625, "ewc_loss_diag": 6.668269634246826e-07, "ewc_loss_parallel": 1.169741153717041e-06, "grad_norm": 1.2365251779556274, "learning_rate": 4.7477744807121665e-07, "loss": 0.734, "mean_token_accuracy": 0.7882151007652283, "num_tokens": 4325092.0, "step": 113 }, { "epoch": 0.014501971759318152, "ewc_loss": 0.00122833251953125, "ewc_loss_diag": 6.854534149169922e-07, "ewc_loss_parallel": 1.1622905731201172e-06, "grad_norm": 1.2000197172164917, "learning_rate": 4.790165324289954e-07, "loss": 0.6665, "mean_token_accuracy": 0.8061025142669678, "num_tokens": 4363459.0, "step": 114 }, { "epoch": 0.014629182037908663, "ewc_loss": 0.001251220703125, "ewc_loss_diag": 7.115304470062256e-07, "ewc_loss_parallel": 1.1846423149108887e-06, "grad_norm": 1.1916821002960205, "learning_rate": 4.832556167867741e-07, "loss": 0.7405, "mean_token_accuracy": 0.7883434295654297, "num_tokens": 4400065.0, "step": 115 }, { "epoch": 0.014756392316499172, "ewc_loss": 0.00128936767578125, "ewc_loss_diag": 7.37607479095459e-07, "ewc_loss_parallel": 1.214444637298584e-06, "grad_norm": 1.1963194608688354, "learning_rate": 4.874947011445528e-07, "loss": 0.7245, "mean_token_accuracy": 0.7902055978775024, "num_tokens": 4437828.0, "step": 116 }, { "epoch": 0.014883602595089683, "ewc_loss": 0.0013427734375, "ewc_loss_diag": 7.562339305877686e-07, "ewc_loss_parallel": 1.2665987014770508e-06, "grad_norm": 1.2471942901611328, "learning_rate": 4.917337855023316e-07, "loss": 0.6929, "mean_token_accuracy": 0.7970073819160461, "num_tokens": 4475556.0, "step": 117 }, { "epoch": 0.015010812873680193, "ewc_loss": 0.00140380859375, "ewc_loss_diag": 7.7858567237854e-07, "ewc_loss_parallel": 1.3262033462524414e-06, "grad_norm": 1.2073583602905273, "learning_rate": 4.959728698601102e-07, "loss": 0.6949, "mean_token_accuracy": 0.8012078404426575, "num_tokens": 4512087.0, "step": 118 }, { "epoch": 0.015138023152270703, "ewc_loss": 0.0014801025390625, "ewc_loss_diag": 8.121132850646973e-07, "ewc_loss_parallel": 1.3932585716247559e-06, "grad_norm": 1.1331896781921387, "learning_rate": 5.00211954217889e-07, "loss": 0.665, "mean_token_accuracy": 0.8037964105606079, "num_tokens": 4552116.0, "step": 119 }, { "epoch": 0.015265233430861214, "ewc_loss": 0.001556396484375, "ewc_loss_diag": 8.419156074523926e-07, "ewc_loss_parallel": 1.475214958190918e-06, "grad_norm": 1.1542843580245972, "learning_rate": 5.044510385756677e-07, "loss": 0.7103, "mean_token_accuracy": 0.7951627373695374, "num_tokens": 4591723.0, "step": 120 }, { "epoch": 0.015392443709451724, "ewc_loss": 0.00159454345703125, "ewc_loss_diag": 8.605420589447021e-07, "ewc_loss_parallel": 1.5124678611755371e-06, "grad_norm": 1.1025251150131226, "learning_rate": 5.086901229334465e-07, "loss": 0.6366, "mean_token_accuracy": 0.8156907558441162, "num_tokens": 4635800.0, "step": 121 }, { "epoch": 0.015519653988042234, "ewc_loss": 0.00167083740234375, "ewc_loss_diag": 9.015202522277832e-07, "ewc_loss_parallel": 1.5795230865478516e-06, "grad_norm": 1.0853596925735474, "learning_rate": 5.129292072912251e-07, "loss": 0.7036, "mean_token_accuracy": 0.7964406609535217, "num_tokens": 4679908.0, "step": 122 }, { "epoch": 0.015646864266632744, "ewc_loss": 0.0017547607421875, "ewc_loss_diag": 9.275972843170166e-07, "ewc_loss_parallel": 1.6614794731140137e-06, "grad_norm": 1.094588041305542, "learning_rate": 5.171682916490039e-07, "loss": 0.6849, "mean_token_accuracy": 0.796344518661499, "num_tokens": 4720120.0, "step": 123 }, { "epoch": 0.015774074545223254, "ewc_loss": 0.0018463134765625, "ewc_loss_diag": 9.685754776000977e-07, "ewc_loss_parallel": 1.7508864402770996e-06, "grad_norm": 1.1007591485977173, "learning_rate": 5.214073760067826e-07, "loss": 0.7074, "mean_token_accuracy": 0.7859645485877991, "num_tokens": 4758777.0, "step": 124 }, { "epoch": 0.015901284823813765, "ewc_loss": 0.00191497802734375, "ewc_loss_diag": 1.0132789611816406e-06, "ewc_loss_parallel": 1.817941665649414e-06, "grad_norm": 1.1259013414382935, "learning_rate": 5.256464603645613e-07, "loss": 0.6682, "mean_token_accuracy": 0.8047929406166077, "num_tokens": 4794781.0, "step": 125 }, { "epoch": 0.016028495102404273, "ewc_loss": 0.001983642578125, "ewc_loss_diag": 1.0207295417785645e-06, "ewc_loss_parallel": 1.8775463104248047e-06, "grad_norm": 1.0572633743286133, "learning_rate": 5.2988554472234e-07, "loss": 0.7156, "mean_token_accuracy": 0.788750946521759, "num_tokens": 4836391.0, "step": 126 }, { "epoch": 0.016155705380994784, "ewc_loss": 0.0020599365234375, "ewc_loss_diag": 1.043081283569336e-06, "ewc_loss_parallel": 1.952052116394043e-06, "grad_norm": 1.1018173694610596, "learning_rate": 5.341246290801187e-07, "loss": 0.6877, "mean_token_accuracy": 0.7983573079109192, "num_tokens": 4874193.0, "step": 127 }, { "epoch": 0.016282915659585295, "ewc_loss": 0.00213623046875, "ewc_loss_diag": 1.0579824447631836e-06, "ewc_loss_parallel": 2.0265579223632812e-06, "grad_norm": 1.0978760719299316, "learning_rate": 5.383637134378975e-07, "loss": 0.6947, "mean_token_accuracy": 0.7925527095794678, "num_tokens": 4909311.0, "step": 128 }, { "epoch": 0.016410125938175806, "ewc_loss": 0.002197265625, "ewc_loss_diag": 1.0579824447631836e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 1.063194990158081, "learning_rate": 5.426027977956762e-07, "loss": 0.6848, "mean_token_accuracy": 0.800143837928772, "num_tokens": 4947409.0, "step": 129 }, { "epoch": 0.016537336216766316, "ewc_loss": 0.0022430419921875, "ewc_loss_diag": 1.0728836059570312e-06, "ewc_loss_parallel": 2.130866050720215e-06, "grad_norm": 1.0690549612045288, "learning_rate": 5.468418821534549e-07, "loss": 0.688, "mean_token_accuracy": 0.7948787808418274, "num_tokens": 4982792.0, "step": 130 }, { "epoch": 0.016664546495356824, "ewc_loss": 0.002288818359375, "ewc_loss_diag": 1.1026859283447266e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 1.0176608562469482, "learning_rate": 5.510809665112336e-07, "loss": 0.6593, "mean_token_accuracy": 0.8027827739715576, "num_tokens": 5020795.0, "step": 131 }, { "epoch": 0.016791756773947335, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 1.1175870895385742e-06, "ewc_loss_parallel": 2.1904706954956055e-06, "grad_norm": 1.0435597896575928, "learning_rate": 5.553200508690124e-07, "loss": 0.6973, "mean_token_accuracy": 0.7934916615486145, "num_tokens": 5061519.0, "step": 132 }, { "epoch": 0.016918967052537846, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 1.1399388313293457e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 1.0798022747039795, "learning_rate": 5.595591352267911e-07, "loss": 0.6929, "mean_token_accuracy": 0.7925995588302612, "num_tokens": 5094608.0, "step": 133 }, { "epoch": 0.017046177331128357, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 1.169741153717041e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 1.0539876222610474, "learning_rate": 5.637982195845697e-07, "loss": 0.6936, "mean_token_accuracy": 0.7966541051864624, "num_tokens": 5131130.0, "step": 134 }, { "epoch": 0.017173387609718864, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 1.214444637298584e-06, "ewc_loss_parallel": 2.2798776626586914e-06, "grad_norm": 1.0206176042556763, "learning_rate": 5.680373039423485e-07, "loss": 0.6765, "mean_token_accuracy": 0.8016785383224487, "num_tokens": 5168921.0, "step": 135 }, { "epoch": 0.017300597888309375, "ewc_loss": 0.002349853515625, "ewc_loss_diag": 1.2293457984924316e-06, "ewc_loss_parallel": 2.2351741790771484e-06, "grad_norm": 1.0362852811813354, "learning_rate": 5.722763883001272e-07, "loss": 0.6737, "mean_token_accuracy": 0.7985043525695801, "num_tokens": 5206417.0, "step": 136 }, { "epoch": 0.017427808166899886, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 1.2665987014770508e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 1.066263198852539, "learning_rate": 5.76515472657906e-07, "loss": 0.7174, "mean_token_accuracy": 0.7870470285415649, "num_tokens": 5239984.0, "step": 137 }, { "epoch": 0.017555018445490397, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 1.296401023864746e-06, "ewc_loss_parallel": 2.2798776626586914e-06, "grad_norm": 1.0248076915740967, "learning_rate": 5.807545570156846e-07, "loss": 0.6905, "mean_token_accuracy": 0.7967191338539124, "num_tokens": 5276444.0, "step": 138 }, { "epoch": 0.017682228724080904, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 1.3187527656555176e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 0.9847487211227417, "learning_rate": 5.849936413734634e-07, "loss": 0.6199, "mean_token_accuracy": 0.8162010908126831, "num_tokens": 5312064.0, "step": 139 }, { "epoch": 0.017809439002671415, "ewc_loss": 0.0023651123046875, "ewc_loss_diag": 1.3560056686401367e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 1.0685330629348755, "learning_rate": 5.892327257312421e-07, "loss": 0.7079, "mean_token_accuracy": 0.7960992455482483, "num_tokens": 5350405.0, "step": 140 }, { "epoch": 0.017936649281261926, "ewc_loss": 0.0023956298828125, "ewc_loss_diag": 1.3783574104309082e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 0.9996314644813538, "learning_rate": 5.934718100890209e-07, "loss": 0.6573, "mean_token_accuracy": 0.806458592414856, "num_tokens": 5388338.0, "step": 141 }, { "epoch": 0.018063859559852437, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 1.4081597328186035e-06, "ewc_loss_parallel": 2.250075340270996e-06, "grad_norm": 0.9889141917228699, "learning_rate": 5.977108944467995e-07, "loss": 0.6601, "mean_token_accuracy": 0.8045278191566467, "num_tokens": 5432228.0, "step": 142 }, { "epoch": 0.018191069838442948, "ewc_loss": 0.0023651123046875, "ewc_loss_diag": 1.430511474609375e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 0.9736770391464233, "learning_rate": 6.019499788045782e-07, "loss": 0.6386, "mean_token_accuracy": 0.8109228610992432, "num_tokens": 5469969.0, "step": 143 }, { "epoch": 0.018318280117033455, "ewc_loss": 0.002349853515625, "ewc_loss_diag": 1.4454126358032227e-06, "ewc_loss_parallel": 2.1904706954956055e-06, "grad_norm": 0.9515095353126526, "learning_rate": 6.06189063162357e-07, "loss": 0.632, "mean_token_accuracy": 0.809156596660614, "num_tokens": 5506262.0, "step": 144 }, { "epoch": 0.018445490395623966, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 1.475214958190918e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 1.023350477218628, "learning_rate": 6.104281475201357e-07, "loss": 0.6516, "mean_token_accuracy": 0.8035569190979004, "num_tokens": 5542870.0, "step": 145 }, { "epoch": 0.018572700674214477, "ewc_loss": 0.0023651123046875, "ewc_loss_diag": 1.4975666999816895e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 0.9607565999031067, "learning_rate": 6.146672318779144e-07, "loss": 0.6634, "mean_token_accuracy": 0.8004462122917175, "num_tokens": 5579958.0, "step": 146 }, { "epoch": 0.018699910952804988, "ewc_loss": 0.0023345947265625, "ewc_loss_diag": 1.5124678611755371e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 0.9761567115783691, "learning_rate": 6.189063162356931e-07, "loss": 0.6772, "mean_token_accuracy": 0.7960069179534912, "num_tokens": 5623306.0, "step": 147 }, { "epoch": 0.018827121231395495, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 1.5348196029663086e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 0.9331137537956238, "learning_rate": 6.231454005934719e-07, "loss": 0.6049, "mean_token_accuracy": 0.8193977475166321, "num_tokens": 5660801.0, "step": 148 }, { "epoch": 0.018954331509986006, "ewc_loss": 0.002288818359375, "ewc_loss_diag": 1.55717134475708e-06, "ewc_loss_parallel": 2.130866050720215e-06, "grad_norm": 0.9680423736572266, "learning_rate": 6.273844849512506e-07, "loss": 0.6375, "mean_token_accuracy": 0.8059722185134888, "num_tokens": 5696625.0, "step": 149 }, { "epoch": 0.019081541788576517, "ewc_loss": 0.002288818359375, "ewc_loss_diag": 1.5795230865478516e-06, "ewc_loss_parallel": 2.130866050720215e-06, "grad_norm": 0.9353109002113342, "learning_rate": 6.316235693090293e-07, "loss": 0.6557, "mean_token_accuracy": 0.8024487495422363, "num_tokens": 5733491.0, "step": 150 }, { "epoch": 0.019208752067167028, "ewc_loss": 0.002288818359375, "ewc_loss_diag": 1.6167759895324707e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 0.9658459424972534, "learning_rate": 6.35862653666808e-07, "loss": 0.6743, "mean_token_accuracy": 0.795793354511261, "num_tokens": 5773550.0, "step": 151 }, { "epoch": 0.01933596234575754, "ewc_loss": 0.002288818359375, "ewc_loss_diag": 1.6391277313232422e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 0.9993330836296082, "learning_rate": 6.401017380245868e-07, "loss": 0.6681, "mean_token_accuracy": 0.7989827394485474, "num_tokens": 5813114.0, "step": 152 }, { "epoch": 0.019463172624348046, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 1.6763806343078613e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 0.9707505106925964, "learning_rate": 6.443408223823655e-07, "loss": 0.6973, "mean_token_accuracy": 0.7920471429824829, "num_tokens": 5852702.0, "step": 153 }, { "epoch": 0.019590382902938557, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 1.7061829566955566e-06, "ewc_loss_parallel": 2.130866050720215e-06, "grad_norm": 0.990646481513977, "learning_rate": 6.485799067401442e-07, "loss": 0.7647, "mean_token_accuracy": 0.7721617221832275, "num_tokens": 5894481.0, "step": 154 }, { "epoch": 0.019717593181529068, "ewc_loss": 0.002288818359375, "ewc_loss_diag": 1.735985279083252e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 0.9474819302558899, "learning_rate": 6.528189910979229e-07, "loss": 0.6212, "mean_token_accuracy": 0.8072469830513, "num_tokens": 5929501.0, "step": 155 }, { "epoch": 0.01984480346011958, "ewc_loss": 0.00225830078125, "ewc_loss_diag": 1.7434358596801758e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 1.4245685338974, "learning_rate": 6.570580754557017e-07, "loss": 0.6205, "mean_token_accuracy": 0.8074449300765991, "num_tokens": 5966407.0, "step": 156 }, { "epoch": 0.019972013738710086, "ewc_loss": 0.00250244140625, "ewc_loss_diag": 1.7508864402770996e-06, "ewc_loss_parallel": 2.339482307434082e-06, "grad_norm": 0.9264846444129944, "learning_rate": 6.612971598134803e-07, "loss": 0.5771, "mean_token_accuracy": 0.8233463764190674, "num_tokens": 6004059.0, "step": 157 }, { "epoch": 0.020099224017300597, "ewc_loss": 0.002349853515625, "ewc_loss_diag": 1.773238182067871e-06, "ewc_loss_parallel": 2.16066837310791e-06, "grad_norm": 0.9305630922317505, "learning_rate": 6.65536244171259e-07, "loss": 0.632, "mean_token_accuracy": 0.8062871694564819, "num_tokens": 6040516.0, "step": 158 }, { "epoch": 0.020226434295891108, "ewc_loss": 0.0022125244140625, "ewc_loss_diag": 1.8030405044555664e-06, "ewc_loss_parallel": 2.0265579223632812e-06, "grad_norm": 0.9452441930770874, "learning_rate": 6.697753285290377e-07, "loss": 0.7029, "mean_token_accuracy": 0.7868205904960632, "num_tokens": 6077801.0, "step": 159 }, { "epoch": 0.02035364457448162, "ewc_loss": 0.002227783203125, "ewc_loss_diag": 1.8402934074401855e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 0.885644257068634, "learning_rate": 6.740144128868165e-07, "loss": 0.6204, "mean_token_accuracy": 0.8110132813453674, "num_tokens": 6114102.0, "step": 160 }, { "epoch": 0.020480854853072127, "ewc_loss": 0.0022125244140625, "ewc_loss_diag": 1.8551945686340332e-06, "ewc_loss_parallel": 2.0265579223632812e-06, "grad_norm": 0.8918651938438416, "learning_rate": 6.782534972445952e-07, "loss": 0.6418, "mean_token_accuracy": 0.8056307435035706, "num_tokens": 6157342.0, "step": 161 }, { "epoch": 0.020608065131662637, "ewc_loss": 0.002197265625, "ewc_loss_diag": 1.8924474716186523e-06, "ewc_loss_parallel": 2.0116567611694336e-06, "grad_norm": 0.9583126902580261, "learning_rate": 6.82492581602374e-07, "loss": 0.643, "mean_token_accuracy": 0.8057153820991516, "num_tokens": 6193058.0, "step": 162 }, { "epoch": 0.02073527541025315, "ewc_loss": 0.002288818359375, "ewc_loss_diag": 1.9371509552001953e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 0.975615382194519, "learning_rate": 6.867316659601526e-07, "loss": 0.6363, "mean_token_accuracy": 0.8108188509941101, "num_tokens": 6223616.0, "step": 163 }, { "epoch": 0.02086248568884366, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 1.9669532775878906e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 0.8967654705047607, "learning_rate": 6.909707503179314e-07, "loss": 0.6363, "mean_token_accuracy": 0.8098862767219543, "num_tokens": 6264920.0, "step": 164 }, { "epoch": 0.02098969596743417, "ewc_loss": 0.00225830078125, "ewc_loss_diag": 1.9818544387817383e-06, "ewc_loss_parallel": 2.0563602447509766e-06, "grad_norm": 0.911925196647644, "learning_rate": 6.952098346757101e-07, "loss": 0.6773, "mean_token_accuracy": 0.7941390872001648, "num_tokens": 6305270.0, "step": 165 }, { "epoch": 0.021116906246024678, "ewc_loss": 0.0022430419921875, "ewc_loss_diag": 2.0116567611694336e-06, "ewc_loss_parallel": 2.041459083557129e-06, "grad_norm": 0.8677566647529602, "learning_rate": 6.994489190334889e-07, "loss": 0.5818, "mean_token_accuracy": 0.8213973045349121, "num_tokens": 6341945.0, "step": 166 }, { "epoch": 0.02124411652461519, "ewc_loss": 0.002197265625, "ewc_loss_diag": 2.041459083557129e-06, "ewc_loss_parallel": 1.996755599975586e-06, "grad_norm": 0.8596996068954468, "learning_rate": 7.036880033912675e-07, "loss": 0.5825, "mean_token_accuracy": 0.823271632194519, "num_tokens": 6381565.0, "step": 167 }, { "epoch": 0.0213713268032057, "ewc_loss": 0.002227783203125, "ewc_loss_diag": 2.0712614059448242e-06, "ewc_loss_parallel": 2.0116567611694336e-06, "grad_norm": 0.8961923122406006, "learning_rate": 7.079270877490463e-07, "loss": 0.6305, "mean_token_accuracy": 0.8084093332290649, "num_tokens": 6419771.0, "step": 168 }, { "epoch": 0.02149853708179621, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 2.086162567138672e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 0.9101546406745911, "learning_rate": 7.12166172106825e-07, "loss": 0.5888, "mean_token_accuracy": 0.8190457224845886, "num_tokens": 6451413.0, "step": 169 }, { "epoch": 0.021625747360386718, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 2.086162567138672e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 0.8748034834861755, "learning_rate": 7.164052564646038e-07, "loss": 0.5944, "mean_token_accuracy": 0.8187262415885925, "num_tokens": 6488883.0, "step": 170 }, { "epoch": 0.02175295763897723, "ewc_loss": 0.002288818359375, "ewc_loss_diag": 2.1010637283325195e-06, "ewc_loss_parallel": 2.0712614059448242e-06, "grad_norm": 0.9484764337539673, "learning_rate": 7.206443408223824e-07, "loss": 0.6447, "mean_token_accuracy": 0.8039639592170715, "num_tokens": 6523669.0, "step": 171 }, { "epoch": 0.02188016791756774, "ewc_loss": 0.002349853515625, "ewc_loss_diag": 2.115964889526367e-06, "ewc_loss_parallel": 2.130866050720215e-06, "grad_norm": 0.8799686431884766, "learning_rate": 7.248834251801612e-07, "loss": 0.6069, "mean_token_accuracy": 0.8130297660827637, "num_tokens": 6561795.0, "step": 172 }, { "epoch": 0.02200737819615825, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 2.115964889526367e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 0.8905463814735413, "learning_rate": 7.291225095379399e-07, "loss": 0.5926, "mean_token_accuracy": 0.8191157579421997, "num_tokens": 6604102.0, "step": 173 }, { "epoch": 0.02213458847474876, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 2.130866050720215e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 0.8926240801811218, "learning_rate": 7.333615938957187e-07, "loss": 0.6768, "mean_token_accuracy": 0.7968350052833557, "num_tokens": 6645412.0, "step": 174 }, { "epoch": 0.02226179875333927, "ewc_loss": 0.0023040771484375, "ewc_loss_diag": 2.16066837310791e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 0.9315106272697449, "learning_rate": 7.376006782534972e-07, "loss": 0.5794, "mean_token_accuracy": 0.8235517740249634, "num_tokens": 6678407.0, "step": 175 }, { "epoch": 0.02238900903192978, "ewc_loss": 0.002349853515625, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 2.130866050720215e-06, "grad_norm": 0.8688061237335205, "learning_rate": 7.41839762611276e-07, "loss": 0.6123, "mean_token_accuracy": 0.8115702867507935, "num_tokens": 6720968.0, "step": 176 }, { "epoch": 0.02251621931052029, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 2.1904706954956055e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 0.8976710438728333, "learning_rate": 7.460788469690547e-07, "loss": 0.6158, "mean_token_accuracy": 0.8147029876708984, "num_tokens": 6757286.0, "step": 177 }, { "epoch": 0.0226434295891108, "ewc_loss": 0.0023345947265625, "ewc_loss_diag": 2.2351741790771484e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 0.9510659575462341, "learning_rate": 7.503179313268335e-07, "loss": 0.6484, "mean_token_accuracy": 0.8011926412582397, "num_tokens": 6790260.0, "step": 178 }, { "epoch": 0.02277063986770131, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 2.2351741790771484e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 0.8767879009246826, "learning_rate": 7.545570156846121e-07, "loss": 0.5914, "mean_token_accuracy": 0.8180845379829407, "num_tokens": 6828882.0, "step": 179 }, { "epoch": 0.02289785014629182, "ewc_loss": 0.0023345947265625, "ewc_loss_diag": 2.2351741790771484e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 0.8790555000305176, "learning_rate": 7.587961000423909e-07, "loss": 0.6064, "mean_token_accuracy": 0.817248523235321, "num_tokens": 6867332.0, "step": 180 }, { "epoch": 0.02302506042488233, "ewc_loss": 0.0023193359375, "ewc_loss_diag": 2.250075340270996e-06, "ewc_loss_parallel": 2.086162567138672e-06, "grad_norm": 0.8689723610877991, "learning_rate": 7.630351844001696e-07, "loss": 0.7193, "mean_token_accuracy": 0.7809861898422241, "num_tokens": 6914101.0, "step": 181 }, { "epoch": 0.02315227070347284, "ewc_loss": 0.0023345947265625, "ewc_loss_diag": 2.250075340270996e-06, "ewc_loss_parallel": 2.1010637283325195e-06, "grad_norm": 0.9144935011863708, "learning_rate": 7.672742687579484e-07, "loss": 0.6138, "mean_token_accuracy": 0.8118707537651062, "num_tokens": 6952981.0, "step": 182 }, { "epoch": 0.02327948098206335, "ewc_loss": 0.00238037109375, "ewc_loss_diag": 2.2798776626586914e-06, "ewc_loss_parallel": 2.1457672119140625e-06, "grad_norm": 0.9101467132568359, "learning_rate": 7.71513353115727e-07, "loss": 0.5918, "mean_token_accuracy": 0.816994309425354, "num_tokens": 6987534.0, "step": 183 }, { "epoch": 0.02340669126065386, "ewc_loss": 0.0024261474609375, "ewc_loss_diag": 2.294778823852539e-06, "ewc_loss_parallel": 2.1904706954956055e-06, "grad_norm": 0.8731935620307922, "learning_rate": 7.757524374735058e-07, "loss": 0.6652, "mean_token_accuracy": 0.7968111634254456, "num_tokens": 7029343.0, "step": 184 }, { "epoch": 0.02353390153924437, "ewc_loss": 0.0023651123046875, "ewc_loss_diag": 2.294778823852539e-06, "ewc_loss_parallel": 2.130866050720215e-06, "grad_norm": 0.8561438322067261, "learning_rate": 7.799915218312845e-07, "loss": 0.6031, "mean_token_accuracy": 0.8144278526306152, "num_tokens": 7067558.0, "step": 185 }, { "epoch": 0.02366111181783488, "ewc_loss": 0.002349853515625, "ewc_loss_diag": 2.3096799850463867e-06, "ewc_loss_parallel": 2.115964889526367e-06, "grad_norm": 0.8839266300201416, "learning_rate": 7.842306061890633e-07, "loss": 0.5817, "mean_token_accuracy": 0.8164492249488831, "num_tokens": 7106164.0, "step": 186 }, { "epoch": 0.023788322096425393, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 2.339482307434082e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 0.8626180291175842, "learning_rate": 7.884696905468419e-07, "loss": 0.6365, "mean_token_accuracy": 0.8063685894012451, "num_tokens": 7145549.0, "step": 187 }, { "epoch": 0.0239155323750159, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 2.384185791015625e-06, "ewc_loss_parallel": 2.16066837310791e-06, "grad_norm": 0.9185997843742371, "learning_rate": 7.927087749046207e-07, "loss": 0.6022, "mean_token_accuracy": 0.8162009716033936, "num_tokens": 7179697.0, "step": 188 }, { "epoch": 0.02404274265360641, "ewc_loss": 0.0024566650390625, "ewc_loss_diag": 2.3990869522094727e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 0.887380838394165, "learning_rate": 7.969478592623994e-07, "loss": 0.6771, "mean_token_accuracy": 0.7920600175857544, "num_tokens": 7220103.0, "step": 189 }, { "epoch": 0.024169952932196922, "ewc_loss": 0.0024566650390625, "ewc_loss_diag": 2.4437904357910156e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 0.8472028970718384, "learning_rate": 8.011869436201782e-07, "loss": 0.535, "mean_token_accuracy": 0.8362030982971191, "num_tokens": 7258059.0, "step": 190 }, { "epoch": 0.024297163210787433, "ewc_loss": 0.002410888671875, "ewc_loss_diag": 2.4586915969848633e-06, "ewc_loss_parallel": 2.16066837310791e-06, "grad_norm": 0.8874380588531494, "learning_rate": 8.054260279779568e-07, "loss": 0.6465, "mean_token_accuracy": 0.8060452938079834, "num_tokens": 7295111.0, "step": 191 }, { "epoch": 0.02442437348937794, "ewc_loss": 0.0024566650390625, "ewc_loss_diag": 2.4884939193725586e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 0.9048478603363037, "learning_rate": 8.096651123357356e-07, "loss": 0.6487, "mean_token_accuracy": 0.8000422716140747, "num_tokens": 7332703.0, "step": 192 }, { "epoch": 0.02455158376796845, "ewc_loss": 0.002471923828125, "ewc_loss_diag": 2.5033950805664062e-06, "ewc_loss_parallel": 2.2351741790771484e-06, "grad_norm": 0.8733810782432556, "learning_rate": 8.139041966935143e-07, "loss": 0.6518, "mean_token_accuracy": 0.8002615571022034, "num_tokens": 7376446.0, "step": 193 }, { "epoch": 0.024678794046558962, "ewc_loss": 0.002471923828125, "ewc_loss_diag": 2.5331974029541016e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 0.884557843208313, "learning_rate": 8.18143281051293e-07, "loss": 0.594, "mean_token_accuracy": 0.8165481090545654, "num_tokens": 7416378.0, "step": 194 }, { "epoch": 0.024806004325149473, "ewc_loss": 0.002471923828125, "ewc_loss_diag": 2.5331974029541016e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 0.8481647968292236, "learning_rate": 8.223823654090716e-07, "loss": 0.6039, "mean_token_accuracy": 0.8163812160491943, "num_tokens": 7458178.0, "step": 195 }, { "epoch": 0.024933214603739984, "ewc_loss": 0.002471923828125, "ewc_loss_diag": 2.562999725341797e-06, "ewc_loss_parallel": 2.205371856689453e-06, "grad_norm": 0.7972276210784912, "learning_rate": 8.266214497668504e-07, "loss": 0.5484, "mean_token_accuracy": 0.8318005800247192, "num_tokens": 7502436.0, "step": 196 }, { "epoch": 0.02506042488233049, "ewc_loss": 0.00244140625, "ewc_loss_diag": 2.5779008865356445e-06, "ewc_loss_parallel": 2.175569534301758e-06, "grad_norm": 0.8506367802619934, "learning_rate": 8.308605341246291e-07, "loss": 0.6065, "mean_token_accuracy": 0.8130166530609131, "num_tokens": 7539625.0, "step": 197 }, { "epoch": 0.025187635160921002, "ewc_loss": 0.0024871826171875, "ewc_loss_diag": 2.592802047729492e-06, "ewc_loss_parallel": 2.2202730178833008e-06, "grad_norm": 0.887686550617218, "learning_rate": 8.350996184824079e-07, "loss": 0.5827, "mean_token_accuracy": 0.8183218240737915, "num_tokens": 7570904.0, "step": 198 }, { "epoch": 0.025314845439511513, "ewc_loss": 0.0025482177734375, "ewc_loss_diag": 2.637505531311035e-06, "ewc_loss_parallel": 2.294778823852539e-06, "grad_norm": 0.8508486747741699, "learning_rate": 8.393387028401865e-07, "loss": 0.5776, "mean_token_accuracy": 0.8222663402557373, "num_tokens": 7608652.0, "step": 199 }, { "epoch": 0.025442055718102024, "ewc_loss": 0.0024871826171875, "ewc_loss_diag": 2.637505531311035e-06, "ewc_loss_parallel": 2.2351741790771484e-06, "grad_norm": 0.9172025918960571, "learning_rate": 8.435777871979653e-07, "loss": 0.6512, "mean_token_accuracy": 0.8035928606987, "num_tokens": 7647759.0, "step": 200 }, { "epoch": 0.02556926599669253, "ewc_loss": 0.0025634765625, "ewc_loss_diag": 2.682209014892578e-06, "ewc_loss_parallel": 2.294778823852539e-06, "grad_norm": 0.9342977404594421, "learning_rate": 8.47816871555744e-07, "loss": 0.5717, "mean_token_accuracy": 0.8199390172958374, "num_tokens": 7678918.0, "step": 201 }, { "epoch": 0.025696476275283042, "ewc_loss": 0.0025787353515625, "ewc_loss_diag": 2.7120113372802734e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 0.8581806421279907, "learning_rate": 8.520559559135228e-07, "loss": 0.5487, "mean_token_accuracy": 0.8286733031272888, "num_tokens": 7715797.0, "step": 202 }, { "epoch": 0.025823686553873553, "ewc_loss": 0.00250244140625, "ewc_loss_diag": 2.726912498474121e-06, "ewc_loss_parallel": 2.2351741790771484e-06, "grad_norm": 0.8439298272132874, "learning_rate": 8.562950402713014e-07, "loss": 0.653, "mean_token_accuracy": 0.8031984567642212, "num_tokens": 7763053.0, "step": 203 }, { "epoch": 0.025950896832464064, "ewc_loss": 0.00250244140625, "ewc_loss_diag": 2.771615982055664e-06, "ewc_loss_parallel": 2.2351741790771484e-06, "grad_norm": 0.9534378051757812, "learning_rate": 8.605341246290802e-07, "loss": 0.619, "mean_token_accuracy": 0.8138309121131897, "num_tokens": 7793956.0, "step": 204 }, { "epoch": 0.026078107111054575, "ewc_loss": 0.0026092529296875, "ewc_loss_diag": 2.7865171432495117e-06, "ewc_loss_parallel": 2.339482307434082e-06, "grad_norm": 0.8614175319671631, "learning_rate": 8.647732089868589e-07, "loss": 0.5973, "mean_token_accuracy": 0.8153544664382935, "num_tokens": 7833997.0, "step": 205 }, { "epoch": 0.026205317389645082, "ewc_loss": 0.002532958984375, "ewc_loss_diag": 2.8312206268310547e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 0.8959882855415344, "learning_rate": 8.690122933446377e-07, "loss": 0.6429, "mean_token_accuracy": 0.8057841658592224, "num_tokens": 7877711.0, "step": 206 }, { "epoch": 0.026332527668235593, "ewc_loss": 0.0025482177734375, "ewc_loss_diag": 2.8461217880249023e-06, "ewc_loss_parallel": 2.2649765014648438e-06, "grad_norm": 0.8990457653999329, "learning_rate": 8.732513777024163e-07, "loss": 0.5654, "mean_token_accuracy": 0.8282528519630432, "num_tokens": 7918136.0, "step": 207 }, { "epoch": 0.026459737946826104, "ewc_loss": 0.0026092529296875, "ewc_loss_diag": 2.8759241104125977e-06, "ewc_loss_parallel": 2.3245811462402344e-06, "grad_norm": 0.8501306176185608, "learning_rate": 8.774904620601951e-07, "loss": 0.6069, "mean_token_accuracy": 0.8155977725982666, "num_tokens": 7960847.0, "step": 208 }, { "epoch": 0.026586948225416615, "ewc_loss": 0.0025787353515625, "ewc_loss_diag": 2.8759241104125977e-06, "ewc_loss_parallel": 2.294778823852539e-06, "grad_norm": 0.8815220594406128, "learning_rate": 8.817295464179738e-07, "loss": 0.5689, "mean_token_accuracy": 0.8209234476089478, "num_tokens": 7996567.0, "step": 209 }, { "epoch": 0.026714158504007122, "ewc_loss": 0.0025787353515625, "ewc_loss_diag": 2.905726432800293e-06, "ewc_loss_parallel": 2.294778823852539e-06, "grad_norm": 0.8906700611114502, "learning_rate": 8.859686307757526e-07, "loss": 0.593, "mean_token_accuracy": 0.8172644376754761, "num_tokens": 8032441.0, "step": 210 }, { "epoch": 0.026841368782597633, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 2.9206275939941406e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 0.8823428153991699, "learning_rate": 8.902077151335312e-07, "loss": 0.625, "mean_token_accuracy": 0.8075093030929565, "num_tokens": 8072020.0, "step": 211 }, { "epoch": 0.026968579061188144, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 2.9653310775756836e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 0.9472223520278931, "learning_rate": 8.944467994913099e-07, "loss": 0.6702, "mean_token_accuracy": 0.7975699305534363, "num_tokens": 8105736.0, "step": 212 }, { "epoch": 0.027095789339778655, "ewc_loss": 0.002685546875, "ewc_loss_diag": 2.995133399963379e-06, "ewc_loss_parallel": 2.384185791015625e-06, "grad_norm": 0.8753838539123535, "learning_rate": 8.986858838490886e-07, "loss": 0.5223, "mean_token_accuracy": 0.8338513374328613, "num_tokens": 8139021.0, "step": 213 }, { "epoch": 0.027222999618369163, "ewc_loss": 0.0026092529296875, "ewc_loss_diag": 3.0100345611572266e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 0.8509573340415955, "learning_rate": 9.029249682068674e-07, "loss": 0.594, "mean_token_accuracy": 0.814732015132904, "num_tokens": 8182361.0, "step": 214 }, { "epoch": 0.027350209896959674, "ewc_loss": 0.002593994140625, "ewc_loss_diag": 3.039836883544922e-06, "ewc_loss_parallel": 2.294778823852539e-06, "grad_norm": 0.8906288743019104, "learning_rate": 9.07164052564646e-07, "loss": 0.6083, "mean_token_accuracy": 0.812373161315918, "num_tokens": 8218131.0, "step": 215 }, { "epoch": 0.027477420175550184, "ewc_loss": 0.0026397705078125, "ewc_loss_diag": 3.069639205932617e-06, "ewc_loss_parallel": 2.339482307434082e-06, "grad_norm": 0.885632336139679, "learning_rate": 9.114031369224248e-07, "loss": 0.5852, "mean_token_accuracy": 0.8227436542510986, "num_tokens": 8255886.0, "step": 216 }, { "epoch": 0.027604630454140695, "ewc_loss": 0.002685546875, "ewc_loss_diag": 3.0994415283203125e-06, "ewc_loss_parallel": 2.384185791015625e-06, "grad_norm": 0.8380799889564514, "learning_rate": 9.156422212802035e-07, "loss": 0.5417, "mean_token_accuracy": 0.8296196460723877, "num_tokens": 8297196.0, "step": 217 }, { "epoch": 0.027731840732731206, "ewc_loss": 0.0026092529296875, "ewc_loss_diag": 3.0994415283203125e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 0.8937220573425293, "learning_rate": 9.198813056379823e-07, "loss": 0.5831, "mean_token_accuracy": 0.8189324736595154, "num_tokens": 8331857.0, "step": 218 }, { "epoch": 0.027859051011321714, "ewc_loss": 0.0027008056640625, "ewc_loss_diag": 3.0994415283203125e-06, "ewc_loss_parallel": 2.3990869522094727e-06, "grad_norm": 0.9873502850532532, "learning_rate": 9.24120389995761e-07, "loss": 0.5778, "mean_token_accuracy": 0.8195312023162842, "num_tokens": 8371120.0, "step": 219 }, { "epoch": 0.027986261289912225, "ewc_loss": 0.002777099609375, "ewc_loss_diag": 3.0994415283203125e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 0.9534544348716736, "learning_rate": 9.283594743535397e-07, "loss": 0.6993, "mean_token_accuracy": 0.7865595817565918, "num_tokens": 8407502.0, "step": 220 }, { "epoch": 0.028113471568502735, "ewc_loss": 0.00274658203125, "ewc_loss_diag": 3.11434268951416e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 0.8676906824111938, "learning_rate": 9.325985587113184e-07, "loss": 0.5513, "mean_token_accuracy": 0.8296608328819275, "num_tokens": 8444360.0, "step": 221 }, { "epoch": 0.028240681847093246, "ewc_loss": 0.00262451171875, "ewc_loss_diag": 3.1441450119018555e-06, "ewc_loss_parallel": 2.3096799850463867e-06, "grad_norm": 0.9020501971244812, "learning_rate": 9.368376430690972e-07, "loss": 0.5634, "mean_token_accuracy": 0.8252862095832825, "num_tokens": 8476749.0, "step": 222 }, { "epoch": 0.028367892125683754, "ewc_loss": 0.002685546875, "ewc_loss_diag": 3.1888484954833984e-06, "ewc_loss_parallel": 2.3692846298217773e-06, "grad_norm": 0.8740898370742798, "learning_rate": 9.410767274268759e-07, "loss": 0.5936, "mean_token_accuracy": 0.8180035948753357, "num_tokens": 8517157.0, "step": 223 }, { "epoch": 0.028495102404274265, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 3.2186508178710938e-06, "ewc_loss_parallel": 2.3990869522094727e-06, "grad_norm": 0.8497171401977539, "learning_rate": 9.453158117846546e-07, "loss": 0.5648, "mean_token_accuracy": 0.8256371021270752, "num_tokens": 8556677.0, "step": 224 }, { "epoch": 0.028622312682864776, "ewc_loss": 0.002685546875, "ewc_loss_diag": 3.248453140258789e-06, "ewc_loss_parallel": 2.3692846298217773e-06, "grad_norm": 0.8687691688537598, "learning_rate": 9.495548961424333e-07, "loss": 0.5999, "mean_token_accuracy": 0.8183950185775757, "num_tokens": 8594335.0, "step": 225 }, { "epoch": 0.028749522961455286, "ewc_loss": 0.0027313232421875, "ewc_loss_diag": 3.293156623840332e-06, "ewc_loss_parallel": 2.3990869522094727e-06, "grad_norm": 0.8736017942428589, "learning_rate": 9.53793980500212e-07, "loss": 0.5211, "mean_token_accuracy": 0.8374336957931519, "num_tokens": 8626259.0, "step": 226 }, { "epoch": 0.028876733240045797, "ewc_loss": 0.0027313232421875, "ewc_loss_diag": 3.293156623840332e-06, "ewc_loss_parallel": 2.3990869522094727e-06, "grad_norm": 0.9025839567184448, "learning_rate": 9.580330648579909e-07, "loss": 0.6816, "mean_token_accuracy": 0.7956506013870239, "num_tokens": 8665481.0, "step": 227 }, { "epoch": 0.029003943518636305, "ewc_loss": 0.0027618408203125, "ewc_loss_diag": 3.3527612686157227e-06, "ewc_loss_parallel": 2.428889274597168e-06, "grad_norm": 0.8537615537643433, "learning_rate": 9.622721492157695e-07, "loss": 0.5479, "mean_token_accuracy": 0.8323878049850464, "num_tokens": 8703136.0, "step": 228 }, { "epoch": 0.029131153797226816, "ewc_loss": 0.002716064453125, "ewc_loss_diag": 3.3974647521972656e-06, "ewc_loss_parallel": 2.384185791015625e-06, "grad_norm": 0.9170428514480591, "learning_rate": 9.665112335735482e-07, "loss": 0.562, "mean_token_accuracy": 0.8244750499725342, "num_tokens": 8734989.0, "step": 229 }, { "epoch": 0.029258364075817327, "ewc_loss": 0.0027923583984375, "ewc_loss_diag": 3.4123659133911133e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 0.8794753551483154, "learning_rate": 9.707503179313269e-07, "loss": 0.5286, "mean_token_accuracy": 0.8336808085441589, "num_tokens": 8768145.0, "step": 230 }, { "epoch": 0.029385574354407838, "ewc_loss": 0.0028076171875, "ewc_loss_diag": 3.427267074584961e-06, "ewc_loss_parallel": 2.4586915969848633e-06, "grad_norm": 0.8909416794776917, "learning_rate": 9.749894022891056e-07, "loss": 0.5814, "mean_token_accuracy": 0.82086181640625, "num_tokens": 8804668.0, "step": 231 }, { "epoch": 0.029512784632998345, "ewc_loss": 0.00274658203125, "ewc_loss_diag": 3.427267074584961e-06, "ewc_loss_parallel": 2.4139881134033203e-06, "grad_norm": 0.8523673415184021, "learning_rate": 9.792284866468842e-07, "loss": 0.6034, "mean_token_accuracy": 0.809325098991394, "num_tokens": 8848343.0, "step": 232 }, { "epoch": 0.029639994911588856, "ewc_loss": 0.0027618408203125, "ewc_loss_diag": 3.4570693969726562e-06, "ewc_loss_parallel": 2.4139881134033203e-06, "grad_norm": 0.8743090033531189, "learning_rate": 9.834675710046631e-07, "loss": 0.5683, "mean_token_accuracy": 0.82021564245224, "num_tokens": 8883794.0, "step": 233 }, { "epoch": 0.029767205190179367, "ewc_loss": 0.0028228759765625, "ewc_loss_diag": 3.471970558166504e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 0.8831401467323303, "learning_rate": 9.877066553624418e-07, "loss": 0.5885, "mean_token_accuracy": 0.8166743516921997, "num_tokens": 8921597.0, "step": 234 }, { "epoch": 0.029894415468769878, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 3.4868717193603516e-06, "ewc_loss_parallel": 2.4884939193725586e-06, "grad_norm": 0.955553412437439, "learning_rate": 9.919457397202205e-07, "loss": 0.609, "mean_token_accuracy": 0.8126741647720337, "num_tokens": 8956260.0, "step": 235 }, { "epoch": 0.030021625747360385, "ewc_loss": 0.0029144287109375, "ewc_loss_diag": 3.4868717193603516e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 0.9715394377708435, "learning_rate": 9.961848240779991e-07, "loss": 0.6808, "mean_token_accuracy": 0.7961462140083313, "num_tokens": 8992492.0, "step": 236 }, { "epoch": 0.030148836025950896, "ewc_loss": 0.0029296875, "ewc_loss_diag": 3.516674041748047e-06, "ewc_loss_parallel": 2.5779008865356445e-06, "grad_norm": 0.9270706176757812, "learning_rate": 1.000423908435778e-06, "loss": 0.5856, "mean_token_accuracy": 0.819272518157959, "num_tokens": 9028300.0, "step": 237 }, { "epoch": 0.030276046304541407, "ewc_loss": 0.00286865234375, "ewc_loss_diag": 3.546476364135742e-06, "ewc_loss_parallel": 2.518296241760254e-06, "grad_norm": 0.867125928401947, "learning_rate": 1.0046629927935567e-06, "loss": 0.5786, "mean_token_accuracy": 0.8211297392845154, "num_tokens": 9066620.0, "step": 238 }, { "epoch": 0.030403256583131918, "ewc_loss": 0.0028076171875, "ewc_loss_diag": 3.5762786865234375e-06, "ewc_loss_parallel": 2.4437904357910156e-06, "grad_norm": 0.8762415051460266, "learning_rate": 1.0089020771513354e-06, "loss": 0.5244, "mean_token_accuracy": 0.8373963832855225, "num_tokens": 9105213.0, "step": 239 }, { "epoch": 0.03053046686172243, "ewc_loss": 0.002838134765625, "ewc_loss_diag": 3.606081008911133e-06, "ewc_loss_parallel": 2.473592758178711e-06, "grad_norm": 0.9139494299888611, "learning_rate": 1.013141161509114e-06, "loss": 0.5848, "mean_token_accuracy": 0.8197089433670044, "num_tokens": 9139673.0, "step": 240 }, { "epoch": 0.030657677140312936, "ewc_loss": 0.0029296875, "ewc_loss_diag": 3.635883331298828e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 0.8999788165092468, "learning_rate": 1.017380245866893e-06, "loss": 0.5866, "mean_token_accuracy": 0.8158171772956848, "num_tokens": 9172435.0, "step": 241 }, { "epoch": 0.030784887418903447, "ewc_loss": 0.0029296875, "ewc_loss_diag": 3.6656856536865234e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 0.9155688881874084, "learning_rate": 1.0216193302246716e-06, "loss": 0.6104, "mean_token_accuracy": 0.8062840700149536, "num_tokens": 9206015.0, "step": 242 }, { "epoch": 0.030912097697493958, "ewc_loss": 0.002960205078125, "ewc_loss_diag": 3.6656856536865234e-06, "ewc_loss_parallel": 2.592802047729492e-06, "grad_norm": 0.8489539623260498, "learning_rate": 1.0258584145824503e-06, "loss": 0.5463, "mean_token_accuracy": 0.8286823034286499, "num_tokens": 9247224.0, "step": 243 }, { "epoch": 0.03103930797608447, "ewc_loss": 0.00286865234375, "ewc_loss_diag": 3.6656856536865234e-06, "ewc_loss_parallel": 2.5033950805664062e-06, "grad_norm": 0.893195390701294, "learning_rate": 1.030097498940229e-06, "loss": 0.5314, "mean_token_accuracy": 0.8330267071723938, "num_tokens": 9282171.0, "step": 244 }, { "epoch": 0.031166518254674976, "ewc_loss": 0.0029296875, "ewc_loss_diag": 3.725290298461914e-06, "ewc_loss_parallel": 2.562999725341797e-06, "grad_norm": 0.8463425636291504, "learning_rate": 1.0343365832980078e-06, "loss": 0.5831, "mean_token_accuracy": 0.818732738494873, "num_tokens": 9324078.0, "step": 245 }, { "epoch": 0.03129372853326549, "ewc_loss": 0.002899169921875, "ewc_loss_diag": 3.7550926208496094e-06, "ewc_loss_parallel": 2.518296241760254e-06, "grad_norm": 0.8705418705940247, "learning_rate": 1.0385756676557865e-06, "loss": 0.579, "mean_token_accuracy": 0.8203511238098145, "num_tokens": 9364516.0, "step": 246 }, { "epoch": 0.031420938811855995, "ewc_loss": 0.00299072265625, "ewc_loss_diag": 3.769993782043457e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 0.873184323310852, "learning_rate": 1.0428147520135652e-06, "loss": 0.5568, "mean_token_accuracy": 0.8269450664520264, "num_tokens": 9403804.0, "step": 247 }, { "epoch": 0.03154814909044651, "ewc_loss": 0.00299072265625, "ewc_loss_diag": 3.7848949432373047e-06, "ewc_loss_parallel": 2.60770320892334e-06, "grad_norm": 0.8838789463043213, "learning_rate": 1.0470538363713438e-06, "loss": 0.5989, "mean_token_accuracy": 0.8133822679519653, "num_tokens": 9441589.0, "step": 248 }, { "epoch": 0.031675359369037016, "ewc_loss": 0.0030059814453125, "ewc_loss_diag": 3.7997961044311523e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 0.899558424949646, "learning_rate": 1.0512929207291225e-06, "loss": 0.5998, "mean_token_accuracy": 0.8130953907966614, "num_tokens": 9480240.0, "step": 249 }, { "epoch": 0.03180256964762753, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 3.814697265625e-06, "ewc_loss_parallel": 2.6673078536987305e-06, "grad_norm": 0.9889190196990967, "learning_rate": 1.0555320050869012e-06, "loss": 0.5613, "mean_token_accuracy": 0.8212476968765259, "num_tokens": 9512704.0, "step": 250 }, { "epoch": 0.03192977992621804, "ewc_loss": 0.003143310546875, "ewc_loss_diag": 3.844499588012695e-06, "ewc_loss_parallel": 2.7567148208618164e-06, "grad_norm": 0.9653248190879822, "learning_rate": 1.05977108944468e-06, "loss": 0.6165, "mean_token_accuracy": 0.8051126003265381, "num_tokens": 9545057.0, "step": 251 }, { "epoch": 0.032056990204808546, "ewc_loss": 0.00311279296875, "ewc_loss_diag": 3.874301910400391e-06, "ewc_loss_parallel": 2.726912498474121e-06, "grad_norm": 0.8760961294174194, "learning_rate": 1.0640101738024587e-06, "loss": 0.5599, "mean_token_accuracy": 0.8266851902008057, "num_tokens": 9585797.0, "step": 252 }, { "epoch": 0.03218420048339906, "ewc_loss": 0.0029754638671875, "ewc_loss_diag": 3.904104232788086e-06, "ewc_loss_parallel": 2.5779008865356445e-06, "grad_norm": 0.8979047536849976, "learning_rate": 1.0682492581602374e-06, "loss": 0.5205, "mean_token_accuracy": 0.837653398513794, "num_tokens": 9617890.0, "step": 253 }, { "epoch": 0.03231141076198957, "ewc_loss": 0.0030364990234375, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 2.637505531311035e-06, "grad_norm": 0.9006310701370239, "learning_rate": 1.072488342518016e-06, "loss": 0.52, "mean_token_accuracy": 0.8340772390365601, "num_tokens": 9652895.0, "step": 254 }, { "epoch": 0.03243862104058008, "ewc_loss": 0.0030517578125, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 2.652406692504883e-06, "grad_norm": 0.9100434184074402, "learning_rate": 1.076727426875795e-06, "loss": 0.5804, "mean_token_accuracy": 0.8184356093406677, "num_tokens": 9691804.0, "step": 255 }, { "epoch": 0.03256583131917059, "ewc_loss": 0.0030975341796875, "ewc_loss_diag": 3.933906555175781e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 0.8771400451660156, "learning_rate": 1.0809665112335736e-06, "loss": 0.6054, "mean_token_accuracy": 0.8142603635787964, "num_tokens": 9736178.0, "step": 256 }, { "epoch": 0.0326930415977611, "ewc_loss": 0.003021240234375, "ewc_loss_diag": 3.9637088775634766e-06, "ewc_loss_parallel": 2.6226043701171875e-06, "grad_norm": 0.8892201781272888, "learning_rate": 1.0852055955913523e-06, "loss": 0.4965, "mean_token_accuracy": 0.8435927629470825, "num_tokens": 9773867.0, "step": 257 }, { "epoch": 0.03282025187635161, "ewc_loss": 0.0030670166015625, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 2.6673078536987305e-06, "grad_norm": 0.8747554421424866, "learning_rate": 1.089444679949131e-06, "loss": 0.6031, "mean_token_accuracy": 0.8132374286651611, "num_tokens": 9813764.0, "step": 258 }, { "epoch": 0.03294746215494212, "ewc_loss": 0.003082275390625, "ewc_loss_diag": 3.993511199951172e-06, "ewc_loss_parallel": 2.682209014892578e-06, "grad_norm": 0.8601885437965393, "learning_rate": 1.0936837643069099e-06, "loss": 0.5579, "mean_token_accuracy": 0.8251882791519165, "num_tokens": 9855447.0, "step": 259 }, { "epoch": 0.03307467243353263, "ewc_loss": 0.0030670166015625, "ewc_loss_diag": 4.023313522338867e-06, "ewc_loss_parallel": 2.6673078536987305e-06, "grad_norm": 0.8790988922119141, "learning_rate": 1.0979228486646885e-06, "loss": 0.5184, "mean_token_accuracy": 0.8398501873016357, "num_tokens": 9889806.0, "step": 260 }, { "epoch": 0.03320188271212314, "ewc_loss": 0.0030975341796875, "ewc_loss_diag": 4.023313522338867e-06, "ewc_loss_parallel": 2.6971101760864258e-06, "grad_norm": 0.9097117781639099, "learning_rate": 1.1021619330224672e-06, "loss": 0.6585, "mean_token_accuracy": 0.7954362630844116, "num_tokens": 9929257.0, "step": 261 }, { "epoch": 0.03332909299071365, "ewc_loss": 0.003143310546875, "ewc_loss_diag": 4.0531158447265625e-06, "ewc_loss_parallel": 2.7418136596679688e-06, "grad_norm": 0.8914580345153809, "learning_rate": 1.1064010173802459e-06, "loss": 0.59, "mean_token_accuracy": 0.8156272172927856, "num_tokens": 9967822.0, "step": 262 }, { "epoch": 0.03345630326930416, "ewc_loss": 0.003143310546875, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 2.726912498474121e-06, "grad_norm": 0.8711107969284058, "learning_rate": 1.1106401017380248e-06, "loss": 0.5212, "mean_token_accuracy": 0.8362653255462646, "num_tokens": 10004873.0, "step": 263 }, { "epoch": 0.03358351354789467, "ewc_loss": 0.003143310546875, "ewc_loss_diag": 4.082918167114258e-06, "ewc_loss_parallel": 2.726912498474121e-06, "grad_norm": 0.9689542055130005, "learning_rate": 1.1148791860958034e-06, "loss": 0.6026, "mean_token_accuracy": 0.8097431659698486, "num_tokens": 10033761.0, "step": 264 }, { "epoch": 0.03371072382648518, "ewc_loss": 0.0032501220703125, "ewc_loss_diag": 4.112720489501953e-06, "ewc_loss_parallel": 2.8312206268310547e-06, "grad_norm": 0.9836970567703247, "learning_rate": 1.1191182704535821e-06, "loss": 0.5893, "mean_token_accuracy": 0.8160487413406372, "num_tokens": 10072586.0, "step": 265 }, { "epoch": 0.03383793410507569, "ewc_loss": 0.0032501220703125, "ewc_loss_diag": 4.112720489501953e-06, "ewc_loss_parallel": 2.8312206268310547e-06, "grad_norm": 0.8846592307090759, "learning_rate": 1.123357354811361e-06, "loss": 0.5822, "mean_token_accuracy": 0.8190954327583313, "num_tokens": 10112210.0, "step": 266 }, { "epoch": 0.0339651443836662, "ewc_loss": 0.0031280517578125, "ewc_loss_diag": 4.1425228118896484e-06, "ewc_loss_parallel": 2.7120113372802734e-06, "grad_norm": 0.8691883087158203, "learning_rate": 1.1275964391691395e-06, "loss": 0.5656, "mean_token_accuracy": 0.8255274295806885, "num_tokens": 10155908.0, "step": 267 }, { "epoch": 0.03409235466225671, "ewc_loss": 0.0031280517578125, "ewc_loss_diag": 4.172325134277344e-06, "ewc_loss_parallel": 2.7120113372802734e-06, "grad_norm": 0.9014133810997009, "learning_rate": 1.1318355235269181e-06, "loss": 0.6175, "mean_token_accuracy": 0.8074909448623657, "num_tokens": 10195944.0, "step": 268 }, { "epoch": 0.03421956494084722, "ewc_loss": 0.003204345703125, "ewc_loss_diag": 4.202127456665039e-06, "ewc_loss_parallel": 2.771615982055664e-06, "grad_norm": 0.8991667032241821, "learning_rate": 1.136074607884697e-06, "loss": 0.604, "mean_token_accuracy": 0.8121789693832397, "num_tokens": 10233602.0, "step": 269 }, { "epoch": 0.03434677521943773, "ewc_loss": 0.00323486328125, "ewc_loss_diag": 4.231929779052734e-06, "ewc_loss_parallel": 2.8014183044433594e-06, "grad_norm": 0.925659716129303, "learning_rate": 1.1403136922424757e-06, "loss": 0.5574, "mean_token_accuracy": 0.8252843618392944, "num_tokens": 10267001.0, "step": 270 }, { "epoch": 0.03447398549802824, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 4.26173210144043e-06, "ewc_loss_parallel": 2.8312206268310547e-06, "grad_norm": 0.8737414479255676, "learning_rate": 1.1445527766002544e-06, "loss": 0.5662, "mean_token_accuracy": 0.819980263710022, "num_tokens": 10308264.0, "step": 271 }, { "epoch": 0.03460119577661875, "ewc_loss": 0.0031890869140625, "ewc_loss_diag": 4.26173210144043e-06, "ewc_loss_parallel": 2.7567148208618164e-06, "grad_norm": 0.9582952260971069, "learning_rate": 1.148791860958033e-06, "loss": 0.6666, "mean_token_accuracy": 0.7946419715881348, "num_tokens": 10343440.0, "step": 272 }, { "epoch": 0.034728406055209264, "ewc_loss": 0.00335693359375, "ewc_loss_diag": 4.26173210144043e-06, "ewc_loss_parallel": 2.9355287551879883e-06, "grad_norm": 0.9368519186973572, "learning_rate": 1.153030945315812e-06, "loss": 0.595, "mean_token_accuracy": 0.8164997100830078, "num_tokens": 10383402.0, "step": 273 }, { "epoch": 0.03485561633379977, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 4.26173210144043e-06, "ewc_loss_parallel": 2.8759241104125977e-06, "grad_norm": 0.8953797817230225, "learning_rate": 1.1572700296735906e-06, "loss": 0.5703, "mean_token_accuracy": 0.819240927696228, "num_tokens": 10421700.0, "step": 274 }, { "epoch": 0.03498282661239028, "ewc_loss": 0.0032196044921875, "ewc_loss_diag": 4.32133674621582e-06, "ewc_loss_parallel": 2.7865171432495117e-06, "grad_norm": 0.9263890385627747, "learning_rate": 1.1615091140313693e-06, "loss": 0.6101, "mean_token_accuracy": 0.8118722438812256, "num_tokens": 10460295.0, "step": 275 }, { "epoch": 0.03511003689098079, "ewc_loss": 0.0032958984375, "ewc_loss_diag": 4.32133674621582e-06, "ewc_loss_parallel": 2.8759241104125977e-06, "grad_norm": 0.8777822852134705, "learning_rate": 1.165748198389148e-06, "loss": 0.5581, "mean_token_accuracy": 0.8258909583091736, "num_tokens": 10501619.0, "step": 276 }, { "epoch": 0.0352372471695713, "ewc_loss": 0.003265380859375, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 2.816319465637207e-06, "grad_norm": 0.8915600180625916, "learning_rate": 1.1699872827469268e-06, "loss": 0.5613, "mean_token_accuracy": 0.8251303434371948, "num_tokens": 10540465.0, "step": 277 }, { "epoch": 0.03536445744816181, "ewc_loss": 0.003326416015625, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 0.8724315166473389, "learning_rate": 1.1742263671047055e-06, "loss": 0.5683, "mean_token_accuracy": 0.822616696357727, "num_tokens": 10580980.0, "step": 278 }, { "epoch": 0.03549166772675232, "ewc_loss": 0.003326416015625, "ewc_loss_diag": 4.351139068603516e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 0.9378242492675781, "learning_rate": 1.1784654514624842e-06, "loss": 0.6409, "mean_token_accuracy": 0.7995902895927429, "num_tokens": 10621332.0, "step": 279 }, { "epoch": 0.03561887800534283, "ewc_loss": 0.0034637451171875, "ewc_loss_diag": 4.380941390991211e-06, "ewc_loss_parallel": 3.0249357223510742e-06, "grad_norm": 0.9776835441589355, "learning_rate": 1.1827045358202629e-06, "loss": 0.5942, "mean_token_accuracy": 0.8181871175765991, "num_tokens": 10654468.0, "step": 280 }, { "epoch": 0.035746088283933344, "ewc_loss": 0.0035552978515625, "ewc_loss_diag": 4.380941390991211e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 0.8948056697845459, "learning_rate": 1.1869436201780417e-06, "loss": 0.5149, "mean_token_accuracy": 0.8390790224075317, "num_tokens": 10692191.0, "step": 281 }, { "epoch": 0.03587329856252385, "ewc_loss": 0.003326416015625, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 2.8908252716064453e-06, "grad_norm": 0.8775851726531982, "learning_rate": 1.1911827045358204e-06, "loss": 0.5551, "mean_token_accuracy": 0.8310303092002869, "num_tokens": 10731568.0, "step": 282 }, { "epoch": 0.03600050884111436, "ewc_loss": 0.00335693359375, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 2.9206275939941406e-06, "grad_norm": 0.9128066897392273, "learning_rate": 1.195421788893599e-06, "loss": 0.6263, "mean_token_accuracy": 0.8095789551734924, "num_tokens": 10772166.0, "step": 283 }, { "epoch": 0.036127719119704874, "ewc_loss": 0.0034942626953125, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 3.0547380447387695e-06, "grad_norm": 0.9097956418991089, "learning_rate": 1.199660873251378e-06, "loss": 0.5686, "mean_token_accuracy": 0.8239852786064148, "num_tokens": 10807527.0, "step": 284 }, { "epoch": 0.03625492939829538, "ewc_loss": 0.0034637451171875, "ewc_loss_diag": 4.4405460357666016e-06, "ewc_loss_parallel": 3.0249357223510742e-06, "grad_norm": 0.8916672468185425, "learning_rate": 1.2038999576091564e-06, "loss": 0.5419, "mean_token_accuracy": 0.8304358720779419, "num_tokens": 10846064.0, "step": 285 }, { "epoch": 0.036382139676885895, "ewc_loss": 0.0034027099609375, "ewc_loss_diag": 4.470348358154297e-06, "ewc_loss_parallel": 2.9653310775756836e-06, "grad_norm": 0.8793213367462158, "learning_rate": 1.208139041966935e-06, "loss": 0.5191, "mean_token_accuracy": 0.8400102853775024, "num_tokens": 10884496.0, "step": 286 }, { "epoch": 0.0365093499554764, "ewc_loss": 0.0034332275390625, "ewc_loss_diag": 4.470348358154297e-06, "ewc_loss_parallel": 2.995133399963379e-06, "grad_norm": 0.9405278563499451, "learning_rate": 1.212378126324714e-06, "loss": 0.5626, "mean_token_accuracy": 0.8219653964042664, "num_tokens": 10917542.0, "step": 287 }, { "epoch": 0.03663656023406691, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 4.500150680541992e-06, "ewc_loss_parallel": 3.129243850708008e-06, "grad_norm": 0.9808501601219177, "learning_rate": 1.2166172106824927e-06, "loss": 0.627, "mean_token_accuracy": 0.8111428618431091, "num_tokens": 10949596.0, "step": 288 }, { "epoch": 0.036763770512657425, "ewc_loss": 0.0037078857421875, "ewc_loss_diag": 4.5299530029296875e-06, "ewc_loss_parallel": 3.248453140258789e-06, "grad_norm": 0.9058064818382263, "learning_rate": 1.2208562950402713e-06, "loss": 0.5331, "mean_token_accuracy": 0.829758882522583, "num_tokens": 10986143.0, "step": 289 }, { "epoch": 0.03689098079124793, "ewc_loss": 0.0034637451171875, "ewc_loss_diag": 4.559755325317383e-06, "ewc_loss_parallel": 3.0100345611572266e-06, "grad_norm": 0.9173493385314941, "learning_rate": 1.22509537939805e-06, "loss": 0.5717, "mean_token_accuracy": 0.8196185827255249, "num_tokens": 11024678.0, "step": 290 }, { "epoch": 0.03701819106983844, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 3.0547380447387695e-06, "grad_norm": 0.8829053640365601, "learning_rate": 1.2293344637558289e-06, "loss": 0.5886, "mean_token_accuracy": 0.8164401054382324, "num_tokens": 11068950.0, "step": 291 }, { "epoch": 0.037145401348428954, "ewc_loss": 0.003509521484375, "ewc_loss_diag": 4.589557647705078e-06, "ewc_loss_parallel": 3.0547380447387695e-06, "grad_norm": 0.9041259288787842, "learning_rate": 1.2335735481136076e-06, "loss": 0.5368, "mean_token_accuracy": 0.8275194764137268, "num_tokens": 11108697.0, "step": 292 }, { "epoch": 0.03727261162701946, "ewc_loss": 0.003570556640625, "ewc_loss_diag": 4.6193599700927734e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 0.9010706543922424, "learning_rate": 1.2378126324713862e-06, "loss": 0.5556, "mean_token_accuracy": 0.8297672271728516, "num_tokens": 11148482.0, "step": 293 }, { "epoch": 0.037399821905609976, "ewc_loss": 0.0036163330078125, "ewc_loss_diag": 4.6193599700927734e-06, "ewc_loss_parallel": 3.159046173095703e-06, "grad_norm": 0.8881252408027649, "learning_rate": 1.242051716829165e-06, "loss": 0.5699, "mean_token_accuracy": 0.8156402707099915, "num_tokens": 11189139.0, "step": 294 }, { "epoch": 0.03752703218420048, "ewc_loss": 0.0035858154296875, "ewc_loss_diag": 4.678964614868164e-06, "ewc_loss_parallel": 3.11434268951416e-06, "grad_norm": 0.8925266265869141, "learning_rate": 1.2462908011869438e-06, "loss": 0.5555, "mean_token_accuracy": 0.8242539167404175, "num_tokens": 11226484.0, "step": 295 }, { "epoch": 0.03765424246279099, "ewc_loss": 0.0036468505859375, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 3.1739473342895508e-06, "grad_norm": 0.8997142910957336, "learning_rate": 1.2505298855447223e-06, "loss": 0.5104, "mean_token_accuracy": 0.8380498886108398, "num_tokens": 11263806.0, "step": 296 }, { "epoch": 0.037781452741381505, "ewc_loss": 0.0036468505859375, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 3.1739473342895508e-06, "grad_norm": 0.9655401110649109, "learning_rate": 1.2547689699025011e-06, "loss": 0.6331, "mean_token_accuracy": 0.8044699430465698, "num_tokens": 11299822.0, "step": 297 }, { "epoch": 0.03790866301997201, "ewc_loss": 0.0038299560546875, "ewc_loss_diag": 4.708766937255859e-06, "ewc_loss_parallel": 3.3527612686157227e-06, "grad_norm": 0.9933217763900757, "learning_rate": 1.2590080542602798e-06, "loss": 0.5942, "mean_token_accuracy": 0.8142819404602051, "num_tokens": 11330796.0, "step": 298 }, { "epoch": 0.03803587329856253, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 0.922745943069458, "learning_rate": 1.2632471386180587e-06, "loss": 0.5848, "mean_token_accuracy": 0.8162676692008972, "num_tokens": 11369529.0, "step": 299 }, { "epoch": 0.038163083577153034, "ewc_loss": 0.0036468505859375, "ewc_loss_diag": 4.738569259643555e-06, "ewc_loss_parallel": 3.1739473342895508e-06, "grad_norm": 0.9484642148017883, "learning_rate": 1.2674862229758372e-06, "loss": 0.5583, "mean_token_accuracy": 0.8317000269889832, "num_tokens": 11405044.0, "step": 300 }, { "epoch": 0.03829029385574354, "ewc_loss": 0.0037384033203125, "ewc_loss_diag": 4.76837158203125e-06, "ewc_loss_parallel": 3.2633543014526367e-06, "grad_norm": 0.9089944362640381, "learning_rate": 1.271725307333616e-06, "loss": 0.5533, "mean_token_accuracy": 0.8284049034118652, "num_tokens": 11445409.0, "step": 301 }, { "epoch": 0.038417504134334056, "ewc_loss": 0.003692626953125, "ewc_loss_diag": 4.76837158203125e-06, "ewc_loss_parallel": 3.2186508178710938e-06, "grad_norm": 0.8585637211799622, "learning_rate": 1.2759643916913947e-06, "loss": 0.548, "mean_token_accuracy": 0.8263908624649048, "num_tokens": 11488285.0, "step": 302 }, { "epoch": 0.03854471441292456, "ewc_loss": 0.003631591796875, "ewc_loss_diag": 4.798173904418945e-06, "ewc_loss_parallel": 3.1441450119018555e-06, "grad_norm": 0.9440265893936157, "learning_rate": 1.2802034760491736e-06, "loss": 0.555, "mean_token_accuracy": 0.8256703615188599, "num_tokens": 11520970.0, "step": 303 }, { "epoch": 0.03867192469151508, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 4.857778549194336e-06, "ewc_loss_parallel": 3.382563591003418e-06, "grad_norm": 0.907875120639801, "learning_rate": 1.284442560406952e-06, "loss": 0.6124, "mean_token_accuracy": 0.806936502456665, "num_tokens": 11563590.0, "step": 304 }, { "epoch": 0.038799134970105585, "ewc_loss": 0.0037994384765625, "ewc_loss_diag": 4.857778549194336e-06, "ewc_loss_parallel": 3.3080577850341797e-06, "grad_norm": 0.9898818135261536, "learning_rate": 1.288681644764731e-06, "loss": 0.6529, "mean_token_accuracy": 0.7948039770126343, "num_tokens": 11598629.0, "step": 305 }, { "epoch": 0.03892634524869609, "ewc_loss": 0.003936767578125, "ewc_loss_diag": 4.857778549194336e-06, "ewc_loss_parallel": 3.4570693969726562e-06, "grad_norm": 0.9337149262428284, "learning_rate": 1.2929207291225096e-06, "loss": 0.5523, "mean_token_accuracy": 0.8296020030975342, "num_tokens": 11635526.0, "step": 306 }, { "epoch": 0.03905355552728661, "ewc_loss": 0.0038299560546875, "ewc_loss_diag": 4.857778549194336e-06, "ewc_loss_parallel": 3.337860107421875e-06, "grad_norm": 0.9473824501037598, "learning_rate": 1.2971598134802885e-06, "loss": 0.5702, "mean_token_accuracy": 0.822593629360199, "num_tokens": 11672974.0, "step": 307 }, { "epoch": 0.039180765805877114, "ewc_loss": 0.0038604736328125, "ewc_loss_diag": 4.887580871582031e-06, "ewc_loss_parallel": 3.3676624298095703e-06, "grad_norm": 0.9591264724731445, "learning_rate": 1.301398897838067e-06, "loss": 0.591, "mean_token_accuracy": 0.816711962223053, "num_tokens": 11708886.0, "step": 308 }, { "epoch": 0.03930797608446762, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 4.887580871582031e-06, "ewc_loss_parallel": 3.4570693969726562e-06, "grad_norm": 0.9144315719604492, "learning_rate": 1.3056379821958458e-06, "loss": 0.6001, "mean_token_accuracy": 0.8176484107971191, "num_tokens": 11749217.0, "step": 309 }, { "epoch": 0.039435186363058136, "ewc_loss": 0.003814697265625, "ewc_loss_diag": 4.9173831939697266e-06, "ewc_loss_parallel": 3.3229589462280273e-06, "grad_norm": 0.9227131009101868, "learning_rate": 1.3098770665536245e-06, "loss": 0.5619, "mean_token_accuracy": 0.8258742690086365, "num_tokens": 11788380.0, "step": 310 }, { "epoch": 0.039562396641648644, "ewc_loss": 0.00390625, "ewc_loss_diag": 4.947185516357422e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 0.9626749753952026, "learning_rate": 1.3141161509114034e-06, "loss": 0.5332, "mean_token_accuracy": 0.8309202790260315, "num_tokens": 11820587.0, "step": 311 }, { "epoch": 0.03968960692023916, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 5.0067901611328125e-06, "ewc_loss_parallel": 3.516674041748047e-06, "grad_norm": 0.9387598633766174, "learning_rate": 1.3183552352691819e-06, "loss": 0.5397, "mean_token_accuracy": 0.8319798707962036, "num_tokens": 11854473.0, "step": 312 }, { "epoch": 0.039816817198829665, "ewc_loss": 0.00396728515625, "ewc_loss_diag": 5.036592483520508e-06, "ewc_loss_parallel": 3.4421682357788086e-06, "grad_norm": 0.9000072479248047, "learning_rate": 1.3225943196269605e-06, "loss": 0.5783, "mean_token_accuracy": 0.8239061236381531, "num_tokens": 11894744.0, "step": 313 }, { "epoch": 0.03994402747742017, "ewc_loss": 0.003875732421875, "ewc_loss_diag": 5.0961971282958984e-06, "ewc_loss_parallel": 3.3527612686157227e-06, "grad_norm": 0.896507203578949, "learning_rate": 1.3268334039847394e-06, "loss": 0.539, "mean_token_accuracy": 0.8306354880332947, "num_tokens": 11932346.0, "step": 314 }, { "epoch": 0.04007123775601069, "ewc_loss": 0.00390625, "ewc_loss_diag": 5.0961971282958984e-06, "ewc_loss_parallel": 3.3974647521972656e-06, "grad_norm": 0.9501348733901978, "learning_rate": 1.331072488342518e-06, "loss": 0.5976, "mean_token_accuracy": 0.8107591867446899, "num_tokens": 11967683.0, "step": 315 }, { "epoch": 0.040198448034601195, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 5.185604095458984e-06, "ewc_loss_parallel": 3.56137752532959e-06, "grad_norm": 0.9025946855545044, "learning_rate": 1.335311572700297e-06, "loss": 0.5842, "mean_token_accuracy": 0.8159042596817017, "num_tokens": 12010521.0, "step": 316 }, { "epoch": 0.04032565831319171, "ewc_loss": 0.003936767578125, "ewc_loss_diag": 5.245208740234375e-06, "ewc_loss_parallel": 3.4123659133911133e-06, "grad_norm": 0.8968826532363892, "learning_rate": 1.3395506570580754e-06, "loss": 0.5555, "mean_token_accuracy": 0.8263534903526306, "num_tokens": 12053317.0, "step": 317 }, { "epoch": 0.040452868591782216, "ewc_loss": 0.003997802734375, "ewc_loss_diag": 5.245208740234375e-06, "ewc_loss_parallel": 3.4570693969726562e-06, "grad_norm": 1.0413833856582642, "learning_rate": 1.3437897414158543e-06, "loss": 0.6128, "mean_token_accuracy": 0.8091039657592773, "num_tokens": 12081589.0, "step": 318 }, { "epoch": 0.040580078870372724, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 5.245208740234375e-06, "ewc_loss_parallel": 3.769993782043457e-06, "grad_norm": 0.9757505655288696, "learning_rate": 1.348028825773633e-06, "loss": 0.5531, "mean_token_accuracy": 0.8250447511672974, "num_tokens": 12114826.0, "step": 319 }, { "epoch": 0.04070728914896324, "ewc_loss": 0.004150390625, "ewc_loss_diag": 5.27501106262207e-06, "ewc_loss_parallel": 3.6209821701049805e-06, "grad_norm": 0.9304830431938171, "learning_rate": 1.3522679101314119e-06, "loss": 0.5344, "mean_token_accuracy": 0.8318167924880981, "num_tokens": 12150769.0, "step": 320 }, { "epoch": 0.040834499427553746, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 5.27501106262207e-06, "ewc_loss_parallel": 3.5017728805541992e-06, "grad_norm": 0.9159615635871887, "learning_rate": 1.3565069944891903e-06, "loss": 0.5556, "mean_token_accuracy": 0.8275966048240662, "num_tokens": 12189035.0, "step": 321 }, { "epoch": 0.04096170970614425, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 5.334615707397461e-06, "ewc_loss_parallel": 3.516674041748047e-06, "grad_norm": 0.9307323694229126, "learning_rate": 1.3607460788469692e-06, "loss": 0.5458, "mean_token_accuracy": 0.8284123539924622, "num_tokens": 12225974.0, "step": 322 }, { "epoch": 0.04108891998473477, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 5.364418029785156e-06, "ewc_loss_parallel": 3.56137752532959e-06, "grad_norm": 0.9319155216217041, "learning_rate": 1.364985163204748e-06, "loss": 0.614, "mean_token_accuracy": 0.8096347451210022, "num_tokens": 12266971.0, "step": 323 }, { "epoch": 0.041216130263325275, "ewc_loss": 0.00408935546875, "ewc_loss_diag": 5.364418029785156e-06, "ewc_loss_parallel": 3.56137752532959e-06, "grad_norm": 0.937536358833313, "learning_rate": 1.3692242475625268e-06, "loss": 0.5257, "mean_token_accuracy": 0.8405312299728394, "num_tokens": 12306365.0, "step": 324 }, { "epoch": 0.04134334054191579, "ewc_loss": 0.004180908203125, "ewc_loss_diag": 5.364418029785156e-06, "ewc_loss_parallel": 3.635883331298828e-06, "grad_norm": 0.9282424449920654, "learning_rate": 1.3734633319203052e-06, "loss": 0.58, "mean_token_accuracy": 0.8288992643356323, "num_tokens": 12345343.0, "step": 325 }, { "epoch": 0.0414705508205063, "ewc_loss": 0.004119873046875, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 3.5762786865234375e-06, "grad_norm": 0.9560063481330872, "learning_rate": 1.3777024162780841e-06, "loss": 0.6453, "mean_token_accuracy": 0.8007923364639282, "num_tokens": 12384532.0, "step": 326 }, { "epoch": 0.041597761099096804, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 5.3942203521728516e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 0.9128920435905457, "learning_rate": 1.3819415006358628e-06, "loss": 0.5085, "mean_token_accuracy": 0.8405840396881104, "num_tokens": 12424955.0, "step": 327 }, { "epoch": 0.04172497137768732, "ewc_loss": 0.004119873046875, "ewc_loss_diag": 5.424022674560547e-06, "ewc_loss_parallel": 3.5762786865234375e-06, "grad_norm": 0.8681803941726685, "learning_rate": 1.3861805849936417e-06, "loss": 0.5288, "mean_token_accuracy": 0.837011992931366, "num_tokens": 12469583.0, "step": 328 }, { "epoch": 0.041852181656277826, "ewc_loss": 0.0040283203125, "ewc_loss_diag": 5.453824996948242e-06, "ewc_loss_parallel": 3.471970558166504e-06, "grad_norm": 0.9632992148399353, "learning_rate": 1.3904196693514201e-06, "loss": 0.575, "mean_token_accuracy": 0.8171857595443726, "num_tokens": 12504997.0, "step": 329 }, { "epoch": 0.04197939193486834, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 5.453824996948242e-06, "ewc_loss_parallel": 3.7401914596557617e-06, "grad_norm": 0.9429887533187866, "learning_rate": 1.394658753709199e-06, "loss": 0.5767, "mean_token_accuracy": 0.8222098350524902, "num_tokens": 12546248.0, "step": 330 }, { "epoch": 0.04210660221345885, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 5.453824996948242e-06, "ewc_loss_parallel": 3.7401914596557617e-06, "grad_norm": 0.9644829034805298, "learning_rate": 1.3988978380669777e-06, "loss": 0.5948, "mean_token_accuracy": 0.8143887519836426, "num_tokens": 12580741.0, "step": 331 }, { "epoch": 0.042233812492049355, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 5.4836273193359375e-06, "ewc_loss_parallel": 3.725290298461914e-06, "grad_norm": 0.922808825969696, "learning_rate": 1.4031369224247562e-06, "loss": 0.601, "mean_token_accuracy": 0.8136541843414307, "num_tokens": 12622539.0, "step": 332 }, { "epoch": 0.04236102277063987, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 5.513429641723633e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 0.956436038017273, "learning_rate": 1.407376006782535e-06, "loss": 0.5558, "mean_token_accuracy": 0.8251458406448364, "num_tokens": 12660137.0, "step": 333 }, { "epoch": 0.04248823304923038, "ewc_loss": 0.004302978515625, "ewc_loss_diag": 5.543231964111328e-06, "ewc_loss_parallel": 3.7401914596557617e-06, "grad_norm": 0.9250050783157349, "learning_rate": 1.4116150911403137e-06, "loss": 0.5822, "mean_token_accuracy": 0.8189293146133423, "num_tokens": 12702341.0, "step": 334 }, { "epoch": 0.04261544332782089, "ewc_loss": 0.004180908203125, "ewc_loss_diag": 5.543231964111328e-06, "ewc_loss_parallel": 3.635883331298828e-06, "grad_norm": 0.9185926914215088, "learning_rate": 1.4158541754980926e-06, "loss": 0.5239, "mean_token_accuracy": 0.8367151021957397, "num_tokens": 12739560.0, "step": 335 }, { "epoch": 0.0427426536064114, "ewc_loss": 0.00421142578125, "ewc_loss_diag": 5.5730342864990234e-06, "ewc_loss_parallel": 3.6656856536865234e-06, "grad_norm": 0.9220436215400696, "learning_rate": 1.420093259855871e-06, "loss": 0.5605, "mean_token_accuracy": 0.8311991691589355, "num_tokens": 12781534.0, "step": 336 }, { "epoch": 0.042869863885001906, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 5.602836608886719e-06, "ewc_loss_parallel": 3.7103891372680664e-06, "grad_norm": 0.9691652059555054, "learning_rate": 1.42433234421365e-06, "loss": 0.5657, "mean_token_accuracy": 0.8232496976852417, "num_tokens": 12815840.0, "step": 337 }, { "epoch": 0.04299707416359242, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 5.632638931274414e-06, "ewc_loss_parallel": 3.7997961044311523e-06, "grad_norm": 0.9775698781013489, "learning_rate": 1.4285714285714286e-06, "loss": 0.5739, "mean_token_accuracy": 0.8223233222961426, "num_tokens": 12849201.0, "step": 338 }, { "epoch": 0.04312428444218293, "ewc_loss": 0.00439453125, "ewc_loss_diag": 5.632638931274414e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 0.971390426158905, "learning_rate": 1.4328105129292075e-06, "loss": 0.5781, "mean_token_accuracy": 0.8219374418258667, "num_tokens": 12888490.0, "step": 339 }, { "epoch": 0.043251494720773435, "ewc_loss": 0.004364013671875, "ewc_loss_diag": 5.632638931274414e-06, "ewc_loss_parallel": 3.7997961044311523e-06, "grad_norm": 0.9276496171951294, "learning_rate": 1.437049597286986e-06, "loss": 0.5526, "mean_token_accuracy": 0.828153133392334, "num_tokens": 12927554.0, "step": 340 }, { "epoch": 0.04337870499936395, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 5.632638931274414e-06, "ewc_loss_parallel": 3.6954879760742188e-06, "grad_norm": 1.0311098098754883, "learning_rate": 1.4412886816447649e-06, "loss": 0.5983, "mean_token_accuracy": 0.8123565912246704, "num_tokens": 12960801.0, "step": 341 }, { "epoch": 0.04350591527795446, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 5.632638931274414e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 0.9883745312690735, "learning_rate": 1.4455277660025435e-06, "loss": 0.5171, "mean_token_accuracy": 0.8360710740089417, "num_tokens": 12993119.0, "step": 342 }, { "epoch": 0.04363312555654497, "ewc_loss": 0.00439453125, "ewc_loss_diag": 5.662441253662109e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 0.9190604090690613, "learning_rate": 1.4497668503603224e-06, "loss": 0.5769, "mean_token_accuracy": 0.8206131458282471, "num_tokens": 13033966.0, "step": 343 }, { "epoch": 0.04376033583513548, "ewc_loss": 0.0042724609375, "ewc_loss_diag": 5.662441253662109e-06, "ewc_loss_parallel": 3.7103891372680664e-06, "grad_norm": 1.0133501291275024, "learning_rate": 1.4540059347181009e-06, "loss": 0.5545, "mean_token_accuracy": 0.8275653123855591, "num_tokens": 13063649.0, "step": 344 }, { "epoch": 0.043887546113725986, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 5.692243576049805e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 0.9465197920799255, "learning_rate": 1.4582450190758798e-06, "loss": 0.6077, "mean_token_accuracy": 0.8108482360839844, "num_tokens": 13106926.0, "step": 345 }, { "epoch": 0.0440147563923165, "ewc_loss": 0.00439453125, "ewc_loss_diag": 5.692243576049805e-06, "ewc_loss_parallel": 3.814697265625e-06, "grad_norm": 0.9379767179489136, "learning_rate": 1.4624841034336584e-06, "loss": 0.5531, "mean_token_accuracy": 0.8279849290847778, "num_tokens": 13142961.0, "step": 346 }, { "epoch": 0.04414196667090701, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 5.751848220825195e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 0.927746593952179, "learning_rate": 1.4667231877914373e-06, "loss": 0.5342, "mean_token_accuracy": 0.8322364687919617, "num_tokens": 13183544.0, "step": 347 }, { "epoch": 0.04426917694949752, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 5.751848220825195e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 0.9353781938552856, "learning_rate": 1.470962272149216e-06, "loss": 0.5508, "mean_token_accuracy": 0.8266153335571289, "num_tokens": 13222609.0, "step": 348 }, { "epoch": 0.04439638722808803, "ewc_loss": 0.004425048828125, "ewc_loss_diag": 5.811452865600586e-06, "ewc_loss_parallel": 3.844499588012695e-06, "grad_norm": 0.966180682182312, "learning_rate": 1.4752013565069945e-06, "loss": 0.5493, "mean_token_accuracy": 0.8288819193840027, "num_tokens": 13258267.0, "step": 349 }, { "epoch": 0.04452359750667854, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 5.811452865600586e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 0.9788193702697754, "learning_rate": 1.4794404408647733e-06, "loss": 0.5563, "mean_token_accuracy": 0.8274585008621216, "num_tokens": 13294681.0, "step": 350 }, { "epoch": 0.04465080778526905, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 5.811452865600586e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 0.9237056970596313, "learning_rate": 1.483679525222552e-06, "loss": 0.5037, "mean_token_accuracy": 0.8395869135856628, "num_tokens": 13335469.0, "step": 351 }, { "epoch": 0.04477801806385956, "ewc_loss": 0.00445556640625, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 3.874301910400391e-06, "grad_norm": 0.9328157305717468, "learning_rate": 1.487918609580331e-06, "loss": 0.4835, "mean_token_accuracy": 0.8430542349815369, "num_tokens": 13372465.0, "step": 352 }, { "epoch": 0.04490522834245007, "ewc_loss": 0.004486083984375, "ewc_loss_diag": 5.841255187988281e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 0.9474906921386719, "learning_rate": 1.4921576939381094e-06, "loss": 0.5326, "mean_token_accuracy": 0.835374116897583, "num_tokens": 13409218.0, "step": 353 }, { "epoch": 0.04503243862104058, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 0.9244088530540466, "learning_rate": 1.4963967782958882e-06, "loss": 0.5012, "mean_token_accuracy": 0.8420341610908508, "num_tokens": 13452548.0, "step": 354 }, { "epoch": 0.04515964889963109, "ewc_loss": 0.004486083984375, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 0.9899604320526123, "learning_rate": 1.500635862653667e-06, "loss": 0.5886, "mean_token_accuracy": 0.8186887502670288, "num_tokens": 13488325.0, "step": 355 }, { "epoch": 0.0452868591782216, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 0.955748438835144, "learning_rate": 1.5048749470114458e-06, "loss": 0.5887, "mean_token_accuracy": 0.818039059638977, "num_tokens": 13528203.0, "step": 356 }, { "epoch": 0.04541406945681211, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 0.9275600910186768, "learning_rate": 1.5091140313692243e-06, "loss": 0.5158, "mean_token_accuracy": 0.8368891477584839, "num_tokens": 13567242.0, "step": 357 }, { "epoch": 0.04554127973540262, "ewc_loss": 0.004486083984375, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 3.904104232788086e-06, "grad_norm": 0.9416757822036743, "learning_rate": 1.5133531157270031e-06, "loss": 0.5671, "mean_token_accuracy": 0.8192784786224365, "num_tokens": 13604521.0, "step": 358 }, { "epoch": 0.04566849001399313, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 5.8710575103759766e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 0.9470186233520508, "learning_rate": 1.5175922000847818e-06, "loss": 0.5283, "mean_token_accuracy": 0.8354102969169617, "num_tokens": 13642787.0, "step": 359 }, { "epoch": 0.04579570029258364, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 5.900859832763672e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 0.9501407146453857, "learning_rate": 1.5218312844425607e-06, "loss": 0.5529, "mean_token_accuracy": 0.8209468126296997, "num_tokens": 13684113.0, "step": 360 }, { "epoch": 0.045922910571174154, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 3.993511199951172e-06, "grad_norm": 0.9563867449760437, "learning_rate": 1.5260703688003392e-06, "loss": 0.4837, "mean_token_accuracy": 0.8433014750480652, "num_tokens": 13719208.0, "step": 361 }, { "epoch": 0.04605012084976466, "ewc_loss": 0.004547119140625, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 0.9216500520706177, "learning_rate": 1.530309453158118e-06, "loss": 0.5045, "mean_token_accuracy": 0.840766191482544, "num_tokens": 13760159.0, "step": 362 }, { "epoch": 0.04617733112835517, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 0.9747579097747803, "learning_rate": 1.5345485375158967e-06, "loss": 0.5919, "mean_token_accuracy": 0.8150867819786072, "num_tokens": 13798375.0, "step": 363 }, { "epoch": 0.04630454140694568, "ewc_loss": 0.004608154296875, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 0.91876620054245, "learning_rate": 1.5387876218736756e-06, "loss": 0.4698, "mean_token_accuracy": 0.850543200969696, "num_tokens": 13841924.0, "step": 364 }, { "epoch": 0.04643175168553619, "ewc_loss": 0.0045166015625, "ewc_loss_diag": 5.930662155151367e-06, "ewc_loss_parallel": 3.933906555175781e-06, "grad_norm": 0.9837459921836853, "learning_rate": 1.543026706231454e-06, "loss": 0.5797, "mean_token_accuracy": 0.8219932317733765, "num_tokens": 13881436.0, "step": 365 }, { "epoch": 0.0465589619641267, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 5.990266799926758e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 1.0052608251571655, "learning_rate": 1.547265790589233e-06, "loss": 0.5925, "mean_token_accuracy": 0.8154741525650024, "num_tokens": 13916295.0, "step": 366 }, { "epoch": 0.04668617224271721, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 5.990266799926758e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 0.9843584299087524, "learning_rate": 1.5515048749470116e-06, "loss": 0.5567, "mean_token_accuracy": 0.8251436352729797, "num_tokens": 13950189.0, "step": 367 }, { "epoch": 0.04681338252130772, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 0.9523493647575378, "learning_rate": 1.55574395930479e-06, "loss": 0.5633, "mean_token_accuracy": 0.824354887008667, "num_tokens": 13988422.0, "step": 368 }, { "epoch": 0.046940592799898234, "ewc_loss": 0.00457763671875, "ewc_loss_diag": 6.020069122314453e-06, "ewc_loss_parallel": 3.9637088775634766e-06, "grad_norm": 0.9318180680274963, "learning_rate": 1.559983043662569e-06, "loss": 0.5606, "mean_token_accuracy": 0.8252021670341492, "num_tokens": 14029199.0, "step": 369 }, { "epoch": 0.04706780307848874, "ewc_loss": 0.004638671875, "ewc_loss_diag": 6.079673767089844e-06, "ewc_loss_parallel": 4.023313522338867e-06, "grad_norm": 0.9701226949691772, "learning_rate": 1.5642221280203476e-06, "loss": 0.5609, "mean_token_accuracy": 0.8271967768669128, "num_tokens": 14067976.0, "step": 370 }, { "epoch": 0.04719501335707925, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 6.139278411865234e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 0.9653925895690918, "learning_rate": 1.5684612123781265e-06, "loss": 0.4846, "mean_token_accuracy": 0.8477668166160583, "num_tokens": 14102858.0, "step": 371 }, { "epoch": 0.04732222363566976, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 6.139278411865234e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 0.9959263801574707, "learning_rate": 1.572700296735905e-06, "loss": 0.5643, "mean_token_accuracy": 0.8293384313583374, "num_tokens": 14138560.0, "step": 372 }, { "epoch": 0.04744943391426027, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 6.16908073425293e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 1.0430067777633667, "learning_rate": 1.5769393810936839e-06, "loss": 0.5534, "mean_token_accuracy": 0.8261268138885498, "num_tokens": 14168007.0, "step": 373 }, { "epoch": 0.047576644192850785, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 6.198883056640625e-06, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 0.981804370880127, "learning_rate": 1.5811784654514625e-06, "loss": 0.4827, "mean_token_accuracy": 0.8491011261940002, "num_tokens": 14201779.0, "step": 374 }, { "epoch": 0.04770385447144129, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 6.198883056640625e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 0.9392017126083374, "learning_rate": 1.5854175498092414e-06, "loss": 0.6113, "mean_token_accuracy": 0.8094732761383057, "num_tokens": 14245917.0, "step": 375 }, { "epoch": 0.0478310647500318, "ewc_loss": 0.004669189453125, "ewc_loss_diag": 6.198883056640625e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 0.9644087553024292, "learning_rate": 1.5896566341670199e-06, "loss": 0.5687, "mean_token_accuracy": 0.8202846050262451, "num_tokens": 14284032.0, "step": 376 }, { "epoch": 0.047958275028622314, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 6.258487701416016e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 1.0146124362945557, "learning_rate": 1.5938957185247988e-06, "loss": 0.6181, "mean_token_accuracy": 0.8043134212493896, "num_tokens": 14322077.0, "step": 377 }, { "epoch": 0.04808548530721282, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 6.258487701416016e-06, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 0.9833000302314758, "learning_rate": 1.5981348028825774e-06, "loss": 0.5672, "mean_token_accuracy": 0.8222323060035706, "num_tokens": 14358031.0, "step": 378 }, { "epoch": 0.048212695585803336, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 6.258487701416016e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 1.0565592050552368, "learning_rate": 1.6023738872403563e-06, "loss": 0.6564, "mean_token_accuracy": 0.7946299314498901, "num_tokens": 14390230.0, "step": 379 }, { "epoch": 0.048339905864393844, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 6.288290023803711e-06, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 0.9768320322036743, "learning_rate": 1.6066129715981348e-06, "loss": 0.5085, "mean_token_accuracy": 0.8402426242828369, "num_tokens": 14427394.0, "step": 380 }, { "epoch": 0.04846711614298435, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 6.288290023803711e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 0.9397717714309692, "learning_rate": 1.6108520559559137e-06, "loss": 0.5052, "mean_token_accuracy": 0.8404301404953003, "num_tokens": 14466606.0, "step": 381 }, { "epoch": 0.048594326421574865, "ewc_loss": 0.00469970703125, "ewc_loss_diag": 6.288290023803711e-06, "ewc_loss_parallel": 4.0531158447265625e-06, "grad_norm": 1.046960473060608, "learning_rate": 1.6150911403136924e-06, "loss": 0.5708, "mean_token_accuracy": 0.8245800137519836, "num_tokens": 14494795.0, "step": 382 }, { "epoch": 0.04872153670016537, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 6.318092346191406e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 0.9619308710098267, "learning_rate": 1.6193302246714712e-06, "loss": 0.5728, "mean_token_accuracy": 0.822637677192688, "num_tokens": 14537400.0, "step": 383 }, { "epoch": 0.04884874697875588, "ewc_loss": 0.004730224609375, "ewc_loss_diag": 6.318092346191406e-06, "ewc_loss_parallel": 4.082918167114258e-06, "grad_norm": 0.9988415837287903, "learning_rate": 1.62356930902925e-06, "loss": 0.5908, "mean_token_accuracy": 0.8229455947875977, "num_tokens": 14575215.0, "step": 384 }, { "epoch": 0.048975957257346395, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 6.377696990966797e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.0256948471069336, "learning_rate": 1.6278083933870286e-06, "loss": 0.6141, "mean_token_accuracy": 0.806985080242157, "num_tokens": 14608486.0, "step": 385 }, { "epoch": 0.0491031675359369, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 6.377696990966797e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 0.9809779524803162, "learning_rate": 1.6320474777448073e-06, "loss": 0.5433, "mean_token_accuracy": 0.8283756375312805, "num_tokens": 14644583.0, "step": 386 }, { "epoch": 0.049230377814527417, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 6.377696990966797e-06, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 0.9838175177574158, "learning_rate": 1.636286562102586e-06, "loss": 0.5665, "mean_token_accuracy": 0.8209415674209595, "num_tokens": 14682976.0, "step": 387 }, { "epoch": 0.049357588093117924, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 6.407499313354492e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 1.100310206413269, "learning_rate": 1.6405256464603648e-06, "loss": 0.5387, "mean_token_accuracy": 0.826938271522522, "num_tokens": 14717926.0, "step": 388 }, { "epoch": 0.04948479837170843, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 6.4373016357421875e-06, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 1.0489610433578491, "learning_rate": 1.6447647308181433e-06, "loss": 0.5789, "mean_token_accuracy": 0.8184577226638794, "num_tokens": 14750484.0, "step": 389 }, { "epoch": 0.049612008650298946, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 0.9779059886932373, "learning_rate": 1.6490038151759222e-06, "loss": 0.5964, "mean_token_accuracy": 0.8149688243865967, "num_tokens": 14790044.0, "step": 390 }, { "epoch": 0.04973921892888945, "ewc_loss": 0.0047607421875, "ewc_loss_diag": 6.467103958129883e-06, "ewc_loss_parallel": 4.112720489501953e-06, "grad_norm": 0.9534146785736084, "learning_rate": 1.6532428995337008e-06, "loss": 0.518, "mean_token_accuracy": 0.8349287509918213, "num_tokens": 14828768.0, "step": 391 }, { "epoch": 0.04986642920747997, "ewc_loss": 0.004791259765625, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 0.9701762795448303, "learning_rate": 1.6574819838914797e-06, "loss": 0.5062, "mean_token_accuracy": 0.8449842929840088, "num_tokens": 14865848.0, "step": 392 }, { "epoch": 0.049993639486070475, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 6.496906280517578e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 0.9809802174568176, "learning_rate": 1.6617210682492582e-06, "loss": 0.5434, "mean_token_accuracy": 0.8286948800086975, "num_tokens": 14903973.0, "step": 393 }, { "epoch": 0.05012084976466098, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 6.5267086029052734e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 0.9925172328948975, "learning_rate": 1.665960152607037e-06, "loss": 0.5782, "mean_token_accuracy": 0.8210346698760986, "num_tokens": 14944958.0, "step": 394 }, { "epoch": 0.0502480600432515, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 6.5267086029052734e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 1.0057950019836426, "learning_rate": 1.6701992369648157e-06, "loss": 0.5691, "mean_token_accuracy": 0.8200975656509399, "num_tokens": 14985924.0, "step": 395 }, { "epoch": 0.050375270321842004, "ewc_loss": 0.0048828125, "ewc_loss_diag": 6.556510925292969e-06, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 1.010898232460022, "learning_rate": 1.6744383213225946e-06, "loss": 0.5348, "mean_token_accuracy": 0.8298236131668091, "num_tokens": 15022346.0, "step": 396 }, { "epoch": 0.05050248060043251, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 6.586313247680664e-06, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 0.9601777195930481, "learning_rate": 1.678677405680373e-06, "loss": 0.4864, "mean_token_accuracy": 0.84346604347229, "num_tokens": 15058656.0, "step": 397 }, { "epoch": 0.050629690879023026, "ewc_loss": 0.00482177734375, "ewc_loss_diag": 6.616115570068359e-06, "ewc_loss_parallel": 4.1425228118896484e-06, "grad_norm": 0.9702800512313843, "learning_rate": 1.682916490038152e-06, "loss": 0.5209, "mean_token_accuracy": 0.8332322835922241, "num_tokens": 15094165.0, "step": 398 }, { "epoch": 0.05075690115761353, "ewc_loss": 0.0048828125, "ewc_loss_diag": 6.645917892456055e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 0.9402665495872498, "learning_rate": 1.6871555743959306e-06, "loss": 0.5023, "mean_token_accuracy": 0.8385355472564697, "num_tokens": 15136979.0, "step": 399 }, { "epoch": 0.05088411143620405, "ewc_loss": 0.0048828125, "ewc_loss_diag": 6.645917892456055e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 0.9588683843612671, "learning_rate": 1.6913946587537095e-06, "loss": 0.5282, "mean_token_accuracy": 0.83326256275177, "num_tokens": 15178765.0, "step": 400 }, { "epoch": 0.051011321714794555, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 6.705522537231445e-06, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 0.9906892776489258, "learning_rate": 1.695633743111488e-06, "loss": 0.5299, "mean_token_accuracy": 0.8348373174667358, "num_tokens": 15215892.0, "step": 401 }, { "epoch": 0.05113853199338506, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 6.705522537231445e-06, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 1.0099822282791138, "learning_rate": 1.6998728274692669e-06, "loss": 0.5017, "mean_token_accuracy": 0.8436505198478699, "num_tokens": 15248693.0, "step": 402 }, { "epoch": 0.05126574227197558, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 6.735324859619141e-06, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 0.9840804934501648, "learning_rate": 1.7041119118270455e-06, "loss": 0.5262, "mean_token_accuracy": 0.8351743817329407, "num_tokens": 15289476.0, "step": 403 }, { "epoch": 0.051392952550566084, "ewc_loss": 0.004913330078125, "ewc_loss_diag": 6.765127182006836e-06, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 1.018567442893982, "learning_rate": 1.708350996184824e-06, "loss": 0.5637, "mean_token_accuracy": 0.8222864866256714, "num_tokens": 15323032.0, "step": 404 }, { "epoch": 0.0515201628291566, "ewc_loss": 0.004974365234375, "ewc_loss_diag": 6.765127182006836e-06, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 0.9600975513458252, "learning_rate": 1.7125900805426029e-06, "loss": 0.5939, "mean_token_accuracy": 0.8137240409851074, "num_tokens": 15369155.0, "step": 405 }, { "epoch": 0.051647373107747106, "ewc_loss": 0.004852294921875, "ewc_loss_diag": 6.8247318267822266e-06, "ewc_loss_parallel": 4.172325134277344e-06, "grad_norm": 0.9914801716804504, "learning_rate": 1.7168291649003816e-06, "loss": 0.5112, "mean_token_accuracy": 0.8350880742073059, "num_tokens": 15402623.0, "step": 406 }, { "epoch": 0.051774583386337614, "ewc_loss": 0.0050048828125, "ewc_loss_diag": 6.854534149169922e-06, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 0.997972846031189, "learning_rate": 1.7210682492581604e-06, "loss": 0.5499, "mean_token_accuracy": 0.8280729055404663, "num_tokens": 15440485.0, "step": 407 }, { "epoch": 0.05190179366492813, "ewc_loss": 0.004974365234375, "ewc_loss_diag": 6.9141387939453125e-06, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 0.990582287311554, "learning_rate": 1.725307333615939e-06, "loss": 0.52, "mean_token_accuracy": 0.8349148631095886, "num_tokens": 15479641.0, "step": 408 }, { "epoch": 0.052029003943518635, "ewc_loss": 0.004974365234375, "ewc_loss_diag": 6.9141387939453125e-06, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 1.0000076293945312, "learning_rate": 1.7295464179737178e-06, "loss": 0.5617, "mean_token_accuracy": 0.8247374296188354, "num_tokens": 15519085.0, "step": 409 }, { "epoch": 0.05215621422210915, "ewc_loss": 0.004974365234375, "ewc_loss_diag": 6.943941116333008e-06, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 0.9782423973083496, "learning_rate": 1.7337855023314965e-06, "loss": 0.4958, "mean_token_accuracy": 0.842944324016571, "num_tokens": 15557642.0, "step": 410 }, { "epoch": 0.05228342450069966, "ewc_loss": 0.004974365234375, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 4.26173210144043e-06, "grad_norm": 0.9848113059997559, "learning_rate": 1.7380245866892753e-06, "loss": 0.496, "mean_token_accuracy": 0.8441821336746216, "num_tokens": 15593178.0, "step": 411 }, { "epoch": 0.052410634779290165, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 7.0035457611083984e-06, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 0.9921701550483704, "learning_rate": 1.7422636710470538e-06, "loss": 0.551, "mean_token_accuracy": 0.826546311378479, "num_tokens": 15630167.0, "step": 412 }, { "epoch": 0.05253784505788068, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 7.033348083496094e-06, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 0.9959946870803833, "learning_rate": 1.7465027554048327e-06, "loss": 0.5816, "mean_token_accuracy": 0.8197736740112305, "num_tokens": 15669380.0, "step": 413 }, { "epoch": 0.05266505533647119, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 7.063150405883789e-06, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 1.0340772867202759, "learning_rate": 1.7507418397626114e-06, "loss": 0.5238, "mean_token_accuracy": 0.836074948310852, "num_tokens": 15701791.0, "step": 414 }, { "epoch": 0.052792265615061694, "ewc_loss": 0.005096435546875, "ewc_loss_diag": 7.063150405883789e-06, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 1.089056134223938, "learning_rate": 1.7549809241203902e-06, "loss": 0.4991, "mean_token_accuracy": 0.840770959854126, "num_tokens": 15738223.0, "step": 415 }, { "epoch": 0.05291947589365221, "ewc_loss": 0.005126953125, "ewc_loss_diag": 7.092952728271484e-06, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 1.2543132305145264, "learning_rate": 1.759220008478169e-06, "loss": 0.5806, "mean_token_accuracy": 0.8198636770248413, "num_tokens": 15773159.0, "step": 416 }, { "epoch": 0.053046686172242716, "ewc_loss": 0.005340576171875, "ewc_loss_diag": 7.092952728271484e-06, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 1.061112880706787, "learning_rate": 1.7634590928359476e-06, "loss": 0.5465, "mean_token_accuracy": 0.8289180994033813, "num_tokens": 15813132.0, "step": 417 }, { "epoch": 0.05317389645083323, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 7.12275505065918e-06, "ewc_loss_parallel": 4.231929779052734e-06, "grad_norm": 0.9669850468635559, "learning_rate": 1.7676981771937263e-06, "loss": 0.5273, "mean_token_accuracy": 0.831460177898407, "num_tokens": 15849752.0, "step": 418 }, { "epoch": 0.05330110672942374, "ewc_loss": 0.00494384765625, "ewc_loss_diag": 7.152557373046875e-06, "ewc_loss_parallel": 4.202127456665039e-06, "grad_norm": 0.9973790645599365, "learning_rate": 1.7719372615515052e-06, "loss": 0.5069, "mean_token_accuracy": 0.8421788215637207, "num_tokens": 15885331.0, "step": 419 }, { "epoch": 0.053428317008014245, "ewc_loss": 0.005126953125, "ewc_loss_diag": 7.152557373046875e-06, "ewc_loss_parallel": 4.380941390991211e-06, "grad_norm": 1.1242388486862183, "learning_rate": 1.7761763459092838e-06, "loss": 0.5473, "mean_token_accuracy": 0.8321453332901001, "num_tokens": 15924649.0, "step": 420 }, { "epoch": 0.05355552728660476, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 7.212162017822266e-06, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 1.0252511501312256, "learning_rate": 1.7804154302670625e-06, "loss": 0.5172, "mean_token_accuracy": 0.8345901370048523, "num_tokens": 15959342.0, "step": 421 }, { "epoch": 0.05368273756519527, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 7.212162017822266e-06, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 0.9651816487312317, "learning_rate": 1.7846545146248412e-06, "loss": 0.5043, "mean_token_accuracy": 0.8404257297515869, "num_tokens": 15998479.0, "step": 422 }, { "epoch": 0.05380994784378578, "ewc_loss": 0.005035400390625, "ewc_loss_diag": 7.212162017822266e-06, "ewc_loss_parallel": 4.291534423828125e-06, "grad_norm": 1.0153402090072632, "learning_rate": 1.7888935989826198e-06, "loss": 0.5476, "mean_token_accuracy": 0.8262083530426025, "num_tokens": 16034197.0, "step": 423 }, { "epoch": 0.05393715812237629, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 7.271766662597656e-06, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 1.0476797819137573, "learning_rate": 1.7931326833403987e-06, "loss": 0.561, "mean_token_accuracy": 0.8214957118034363, "num_tokens": 16071448.0, "step": 424 }, { "epoch": 0.054064368400966796, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 7.271766662597656e-06, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 0.9515451788902283, "learning_rate": 1.7973717676981772e-06, "loss": 0.4856, "mean_token_accuracy": 0.8484388589859009, "num_tokens": 16116007.0, "step": 425 }, { "epoch": 0.05419157867955731, "ewc_loss": 0.00506591796875, "ewc_loss_diag": 7.271766662597656e-06, "ewc_loss_parallel": 4.32133674621582e-06, "grad_norm": 1.0162022113800049, "learning_rate": 1.801610852055956e-06, "loss": 0.5208, "mean_token_accuracy": 0.8376238346099854, "num_tokens": 16158070.0, "step": 426 }, { "epoch": 0.05431878895814782, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 1.0165425539016724, "learning_rate": 1.8058499364137347e-06, "loss": 0.5611, "mean_token_accuracy": 0.8271646499633789, "num_tokens": 16196310.0, "step": 427 }, { "epoch": 0.054445999236738325, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 0.9825855493545532, "learning_rate": 1.8100890207715136e-06, "loss": 0.6076, "mean_token_accuracy": 0.8071213960647583, "num_tokens": 16237927.0, "step": 428 }, { "epoch": 0.05457320951532884, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 1.0237067937850952, "learning_rate": 1.814328105129292e-06, "loss": 0.5623, "mean_token_accuracy": 0.8214062452316284, "num_tokens": 16274865.0, "step": 429 }, { "epoch": 0.05470041979391935, "ewc_loss": 0.005218505859375, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 4.500150680541992e-06, "grad_norm": 1.0862033367156982, "learning_rate": 1.818567189487071e-06, "loss": 0.5392, "mean_token_accuracy": 0.8254191875457764, "num_tokens": 16302503.0, "step": 430 }, { "epoch": 0.05482763007250986, "ewc_loss": 0.00531005859375, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 1.0217574834823608, "learning_rate": 1.8228062738448496e-06, "loss": 0.5937, "mean_token_accuracy": 0.8151404857635498, "num_tokens": 16342472.0, "step": 431 }, { "epoch": 0.05495484035110037, "ewc_loss": 0.005157470703125, "ewc_loss_diag": 7.3015689849853516e-06, "ewc_loss_parallel": 4.410743713378906e-06, "grad_norm": 1.001373529434204, "learning_rate": 1.8270453582026285e-06, "loss": 0.5534, "mean_token_accuracy": 0.8250479698181152, "num_tokens": 16378423.0, "step": 432 }, { "epoch": 0.055082050629690876, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 7.331371307373047e-06, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 0.9943616390228271, "learning_rate": 1.831284442560407e-06, "loss": 0.5228, "mean_token_accuracy": 0.8361390829086304, "num_tokens": 16418430.0, "step": 433 }, { "epoch": 0.05520926090828139, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 7.331371307373047e-06, "ewc_loss_parallel": 4.4405460357666016e-06, "grad_norm": 0.9992492198944092, "learning_rate": 1.8355235269181859e-06, "loss": 0.5353, "mean_token_accuracy": 0.8319023847579956, "num_tokens": 16460984.0, "step": 434 }, { "epoch": 0.0553364711868719, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 7.331371307373047e-06, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 1.0342319011688232, "learning_rate": 1.8397626112759646e-06, "loss": 0.5134, "mean_token_accuracy": 0.8378477096557617, "num_tokens": 16493460.0, "step": 435 }, { "epoch": 0.05546368146546241, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 7.331371307373047e-06, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 1.0194101333618164, "learning_rate": 1.8440016956337434e-06, "loss": 0.4916, "mean_token_accuracy": 0.8426316976547241, "num_tokens": 16528998.0, "step": 436 }, { "epoch": 0.05559089174405292, "ewc_loss": 0.0052490234375, "ewc_loss_diag": 7.331371307373047e-06, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 1.0291619300842285, "learning_rate": 1.848240779991522e-06, "loss": 0.5168, "mean_token_accuracy": 0.8363533616065979, "num_tokens": 16564041.0, "step": 437 }, { "epoch": 0.05571810202264343, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 7.331371307373047e-06, "ewc_loss_parallel": 4.559755325317383e-06, "grad_norm": 1.0715659856796265, "learning_rate": 1.8524798643493008e-06, "loss": 0.59, "mean_token_accuracy": 0.8130837082862854, "num_tokens": 16597967.0, "step": 438 }, { "epoch": 0.05584531230123394, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 1.0276618003845215, "learning_rate": 1.8567189487070795e-06, "loss": 0.5745, "mean_token_accuracy": 0.8215750455856323, "num_tokens": 16636756.0, "step": 439 }, { "epoch": 0.05597252257982445, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 0.9835512638092041, "learning_rate": 1.860958033064858e-06, "loss": 0.4869, "mean_token_accuracy": 0.8438262939453125, "num_tokens": 16675811.0, "step": 440 }, { "epoch": 0.05609973285841496, "ewc_loss": 0.00518798828125, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 4.470348358154297e-06, "grad_norm": 1.023652195930481, "learning_rate": 1.8651971174226368e-06, "loss": 0.5707, "mean_token_accuracy": 0.8231430053710938, "num_tokens": 16713714.0, "step": 441 }, { "epoch": 0.05622694313700547, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 1.00806725025177, "learning_rate": 1.8694362017804155e-06, "loss": 0.5268, "mean_token_accuracy": 0.8333024978637695, "num_tokens": 16754317.0, "step": 442 }, { "epoch": 0.05635415341559598, "ewc_loss": 0.00531005859375, "ewc_loss_diag": 7.3909759521484375e-06, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 0.9992324709892273, "learning_rate": 1.8736752861381944e-06, "loss": 0.4881, "mean_token_accuracy": 0.8456978797912598, "num_tokens": 16793189.0, "step": 443 }, { "epoch": 0.05648136369418649, "ewc_loss": 0.00531005859375, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 1.032397747039795, "learning_rate": 1.8779143704959728e-06, "loss": 0.4977, "mean_token_accuracy": 0.8414134979248047, "num_tokens": 16825046.0, "step": 444 }, { "epoch": 0.056608573972777, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 7.450580596923828e-06, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 1.0666254758834839, "learning_rate": 1.8821534548537517e-06, "loss": 0.5603, "mean_token_accuracy": 0.8257905840873718, "num_tokens": 16857769.0, "step": 445 }, { "epoch": 0.05673578425136751, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 7.510185241699219e-06, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 1.005755066871643, "learning_rate": 1.8863925392115304e-06, "loss": 0.5183, "mean_token_accuracy": 0.8381471633911133, "num_tokens": 16900878.0, "step": 446 }, { "epoch": 0.05686299452995802, "ewc_loss": 0.005279541015625, "ewc_loss_diag": 7.510185241699219e-06, "ewc_loss_parallel": 4.5299530029296875e-06, "grad_norm": 0.9980320334434509, "learning_rate": 1.8906316235693093e-06, "loss": 0.5324, "mean_token_accuracy": 0.8306676149368286, "num_tokens": 16942832.0, "step": 447 }, { "epoch": 0.05699020480854853, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 7.569789886474609e-06, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 1.056121826171875, "learning_rate": 1.8948707079270877e-06, "loss": 0.5362, "mean_token_accuracy": 0.8246848583221436, "num_tokens": 16975609.0, "step": 448 }, { "epoch": 0.057117415087139044, "ewc_loss": 0.0054931640625, "ewc_loss_diag": 7.569789886474609e-06, "ewc_loss_parallel": 4.738569259643555e-06, "grad_norm": 1.0298386812210083, "learning_rate": 1.8991097922848666e-06, "loss": 0.5071, "mean_token_accuracy": 0.8389702439308167, "num_tokens": 17012620.0, "step": 449 }, { "epoch": 0.05724462536572955, "ewc_loss": 0.005401611328125, "ewc_loss_diag": 7.62939453125e-06, "ewc_loss_parallel": 4.649162292480469e-06, "grad_norm": 1.0196189880371094, "learning_rate": 1.9033488766426453e-06, "loss": 0.5517, "mean_token_accuracy": 0.8232125043869019, "num_tokens": 17053820.0, "step": 450 }, { "epoch": 0.05737183564432006, "ewc_loss": 0.005462646484375, "ewc_loss_diag": 7.62939453125e-06, "ewc_loss_parallel": 4.708766937255859e-06, "grad_norm": 1.0261677503585815, "learning_rate": 1.907587961000424e-06, "loss": 0.5053, "mean_token_accuracy": 0.8429585695266724, "num_tokens": 17091885.0, "step": 451 }, { "epoch": 0.05749904592291057, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 1.0204086303710938, "learning_rate": 1.911827045358203e-06, "loss": 0.552, "mean_token_accuracy": 0.8267094492912292, "num_tokens": 17133432.0, "step": 452 }, { "epoch": 0.05762625620150108, "ewc_loss": 0.00543212890625, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 4.678964614868164e-06, "grad_norm": 1.0806487798690796, "learning_rate": 1.9160661297159817e-06, "loss": 0.574, "mean_token_accuracy": 0.8221526145935059, "num_tokens": 17171845.0, "step": 453 }, { "epoch": 0.057753466480091595, "ewc_loss": 0.005706787109375, "ewc_loss_diag": 7.748603820800781e-06, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 1.0805507898330688, "learning_rate": 1.92030521407376e-06, "loss": 0.5639, "mean_token_accuracy": 0.8204793930053711, "num_tokens": 17206137.0, "step": 454 }, { "epoch": 0.0578806767586821, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 4.798173904418945e-06, "grad_norm": 1.002273440361023, "learning_rate": 1.924544298431539e-06, "loss": 0.5436, "mean_token_accuracy": 0.8320398926734924, "num_tokens": 17247619.0, "step": 455 }, { "epoch": 0.05800788703727261, "ewc_loss": 0.00537109375, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 4.589557647705078e-06, "grad_norm": 1.0717040300369263, "learning_rate": 1.9287833827893175e-06, "loss": 0.5651, "mean_token_accuracy": 0.8214725852012634, "num_tokens": 17285053.0, "step": 456 }, { "epoch": 0.058135097315863124, "ewc_loss": 0.005706787109375, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 1.0631096363067627, "learning_rate": 1.9330224671470964e-06, "loss": 0.5398, "mean_token_accuracy": 0.8284243941307068, "num_tokens": 17320303.0, "step": 457 }, { "epoch": 0.05826230759445363, "ewc_loss": 0.00555419921875, "ewc_loss_diag": 7.867813110351562e-06, "ewc_loss_parallel": 4.76837158203125e-06, "grad_norm": 1.070616364479065, "learning_rate": 1.9372615515048753e-06, "loss": 0.5907, "mean_token_accuracy": 0.8195558190345764, "num_tokens": 17354102.0, "step": 458 }, { "epoch": 0.05838951787304414, "ewc_loss": 0.005645751953125, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 4.857778549194336e-06, "grad_norm": 1.0413622856140137, "learning_rate": 1.9415006358626538e-06, "loss": 0.5303, "mean_token_accuracy": 0.8328633904457092, "num_tokens": 17389958.0, "step": 459 }, { "epoch": 0.05851672815163465, "ewc_loss": 0.005584716796875, "ewc_loss_diag": 7.987022399902344e-06, "ewc_loss_parallel": 4.798173904418945e-06, "grad_norm": 1.074398398399353, "learning_rate": 1.9457397202204326e-06, "loss": 0.5177, "mean_token_accuracy": 0.8352367877960205, "num_tokens": 17421488.0, "step": 460 }, { "epoch": 0.05864393843022516, "ewc_loss": 0.005706787109375, "ewc_loss_diag": 8.046627044677734e-06, "ewc_loss_parallel": 4.9173831939697266e-06, "grad_norm": 1.0380693674087524, "learning_rate": 1.949978804578211e-06, "loss": 0.4801, "mean_token_accuracy": 0.8494483232498169, "num_tokens": 17455172.0, "step": 461 }, { "epoch": 0.058771148708815675, "ewc_loss": 0.005615234375, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 4.827976226806641e-06, "grad_norm": 1.0308091640472412, "learning_rate": 1.95421788893599e-06, "loss": 0.5784, "mean_token_accuracy": 0.8174746036529541, "num_tokens": 17494798.0, "step": 462 }, { "epoch": 0.05889835898740618, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 4.887580871582031e-06, "grad_norm": 1.0750916004180908, "learning_rate": 1.9584569732937684e-06, "loss": 0.5423, "mean_token_accuracy": 0.8342622518539429, "num_tokens": 17527619.0, "step": 463 }, { "epoch": 0.05902556926599669, "ewc_loss": 0.005859375, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 1.018432378768921, "learning_rate": 1.9626960576515473e-06, "loss": 0.4618, "mean_token_accuracy": 0.853185772895813, "num_tokens": 17564299.0, "step": 464 }, { "epoch": 0.059152779544587204, "ewc_loss": 0.00567626953125, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 4.857778549194336e-06, "grad_norm": 1.0615757703781128, "learning_rate": 1.9669351420093262e-06, "loss": 0.5772, "mean_token_accuracy": 0.8205696940422058, "num_tokens": 17598333.0, "step": 465 }, { "epoch": 0.05927998982317771, "ewc_loss": 0.005859375, "ewc_loss_diag": 8.106231689453125e-06, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 1.0798090696334839, "learning_rate": 1.971174226367105e-06, "loss": 0.5597, "mean_token_accuracy": 0.8270426392555237, "num_tokens": 17642106.0, "step": 466 }, { "epoch": 0.059407200101768226, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 8.225440979003906e-06, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 1.0442970991134644, "learning_rate": 1.9754133107248836e-06, "loss": 0.5016, "mean_token_accuracy": 0.8369603753089905, "num_tokens": 17676412.0, "step": 467 }, { "epoch": 0.059534410380358734, "ewc_loss": 0.005767822265625, "ewc_loss_diag": 8.225440979003906e-06, "ewc_loss_parallel": 4.947185516357422e-06, "grad_norm": 1.0488195419311523, "learning_rate": 1.9796523950826624e-06, "loss": 0.5836, "mean_token_accuracy": 0.8190939426422119, "num_tokens": 17717079.0, "step": 468 }, { "epoch": 0.05966162065894924, "ewc_loss": 0.005859375, "ewc_loss_diag": 8.225440979003906e-06, "ewc_loss_parallel": 5.036592483520508e-06, "grad_norm": 1.056382656097412, "learning_rate": 1.983891479440441e-06, "loss": 0.4742, "mean_token_accuracy": 0.8495392799377441, "num_tokens": 17753603.0, "step": 469 }, { "epoch": 0.059788830937539755, "ewc_loss": 0.005828857421875, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 5.0067901611328125e-06, "grad_norm": 1.017468810081482, "learning_rate": 1.98813056379822e-06, "loss": 0.5399, "mean_token_accuracy": 0.8311537504196167, "num_tokens": 17795960.0, "step": 470 }, { "epoch": 0.05991604121613026, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 1.0847804546356201, "learning_rate": 1.9923696481559983e-06, "loss": 0.5573, "mean_token_accuracy": 0.8260818719863892, "num_tokens": 17829298.0, "step": 471 }, { "epoch": 0.06004325149472077, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 8.285045623779297e-06, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 1.0497801303863525, "learning_rate": 1.996608732513777e-06, "loss": 0.5933, "mean_token_accuracy": 0.8152779936790466, "num_tokens": 17872027.0, "step": 472 }, { "epoch": 0.060170461773311285, "ewc_loss": 0.00579833984375, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 4.976987838745117e-06, "grad_norm": 1.0123696327209473, "learning_rate": 2.000847816871556e-06, "loss": 0.4805, "mean_token_accuracy": 0.8443265557289124, "num_tokens": 17911846.0, "step": 473 }, { "epoch": 0.06029767205190179, "ewc_loss": 0.005859375, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 5.0067901611328125e-06, "grad_norm": 1.0505648851394653, "learning_rate": 2.005086901229335e-06, "loss": 0.4759, "mean_token_accuracy": 0.8477553129196167, "num_tokens": 17946095.0, "step": 474 }, { "epoch": 0.060424882330492306, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 1.058504343032837, "learning_rate": 2.0093259855871134e-06, "loss": 0.5069, "mean_token_accuracy": 0.839136004447937, "num_tokens": 17983232.0, "step": 475 }, { "epoch": 0.060552092609082814, "ewc_loss": 0.0059814453125, "ewc_loss_diag": 8.404254913330078e-06, "ewc_loss_parallel": 5.155801773071289e-06, "grad_norm": 1.0566366910934448, "learning_rate": 2.0135650699448923e-06, "loss": 0.5959, "mean_token_accuracy": 0.8113133907318115, "num_tokens": 18022455.0, "step": 476 }, { "epoch": 0.06067930288767332, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 1.0698378086090088, "learning_rate": 2.0178041543026707e-06, "loss": 0.5611, "mean_token_accuracy": 0.8239954710006714, "num_tokens": 18059844.0, "step": 477 }, { "epoch": 0.060806513166263836, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 8.463859558105469e-06, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 1.066954493522644, "learning_rate": 2.022043238660449e-06, "loss": 0.6116, "mean_token_accuracy": 0.809255838394165, "num_tokens": 18099569.0, "step": 478 }, { "epoch": 0.06093372344485434, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 8.58306884765625e-06, "ewc_loss_parallel": 5.185604095458984e-06, "grad_norm": 1.075668215751648, "learning_rate": 2.026282323018228e-06, "loss": 0.5723, "mean_token_accuracy": 0.8201876282691956, "num_tokens": 18134747.0, "step": 479 }, { "epoch": 0.06106093372344486, "ewc_loss": 0.006072998046875, "ewc_loss_diag": 8.58306884765625e-06, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 1.0461561679840088, "learning_rate": 2.030521407376007e-06, "loss": 0.5801, "mean_token_accuracy": 0.8217867016792297, "num_tokens": 18173989.0, "step": 480 }, { "epoch": 0.061188144002035365, "ewc_loss": 0.00592041015625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 5.066394805908203e-06, "grad_norm": 1.1279969215393066, "learning_rate": 2.034760491733786e-06, "loss": 0.5527, "mean_token_accuracy": 0.8247318267822266, "num_tokens": 18205065.0, "step": 481 }, { "epoch": 0.06131535428062587, "ewc_loss": 0.006256103515625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 1.0450435876846313, "learning_rate": 2.0389995760915643e-06, "loss": 0.5013, "mean_token_accuracy": 0.8421359062194824, "num_tokens": 18242999.0, "step": 482 }, { "epoch": 0.06144256455921639, "ewc_loss": 0.005950927734375, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 5.0961971282958984e-06, "grad_norm": 1.057969093322754, "learning_rate": 2.043238660449343e-06, "loss": 0.5838, "mean_token_accuracy": 0.8175108432769775, "num_tokens": 18280642.0, "step": 483 }, { "epoch": 0.061569774837806894, "ewc_loss": 0.006072998046875, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 5.21540641784668e-06, "grad_norm": 1.045637845993042, "learning_rate": 2.0474777448071216e-06, "loss": 0.5072, "mean_token_accuracy": 0.8412634134292603, "num_tokens": 18317894.0, "step": 484 }, { "epoch": 0.0616969851163974, "ewc_loss": 0.006103515625, "ewc_loss_diag": 8.64267349243164e-06, "ewc_loss_parallel": 5.245208740234375e-06, "grad_norm": 1.0791369676589966, "learning_rate": 2.0517168291649005e-06, "loss": 0.536, "mean_token_accuracy": 0.8307623267173767, "num_tokens": 18350481.0, "step": 485 }, { "epoch": 0.061824195394987916, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 8.702278137207031e-06, "ewc_loss_parallel": 5.364418029785156e-06, "grad_norm": 1.1251139640808105, "learning_rate": 2.0559559135226794e-06, "loss": 0.5442, "mean_token_accuracy": 0.8245187997817993, "num_tokens": 18384390.0, "step": 486 }, { "epoch": 0.06195140567357842, "ewc_loss": 0.00634765625, "ewc_loss_diag": 8.702278137207031e-06, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 1.0364640951156616, "learning_rate": 2.060194997880458e-06, "loss": 0.5042, "mean_token_accuracy": 0.8375482559204102, "num_tokens": 18423958.0, "step": 487 }, { "epoch": 0.06207861595216894, "ewc_loss": 0.00604248046875, "ewc_loss_diag": 8.702278137207031e-06, "ewc_loss_parallel": 5.155801773071289e-06, "grad_norm": 1.0420405864715576, "learning_rate": 2.0644340822382368e-06, "loss": 0.4688, "mean_token_accuracy": 0.850232720375061, "num_tokens": 18459622.0, "step": 488 }, { "epoch": 0.062205826230759445, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 1.0751714706420898, "learning_rate": 2.0686731665960156e-06, "loss": 0.5259, "mean_token_accuracy": 0.836747407913208, "num_tokens": 18495568.0, "step": 489 }, { "epoch": 0.06233303650934995, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 1.0520814657211304, "learning_rate": 2.072912250953794e-06, "loss": 0.4925, "mean_token_accuracy": 0.8476216793060303, "num_tokens": 18534440.0, "step": 490 }, { "epoch": 0.06246024678794047, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 1.0635986328125, "learning_rate": 2.077151335311573e-06, "loss": 0.5285, "mean_token_accuracy": 0.8319230079650879, "num_tokens": 18570583.0, "step": 491 }, { "epoch": 0.06258745706653097, "ewc_loss": 0.00634765625, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 1.0511932373046875, "learning_rate": 2.0813904196693514e-06, "loss": 0.5065, "mean_token_accuracy": 0.8408750891685486, "num_tokens": 18611801.0, "step": 492 }, { "epoch": 0.06271466734512149, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 1.0873196125030518, "learning_rate": 2.0856295040271303e-06, "loss": 0.564, "mean_token_accuracy": 0.8197592496871948, "num_tokens": 18646985.0, "step": 493 }, { "epoch": 0.06284187762371199, "ewc_loss": 0.00634765625, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 1.0691421031951904, "learning_rate": 2.0898685883849092e-06, "loss": 0.5439, "mean_token_accuracy": 0.832901656627655, "num_tokens": 18688818.0, "step": 494 }, { "epoch": 0.0629690879023025, "ewc_loss": 0.006317138671875, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 1.055321455001831, "learning_rate": 2.0941076727426877e-06, "loss": 0.524, "mean_token_accuracy": 0.8362721800804138, "num_tokens": 18727208.0, "step": 495 }, { "epoch": 0.06309629818089302, "ewc_loss": 0.00628662109375, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 5.3942203521728516e-06, "grad_norm": 1.0160160064697266, "learning_rate": 2.0983467571004666e-06, "loss": 0.5087, "mean_token_accuracy": 0.8385360240936279, "num_tokens": 18772174.0, "step": 496 }, { "epoch": 0.06322350845948353, "ewc_loss": 0.0062255859375, "ewc_loss_diag": 8.821487426757812e-06, "ewc_loss_parallel": 5.334615707397461e-06, "grad_norm": 1.0640711784362793, "learning_rate": 2.102585841458245e-06, "loss": 0.5439, "mean_token_accuracy": 0.8290042281150818, "num_tokens": 18808111.0, "step": 497 }, { "epoch": 0.06335071873807403, "ewc_loss": 0.006378173828125, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 1.0880545377731323, "learning_rate": 2.106824925816024e-06, "loss": 0.5004, "mean_token_accuracy": 0.8356111645698547, "num_tokens": 18841912.0, "step": 498 }, { "epoch": 0.06347792901666455, "ewc_loss": 0.006378173828125, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 1.081151008605957, "learning_rate": 2.1110640101738024e-06, "loss": 0.5104, "mean_token_accuracy": 0.8404209613800049, "num_tokens": 18877694.0, "step": 499 }, { "epoch": 0.06360513929525506, "ewc_loss": 0.00634765625, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 1.0676864385604858, "learning_rate": 2.1153030945315813e-06, "loss": 0.5153, "mean_token_accuracy": 0.8369312286376953, "num_tokens": 18912253.0, "step": 500 }, { "epoch": 0.06373234957384556, "ewc_loss": 0.006317138671875, "ewc_loss_diag": 8.940696716308594e-06, "ewc_loss_parallel": 5.424022674560547e-06, "grad_norm": 1.0845892429351807, "learning_rate": 2.11954217888936e-06, "loss": 0.5165, "mean_token_accuracy": 0.8379516005516052, "num_tokens": 18945700.0, "step": 501 }, { "epoch": 0.06385955985243608, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 1.0660881996154785, "learning_rate": 2.123781263247139e-06, "loss": 0.5414, "mean_token_accuracy": 0.8288091421127319, "num_tokens": 18985260.0, "step": 502 }, { "epoch": 0.06398677013102659, "ewc_loss": 0.006378173828125, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 1.066961407661438, "learning_rate": 2.1280203476049175e-06, "loss": 0.6119, "mean_token_accuracy": 0.8057048320770264, "num_tokens": 19025364.0, "step": 503 }, { "epoch": 0.06411398040961709, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 1.0618044137954712, "learning_rate": 2.1322594319626964e-06, "loss": 0.5483, "mean_token_accuracy": 0.8268086910247803, "num_tokens": 19064262.0, "step": 504 }, { "epoch": 0.0642411906882076, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 9.059906005859375e-06, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 1.1373097896575928, "learning_rate": 2.136498516320475e-06, "loss": 0.5503, "mean_token_accuracy": 0.8290711045265198, "num_tokens": 19094780.0, "step": 505 }, { "epoch": 0.06436840096679812, "ewc_loss": 0.006622314453125, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 1.0877854824066162, "learning_rate": 2.1407376006782537e-06, "loss": 0.4889, "mean_token_accuracy": 0.8454602360725403, "num_tokens": 19131166.0, "step": 506 }, { "epoch": 0.06449561124538863, "ewc_loss": 0.006378173828125, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 1.0554099082946777, "learning_rate": 2.144976685036032e-06, "loss": 0.5025, "mean_token_accuracy": 0.8424124121665955, "num_tokens": 19167371.0, "step": 507 }, { "epoch": 0.06462282152397913, "ewc_loss": 0.006378173828125, "ewc_loss_diag": 9.119510650634766e-06, "ewc_loss_parallel": 5.453824996948242e-06, "grad_norm": 1.0433436632156372, "learning_rate": 2.149215769393811e-06, "loss": 0.5526, "mean_token_accuracy": 0.8262315392494202, "num_tokens": 19206751.0, "step": 508 }, { "epoch": 0.06475003180256965, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 1.0701924562454224, "learning_rate": 2.15345485375159e-06, "loss": 0.5081, "mean_token_accuracy": 0.8386794328689575, "num_tokens": 19243227.0, "step": 509 }, { "epoch": 0.06487724208116016, "ewc_loss": 0.006439208984375, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 1.0574562549591064, "learning_rate": 2.157693938109369e-06, "loss": 0.5607, "mean_token_accuracy": 0.826590895652771, "num_tokens": 19285050.0, "step": 510 }, { "epoch": 0.06500445235975066, "ewc_loss": 0.006439208984375, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 1.0454027652740479, "learning_rate": 2.1619330224671473e-06, "loss": 0.5448, "mean_token_accuracy": 0.8286300301551819, "num_tokens": 19328780.0, "step": 511 }, { "epoch": 0.06513166263834118, "ewc_loss": 0.00640869140625, "ewc_loss_diag": 9.179115295410156e-06, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 1.04952073097229, "learning_rate": 2.166172106824926e-06, "loss": 0.5406, "mean_token_accuracy": 0.8330482244491577, "num_tokens": 19373011.0, "step": 512 }, { "epoch": 0.06525887291693169, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 5.543231964111328e-06, "grad_norm": 1.0872387886047363, "learning_rate": 2.1704111911827046e-06, "loss": 0.5311, "mean_token_accuracy": 0.8310163617134094, "num_tokens": 19409641.0, "step": 513 }, { "epoch": 0.0653860831955222, "ewc_loss": 0.006591796875, "ewc_loss_diag": 9.298324584960938e-06, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 1.0750128030776978, "learning_rate": 2.174650275540483e-06, "loss": 0.5715, "mean_token_accuracy": 0.8242740631103516, "num_tokens": 19451285.0, "step": 514 }, { "epoch": 0.06551329347411271, "ewc_loss": 0.006500244140625, "ewc_loss_diag": 9.357929229736328e-06, "ewc_loss_parallel": 5.543231964111328e-06, "grad_norm": 1.093761920928955, "learning_rate": 2.178889359898262e-06, "loss": 0.5688, "mean_token_accuracy": 0.8244665265083313, "num_tokens": 19487584.0, "step": 515 }, { "epoch": 0.06564050375270322, "ewc_loss": 0.006561279296875, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 5.602836608886719e-06, "grad_norm": 1.063757061958313, "learning_rate": 2.183128444256041e-06, "loss": 0.577, "mean_token_accuracy": 0.823347270488739, "num_tokens": 19531200.0, "step": 516 }, { "epoch": 0.06576771403129372, "ewc_loss": 0.006439208984375, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 5.4836273193359375e-06, "grad_norm": 1.0587379932403564, "learning_rate": 2.1873675286138197e-06, "loss": 0.5765, "mean_token_accuracy": 0.8163148164749146, "num_tokens": 19573126.0, "step": 517 }, { "epoch": 0.06589492430988424, "ewc_loss": 0.006500244140625, "ewc_loss_diag": 9.417533874511719e-06, "ewc_loss_parallel": 5.543231964111328e-06, "grad_norm": 1.071769118309021, "learning_rate": 2.191606612971598e-06, "loss": 0.5911, "mean_token_accuracy": 0.8149240612983704, "num_tokens": 19613499.0, "step": 518 }, { "epoch": 0.06602213458847475, "ewc_loss": 0.00653076171875, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 5.5730342864990234e-06, "grad_norm": 1.0806423425674438, "learning_rate": 2.195845697329377e-06, "loss": 0.5802, "mean_token_accuracy": 0.8132896423339844, "num_tokens": 19652459.0, "step": 519 }, { "epoch": 0.06614934486706527, "ewc_loss": 0.006591796875, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 5.632638931274414e-06, "grad_norm": 1.0650044679641724, "learning_rate": 2.2000847816871556e-06, "loss": 0.462, "mean_token_accuracy": 0.8514019250869751, "num_tokens": 19691874.0, "step": 520 }, { "epoch": 0.06627655514565577, "ewc_loss": 0.0064697265625, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 5.513429641723633e-06, "grad_norm": 1.0544936656951904, "learning_rate": 2.2043238660449344e-06, "loss": 0.5108, "mean_token_accuracy": 0.8405937552452087, "num_tokens": 19731424.0, "step": 521 }, { "epoch": 0.06640376542424628, "ewc_loss": 0.006500244140625, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 5.543231964111328e-06, "grad_norm": 1.0643565654754639, "learning_rate": 2.2085629504027133e-06, "loss": 0.5042, "mean_token_accuracy": 0.837889552116394, "num_tokens": 19768530.0, "step": 522 }, { "epoch": 0.0665309757028368, "ewc_loss": 0.006561279296875, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 5.602836608886719e-06, "grad_norm": 1.081713318824768, "learning_rate": 2.2128020347604918e-06, "loss": 0.5393, "mean_token_accuracy": 0.8298978805541992, "num_tokens": 19807836.0, "step": 523 }, { "epoch": 0.0666581859814273, "ewc_loss": 0.00665283203125, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 5.692243576049805e-06, "grad_norm": 1.0661870241165161, "learning_rate": 2.2170411191182707e-06, "loss": 0.5186, "mean_token_accuracy": 0.835808515548706, "num_tokens": 19854338.0, "step": 524 }, { "epoch": 0.06678539626001781, "ewc_loss": 0.006622314453125, "ewc_loss_diag": 9.47713851928711e-06, "ewc_loss_parallel": 5.662441253662109e-06, "grad_norm": 1.1128077507019043, "learning_rate": 2.2212802034760496e-06, "loss": 0.5367, "mean_token_accuracy": 0.8285495042800903, "num_tokens": 19888469.0, "step": 525 }, { "epoch": 0.06691260653860832, "ewc_loss": 0.0068359375, "ewc_loss_diag": 9.59634780883789e-06, "ewc_loss_parallel": 5.900859832763672e-06, "grad_norm": 1.1552244424819946, "learning_rate": 2.225519287833828e-06, "loss": 0.587, "mean_token_accuracy": 0.8189747333526611, "num_tokens": 19925137.0, "step": 526 }, { "epoch": 0.06703981681719882, "ewc_loss": 0.00677490234375, "ewc_loss_diag": 9.59634780883789e-06, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 1.105138897895813, "learning_rate": 2.229758372191607e-06, "loss": 0.545, "mean_token_accuracy": 0.828352689743042, "num_tokens": 19959658.0, "step": 527 }, { "epoch": 0.06716702709578934, "ewc_loss": 0.006591796875, "ewc_loss_diag": 9.59634780883789e-06, "ewc_loss_parallel": 5.602836608886719e-06, "grad_norm": 1.0666946172714233, "learning_rate": 2.2339974565493854e-06, "loss": 0.5033, "mean_token_accuracy": 0.8405275940895081, "num_tokens": 19998702.0, "step": 528 }, { "epoch": 0.06729423737437985, "ewc_loss": 0.006591796875, "ewc_loss_diag": 9.59634780883789e-06, "ewc_loss_parallel": 5.602836608886719e-06, "grad_norm": 1.109465479850769, "learning_rate": 2.2382365409071642e-06, "loss": 0.5737, "mean_token_accuracy": 0.8234475255012512, "num_tokens": 20038406.0, "step": 529 }, { "epoch": 0.06742144765297035, "ewc_loss": 0.00677490234375, "ewc_loss_diag": 9.59634780883789e-06, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 1.1172058582305908, "learning_rate": 2.242475625264943e-06, "loss": 0.5292, "mean_token_accuracy": 0.8319068551063538, "num_tokens": 20073185.0, "step": 530 }, { "epoch": 0.06754865793156087, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 5.841255187988281e-06, "grad_norm": 1.099149465560913, "learning_rate": 2.246714709622722e-06, "loss": 0.5107, "mean_token_accuracy": 0.8398405313491821, "num_tokens": 20106828.0, "step": 531 }, { "epoch": 0.06767586821015138, "ewc_loss": 0.0068359375, "ewc_loss_diag": 9.655952453613281e-06, "ewc_loss_parallel": 5.8710575103759766e-06, "grad_norm": 1.1502959728240967, "learning_rate": 2.2509537939805005e-06, "loss": 0.5139, "mean_token_accuracy": 0.836593747138977, "num_tokens": 20137541.0, "step": 532 }, { "epoch": 0.0678030784887419, "ewc_loss": 0.007080078125, "ewc_loss_diag": 9.715557098388672e-06, "ewc_loss_parallel": 6.109476089477539e-06, "grad_norm": 1.1151293516159058, "learning_rate": 2.255192878338279e-06, "loss": 0.5272, "mean_token_accuracy": 0.8321871161460876, "num_tokens": 20174452.0, "step": 533 }, { "epoch": 0.0679302887673324, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 9.715557098388672e-06, "ewc_loss_parallel": 5.841255187988281e-06, "grad_norm": 1.1370834112167358, "learning_rate": 2.259431962696058e-06, "loss": 0.5432, "mean_token_accuracy": 0.8312787413597107, "num_tokens": 20212195.0, "step": 534 }, { "epoch": 0.06805749904592291, "ewc_loss": 0.007110595703125, "ewc_loss_diag": 9.715557098388672e-06, "ewc_loss_parallel": 6.139278411865234e-06, "grad_norm": 1.0784586668014526, "learning_rate": 2.2636710470538363e-06, "loss": 0.4572, "mean_token_accuracy": 0.8476442098617554, "num_tokens": 20248344.0, "step": 535 }, { "epoch": 0.06818470932451343, "ewc_loss": 0.00677490234375, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 5.811452865600586e-06, "grad_norm": 1.0514085292816162, "learning_rate": 2.267910131411615e-06, "loss": 0.5012, "mean_token_accuracy": 0.8421282768249512, "num_tokens": 20289874.0, "step": 536 }, { "epoch": 0.06831191960310393, "ewc_loss": 0.00689697265625, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 5.930662155151367e-06, "grad_norm": 1.1444813013076782, "learning_rate": 2.272149215769394e-06, "loss": 0.5132, "mean_token_accuracy": 0.834784746170044, "num_tokens": 20319582.0, "step": 537 }, { "epoch": 0.06843912988169444, "ewc_loss": 0.0074462890625, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 1.1832356452941895, "learning_rate": 2.276388300127173e-06, "loss": 0.4928, "mean_token_accuracy": 0.840286135673523, "num_tokens": 20356944.0, "step": 538 }, { "epoch": 0.06856634016028496, "ewc_loss": 0.007232666015625, "ewc_loss_diag": 9.775161743164062e-06, "ewc_loss_parallel": 6.258487701416016e-06, "grad_norm": 1.066515564918518, "learning_rate": 2.2806273844849514e-06, "loss": 0.545, "mean_token_accuracy": 0.8251410722732544, "num_tokens": 20398099.0, "step": 539 }, { "epoch": 0.06869355043887546, "ewc_loss": 0.006805419921875, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 5.841255187988281e-06, "grad_norm": 1.0932083129882812, "learning_rate": 2.2848664688427303e-06, "loss": 0.5165, "mean_token_accuracy": 0.833968997001648, "num_tokens": 20435155.0, "step": 540 }, { "epoch": 0.06882076071746597, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 6.22868537902832e-06, "grad_norm": 1.1642454862594604, "learning_rate": 2.2891055532005087e-06, "loss": 0.5753, "mean_token_accuracy": 0.8208611011505127, "num_tokens": 20468369.0, "step": 541 }, { "epoch": 0.06894797099605648, "ewc_loss": 0.007476806640625, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 1.157009243965149, "learning_rate": 2.2933446375582876e-06, "loss": 0.5507, "mean_token_accuracy": 0.8307543992996216, "num_tokens": 20500303.0, "step": 542 }, { "epoch": 0.06907518127464699, "ewc_loss": 0.007293701171875, "ewc_loss_diag": 9.834766387939453e-06, "ewc_loss_parallel": 6.318092346191406e-06, "grad_norm": 1.0872445106506348, "learning_rate": 2.297583721916066e-06, "loss": 0.503, "mean_token_accuracy": 0.8429348468780518, "num_tokens": 20541290.0, "step": 543 }, { "epoch": 0.0692023915532375, "ewc_loss": 0.00714111328125, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 1.0736334323883057, "learning_rate": 2.301822806273845e-06, "loss": 0.523, "mean_token_accuracy": 0.8361847400665283, "num_tokens": 20583486.0, "step": 544 }, { "epoch": 0.06932960183182801, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 9.953975677490234e-06, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 1.1468849182128906, "learning_rate": 2.306061890631624e-06, "loss": 0.5748, "mean_token_accuracy": 0.8229312896728516, "num_tokens": 20618655.0, "step": 545 }, { "epoch": 0.06945681211041853, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 6.616115570068359e-06, "grad_norm": 1.1094858646392822, "learning_rate": 2.3103009749894027e-06, "loss": 0.5283, "mean_token_accuracy": 0.834108829498291, "num_tokens": 20655733.0, "step": 546 }, { "epoch": 0.06958402238900903, "ewc_loss": 0.007171630859375, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 6.16908073425293e-06, "grad_norm": 1.06533682346344, "learning_rate": 2.314540059347181e-06, "loss": 0.556, "mean_token_accuracy": 0.825668215751648, "num_tokens": 20699338.0, "step": 547 }, { "epoch": 0.06971123266759954, "ewc_loss": 0.007293701171875, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 6.288290023803711e-06, "grad_norm": 1.139801263809204, "learning_rate": 2.31877914370496e-06, "loss": 0.5801, "mean_token_accuracy": 0.8180627822875977, "num_tokens": 20731841.0, "step": 548 }, { "epoch": 0.06983844294619006, "ewc_loss": 0.007659912109375, "ewc_loss_diag": 1.0013580322265625e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 1.1524499654769897, "learning_rate": 2.3230182280627385e-06, "loss": 0.502, "mean_token_accuracy": 0.8420140743255615, "num_tokens": 20761097.0, "step": 549 }, { "epoch": 0.06996565322478056, "ewc_loss": 0.007476806640625, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 1.091667652130127, "learning_rate": 2.3272573124205174e-06, "loss": 0.4775, "mean_token_accuracy": 0.8475596904754639, "num_tokens": 20797013.0, "step": 550 }, { "epoch": 0.07009286350337107, "ewc_loss": 0.0072021484375, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 6.198883056640625e-06, "grad_norm": 1.0966852903366089, "learning_rate": 2.331496396778296e-06, "loss": 0.5132, "mean_token_accuracy": 0.8343713283538818, "num_tokens": 20833727.0, "step": 551 }, { "epoch": 0.07022007378196159, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 1.0799068212509155, "learning_rate": 2.3357354811360748e-06, "loss": 0.5423, "mean_token_accuracy": 0.831688404083252, "num_tokens": 20877587.0, "step": 552 }, { "epoch": 0.07034728406055209, "ewc_loss": 0.00732421875, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 6.318092346191406e-06, "grad_norm": 1.1587200164794922, "learning_rate": 2.3399745654938537e-06, "loss": 0.5513, "mean_token_accuracy": 0.8261982202529907, "num_tokens": 20909623.0, "step": 553 }, { "epoch": 0.0704744943391426, "ewc_loss": 0.007781982421875, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 1.132014513015747, "learning_rate": 2.344213649851632e-06, "loss": 0.5332, "mean_token_accuracy": 0.8333286046981812, "num_tokens": 20946348.0, "step": 554 }, { "epoch": 0.07060170461773312, "ewc_loss": 0.0074462890625, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 6.4373016357421875e-06, "grad_norm": 1.0946305990219116, "learning_rate": 2.348452734209411e-06, "loss": 0.6134, "mean_token_accuracy": 0.8105435967445374, "num_tokens": 20989831.0, "step": 555 }, { "epoch": 0.07072891489632362, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 1.1179139614105225, "learning_rate": 2.3526918185671895e-06, "loss": 0.5614, "mean_token_accuracy": 0.8216352462768555, "num_tokens": 21027007.0, "step": 556 }, { "epoch": 0.07085612517491413, "ewc_loss": 0.007537841796875, "ewc_loss_diag": 1.0073184967041016e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 1.09519624710083, "learning_rate": 2.3569309029249684e-06, "loss": 0.51, "mean_token_accuracy": 0.8385171890258789, "num_tokens": 21066488.0, "step": 557 }, { "epoch": 0.07098333545350465, "ewc_loss": 0.0074462890625, "ewc_loss_diag": 1.0192394256591797e-05, "ewc_loss_parallel": 6.407499313354492e-06, "grad_norm": 1.1090803146362305, "learning_rate": 2.3611699872827472e-06, "loss": 0.5994, "mean_token_accuracy": 0.8133143782615662, "num_tokens": 21109049.0, "step": 558 }, { "epoch": 0.07111054573209516, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.0192394256591797e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 1.1232670545578003, "learning_rate": 2.3654090716405257e-06, "loss": 0.5428, "mean_token_accuracy": 0.829046368598938, "num_tokens": 21146661.0, "step": 559 }, { "epoch": 0.07123775601068566, "ewc_loss": 0.007568359375, "ewc_loss_diag": 1.0192394256591797e-05, "ewc_loss_parallel": 6.5267086029052734e-06, "grad_norm": 1.1439908742904663, "learning_rate": 2.3696481559983046e-06, "loss": 0.5596, "mean_token_accuracy": 0.8212586045265198, "num_tokens": 21179568.0, "step": 560 }, { "epoch": 0.07136496628927617, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.0192394256591797e-05, "ewc_loss_parallel": 6.67572021484375e-06, "grad_norm": 1.1321450471878052, "learning_rate": 2.3738872403560835e-06, "loss": 0.5625, "mean_token_accuracy": 0.8224833011627197, "num_tokens": 21216535.0, "step": 561 }, { "epoch": 0.07149217656786669, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.0192394256591797e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 1.1174923181533813, "learning_rate": 2.378126324713862e-06, "loss": 0.5236, "mean_token_accuracy": 0.8329854011535645, "num_tokens": 21253887.0, "step": 562 }, { "epoch": 0.07161938684645719, "ewc_loss": 0.00750732421875, "ewc_loss_diag": 1.0192394256591797e-05, "ewc_loss_parallel": 6.467103958129883e-06, "grad_norm": 1.1173458099365234, "learning_rate": 2.382365409071641e-06, "loss": 0.5285, "mean_token_accuracy": 0.8336948752403259, "num_tokens": 21289813.0, "step": 563 }, { "epoch": 0.0717465971250477, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.0192394256591797e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 1.1047126054763794, "learning_rate": 2.3866044934294193e-06, "loss": 0.4809, "mean_token_accuracy": 0.8459264636039734, "num_tokens": 21326121.0, "step": 564 }, { "epoch": 0.07187380740363822, "ewc_loss": 0.007537841796875, "ewc_loss_diag": 1.0311603546142578e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 1.125774621963501, "learning_rate": 2.390843577787198e-06, "loss": 0.5745, "mean_token_accuracy": 0.8186521530151367, "num_tokens": 21363644.0, "step": 565 }, { "epoch": 0.07200101768222872, "ewc_loss": 0.0076904296875, "ewc_loss_diag": 1.0311603546142578e-05, "ewc_loss_parallel": 6.645917892456055e-06, "grad_norm": 1.074968934059143, "learning_rate": 2.395082662144977e-06, "loss": 0.4604, "mean_token_accuracy": 0.8547539710998535, "num_tokens": 21407119.0, "step": 566 }, { "epoch": 0.07212822796081923, "ewc_loss": 0.007415771484375, "ewc_loss_diag": 1.0371208190917969e-05, "ewc_loss_parallel": 6.377696990966797e-06, "grad_norm": 1.1064039468765259, "learning_rate": 2.399321746502756e-06, "loss": 0.5346, "mean_token_accuracy": 0.8328756093978882, "num_tokens": 21447648.0, "step": 567 }, { "epoch": 0.07225543823940975, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.0371208190917969e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 1.1354035139083862, "learning_rate": 2.4035608308605344e-06, "loss": 0.5758, "mean_token_accuracy": 0.8270009756088257, "num_tokens": 21488335.0, "step": 568 }, { "epoch": 0.07238264851800025, "ewc_loss": 0.007781982421875, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 1.1517012119293213, "learning_rate": 2.407799915218313e-06, "loss": 0.5783, "mean_token_accuracy": 0.8170046806335449, "num_tokens": 21525987.0, "step": 569 }, { "epoch": 0.07250985879659076, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 6.765127182006836e-06, "grad_norm": 1.1025432348251343, "learning_rate": 2.4120389995760917e-06, "loss": 0.4704, "mean_token_accuracy": 0.8542169332504272, "num_tokens": 21564671.0, "step": 570 }, { "epoch": 0.07263706907518128, "ewc_loss": 0.007537841796875, "ewc_loss_diag": 1.043081283569336e-05, "ewc_loss_parallel": 6.496906280517578e-06, "grad_norm": 1.0820344686508179, "learning_rate": 2.41627808393387e-06, "loss": 0.5165, "mean_token_accuracy": 0.8344590663909912, "num_tokens": 21608590.0, "step": 571 }, { "epoch": 0.07276427935377179, "ewc_loss": 0.00762939453125, "ewc_loss_diag": 1.049041748046875e-05, "ewc_loss_parallel": 6.586313247680664e-06, "grad_norm": 1.1562683582305908, "learning_rate": 2.420517168291649e-06, "loss": 0.5376, "mean_token_accuracy": 0.8345513343811035, "num_tokens": 21641684.0, "step": 572 }, { "epoch": 0.07289148963236229, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.049041748046875e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 1.128851294517517, "learning_rate": 2.424756252649428e-06, "loss": 0.4782, "mean_token_accuracy": 0.8466323614120483, "num_tokens": 21674923.0, "step": 573 }, { "epoch": 0.0730186999109528, "ewc_loss": 0.00775146484375, "ewc_loss_diag": 1.049041748046875e-05, "ewc_loss_parallel": 6.705522537231445e-06, "grad_norm": 1.1186221837997437, "learning_rate": 2.428995337007207e-06, "loss": 0.5639, "mean_token_accuracy": 0.8255631327629089, "num_tokens": 21714006.0, "step": 574 }, { "epoch": 0.07314591018954332, "ewc_loss": 0.00787353515625, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 6.794929504394531e-06, "grad_norm": 1.1376763582229614, "learning_rate": 2.4332344213649853e-06, "loss": 0.5163, "mean_token_accuracy": 0.8416990041732788, "num_tokens": 21750580.0, "step": 575 }, { "epoch": 0.07327312046813382, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 6.854534149169922e-06, "grad_norm": 1.1252774000167847, "learning_rate": 2.437473505722764e-06, "loss": 0.4914, "mean_token_accuracy": 0.8430631160736084, "num_tokens": 21787400.0, "step": 576 }, { "epoch": 0.07340033074672433, "ewc_loss": 0.0078125, "ewc_loss_diag": 1.0609626770019531e-05, "ewc_loss_parallel": 6.735324859619141e-06, "grad_norm": 1.1322160959243774, "learning_rate": 2.4417125900805427e-06, "loss": 0.507, "mean_token_accuracy": 0.8347032070159912, "num_tokens": 21823159.0, "step": 577 }, { "epoch": 0.07352754102531485, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.0728836059570312e-05, "ewc_loss_parallel": 6.854534149169922e-06, "grad_norm": 1.1479475498199463, "learning_rate": 2.4459516744383215e-06, "loss": 0.5459, "mean_token_accuracy": 0.8296146988868713, "num_tokens": 21858067.0, "step": 578 }, { "epoch": 0.07365475130390535, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.0728836059570312e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 1.129799485206604, "learning_rate": 2.4501907587961e-06, "loss": 0.5057, "mean_token_accuracy": 0.8378758430480957, "num_tokens": 21893332.0, "step": 579 }, { "epoch": 0.07378196158249586, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.0728836059570312e-05, "ewc_loss_parallel": 6.854534149169922e-06, "grad_norm": 1.1422706842422485, "learning_rate": 2.454429843153879e-06, "loss": 0.6173, "mean_token_accuracy": 0.8072946071624756, "num_tokens": 21932696.0, "step": 580 }, { "epoch": 0.07390917186108638, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.0728836059570312e-05, "ewc_loss_parallel": 6.9141387939453125e-06, "grad_norm": 1.1651626825332642, "learning_rate": 2.4586689275116578e-06, "loss": 0.5301, "mean_token_accuracy": 0.8268688917160034, "num_tokens": 21964825.0, "step": 581 }, { "epoch": 0.07403638213967688, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 6.943941116333008e-06, "grad_norm": 1.1235666275024414, "learning_rate": 2.4629080118694367e-06, "loss": 0.4916, "mean_token_accuracy": 0.8430242538452148, "num_tokens": 22003396.0, "step": 582 }, { "epoch": 0.0741635924182674, "ewc_loss": 0.0079345703125, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 6.8247318267822266e-06, "grad_norm": 1.109004259109497, "learning_rate": 2.467147096227215e-06, "loss": 0.4988, "mean_token_accuracy": 0.8426738977432251, "num_tokens": 22045107.0, "step": 583 }, { "epoch": 0.07429080269685791, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 1.1251252889633179, "learning_rate": 2.471386180584994e-06, "loss": 0.4737, "mean_token_accuracy": 0.8499193787574768, "num_tokens": 22082754.0, "step": 584 }, { "epoch": 0.07441801297544842, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.0788440704345703e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 1.1279560327529907, "learning_rate": 2.4756252649427725e-06, "loss": 0.4937, "mean_token_accuracy": 0.8422837257385254, "num_tokens": 22122501.0, "step": 585 }, { "epoch": 0.07454522325403892, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.0848045349121094e-05, "ewc_loss_parallel": 6.9141387939453125e-06, "grad_norm": 1.1318025588989258, "learning_rate": 2.4798643493005513e-06, "loss": 0.5312, "mean_token_accuracy": 0.834469199180603, "num_tokens": 22158110.0, "step": 586 }, { "epoch": 0.07467243353262944, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.0848045349121094e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 1.15122652053833, "learning_rate": 2.48410343365833e-06, "loss": 0.4817, "mean_token_accuracy": 0.8414551615715027, "num_tokens": 22190097.0, "step": 587 }, { "epoch": 0.07479964381121995, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.0848045349121094e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 1.1518243551254272, "learning_rate": 2.4883425180161087e-06, "loss": 0.5565, "mean_token_accuracy": 0.8257361650466919, "num_tokens": 22227568.0, "step": 588 }, { "epoch": 0.07492685408981045, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.0848045349121094e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 1.1495676040649414, "learning_rate": 2.4925816023738876e-06, "loss": 0.5187, "mean_token_accuracy": 0.839045524597168, "num_tokens": 22265376.0, "step": 589 }, { "epoch": 0.07505406436840097, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.0907649993896484e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 1.1091313362121582, "learning_rate": 2.496820686731666e-06, "loss": 0.5422, "mean_token_accuracy": 0.83158940076828, "num_tokens": 22310398.0, "step": 590 }, { "epoch": 0.07518127464699148, "ewc_loss": 0.00799560546875, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 6.884336471557617e-06, "grad_norm": 1.1503173112869263, "learning_rate": 2.5010597710894445e-06, "loss": 0.4482, "mean_token_accuracy": 0.8569149971008301, "num_tokens": 22341413.0, "step": 591 }, { "epoch": 0.07530848492558198, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 1.128411054611206, "learning_rate": 2.5052988554472234e-06, "loss": 0.5227, "mean_token_accuracy": 0.8350212574005127, "num_tokens": 22383228.0, "step": 592 }, { "epoch": 0.0754356952041725, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 1.1372060775756836, "learning_rate": 2.5095379398050023e-06, "loss": 0.542, "mean_token_accuracy": 0.8309032917022705, "num_tokens": 22423750.0, "step": 593 }, { "epoch": 0.07556290548276301, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 7.0035457611083984e-06, "grad_norm": 1.1824944019317627, "learning_rate": 2.513777024162781e-06, "loss": 0.4987, "mean_token_accuracy": 0.8408281803131104, "num_tokens": 22454717.0, "step": 594 }, { "epoch": 0.07569011576135352, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 1.126251220703125, "learning_rate": 2.5180161085205596e-06, "loss": 0.4986, "mean_token_accuracy": 0.8431358337402344, "num_tokens": 22492798.0, "step": 595 }, { "epoch": 0.07581732603994402, "ewc_loss": 0.008056640625, "ewc_loss_diag": 1.0967254638671875e-05, "ewc_loss_parallel": 6.973743438720703e-06, "grad_norm": 1.1378052234649658, "learning_rate": 2.5222551928783385e-06, "loss": 0.542, "mean_token_accuracy": 0.8305040597915649, "num_tokens": 22530774.0, "step": 596 }, { "epoch": 0.07594453631853454, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.1026859283447266e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 1.143975019454956, "learning_rate": 2.5264942772361174e-06, "loss": 0.5213, "mean_token_accuracy": 0.8350588083267212, "num_tokens": 22566355.0, "step": 597 }, { "epoch": 0.07607174659712505, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 1.1256003379821777, "learning_rate": 2.5307333615938963e-06, "loss": 0.5067, "mean_token_accuracy": 0.8420118093490601, "num_tokens": 22603633.0, "step": 598 }, { "epoch": 0.07619895687571555, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.1086463928222656e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 1.151041865348816, "learning_rate": 2.5349724459516743e-06, "loss": 0.5359, "mean_token_accuracy": 0.8335263729095459, "num_tokens": 22641481.0, "step": 599 }, { "epoch": 0.07632616715430607, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.1146068572998047e-05, "ewc_loss_parallel": 7.092952728271484e-06, "grad_norm": 1.1635955572128296, "learning_rate": 2.539211530309453e-06, "loss": 0.4789, "mean_token_accuracy": 0.8480397462844849, "num_tokens": 22672853.0, "step": 600 }, { "epoch": 0.07645337743289658, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 1.1770484447479248, "learning_rate": 2.543450614667232e-06, "loss": 0.5522, "mean_token_accuracy": 0.8248027563095093, "num_tokens": 22704851.0, "step": 601 }, { "epoch": 0.07658058771148708, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 1.1653443574905396, "learning_rate": 2.547689699025011e-06, "loss": 0.5713, "mean_token_accuracy": 0.820326030254364, "num_tokens": 22742578.0, "step": 602 }, { "epoch": 0.0767077979900776, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 1.1140375137329102, "learning_rate": 2.5519287833827894e-06, "loss": 0.5132, "mean_token_accuracy": 0.834548830986023, "num_tokens": 22783960.0, "step": 603 }, { "epoch": 0.07683500826866811, "ewc_loss": 0.0081787109375, "ewc_loss_diag": 1.1265277862548828e-05, "ewc_loss_parallel": 7.063150405883789e-06, "grad_norm": 1.1729861497879028, "learning_rate": 2.5561678677405683e-06, "loss": 0.4753, "mean_token_accuracy": 0.8465027809143066, "num_tokens": 22813643.0, "step": 604 }, { "epoch": 0.07696221854725861, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 1.1643257141113281, "learning_rate": 2.560406952098347e-06, "loss": 0.5657, "mean_token_accuracy": 0.8200214505195618, "num_tokens": 22850290.0, "step": 605 }, { "epoch": 0.07708942882584913, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.138448715209961e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 1.1476263999938965, "learning_rate": 2.5646460364561252e-06, "loss": 0.5194, "mean_token_accuracy": 0.8331431150436401, "num_tokens": 22894009.0, "step": 606 }, { "epoch": 0.07721663910443964, "ewc_loss": 0.00823974609375, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 7.12275505065918e-06, "grad_norm": 1.1693706512451172, "learning_rate": 2.568885120813904e-06, "loss": 0.5184, "mean_token_accuracy": 0.8353092670440674, "num_tokens": 22931418.0, "step": 607 }, { "epoch": 0.07734384938303016, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 7.212162017822266e-06, "grad_norm": 1.1392152309417725, "learning_rate": 2.573124205171683e-06, "loss": 0.5084, "mean_token_accuracy": 0.841187596321106, "num_tokens": 22970074.0, "step": 608 }, { "epoch": 0.07747105966162066, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 7.18235969543457e-06, "grad_norm": 1.146921992301941, "learning_rate": 2.577363289529462e-06, "loss": 0.5422, "mean_token_accuracy": 0.8323677182197571, "num_tokens": 23008185.0, "step": 609 }, { "epoch": 0.07759826994021117, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 7.271766662597656e-06, "grad_norm": 1.1554453372955322, "learning_rate": 2.5816023738872403e-06, "loss": 0.4609, "mean_token_accuracy": 0.8511192202568054, "num_tokens": 23039667.0, "step": 610 }, { "epoch": 0.07772548021880168, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 7.212162017822266e-06, "grad_norm": 1.1852033138275146, "learning_rate": 2.5858414582450192e-06, "loss": 0.5286, "mean_token_accuracy": 0.8340157270431519, "num_tokens": 23075012.0, "step": 611 }, { "epoch": 0.07785269049739219, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 1.160534381866455, "learning_rate": 2.590080542602798e-06, "loss": 0.4614, "mean_token_accuracy": 0.8512164354324341, "num_tokens": 23108397.0, "step": 612 }, { "epoch": 0.0779799007759827, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 7.241964340209961e-06, "grad_norm": 1.1579487323760986, "learning_rate": 2.594319626960577e-06, "loss": 0.5377, "mean_token_accuracy": 0.8354736566543579, "num_tokens": 23146419.0, "step": 613 }, { "epoch": 0.07810711105457321, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.1444091796875e-05, "ewc_loss_parallel": 7.271766662597656e-06, "grad_norm": 1.1144218444824219, "learning_rate": 2.598558711318355e-06, "loss": 0.5327, "mean_token_accuracy": 0.8336114883422852, "num_tokens": 23191750.0, "step": 614 }, { "epoch": 0.07823432133316371, "ewc_loss": 0.00830078125, "ewc_loss_diag": 1.150369644165039e-05, "ewc_loss_parallel": 7.152557373046875e-06, "grad_norm": 1.1539729833602905, "learning_rate": 2.602797795676134e-06, "loss": 0.5426, "mean_token_accuracy": 0.8291205167770386, "num_tokens": 23230199.0, "step": 615 }, { "epoch": 0.07836153161175423, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 1.1747432947158813, "learning_rate": 2.607036880033913e-06, "loss": 0.5311, "mean_token_accuracy": 0.8266565799713135, "num_tokens": 23264184.0, "step": 616 }, { "epoch": 0.07848874189034474, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 1.1691336631774902, "learning_rate": 2.6112759643916917e-06, "loss": 0.4867, "mean_token_accuracy": 0.846025288105011, "num_tokens": 23295543.0, "step": 617 }, { "epoch": 0.07861595216893524, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.1563301086425781e-05, "ewc_loss_parallel": 7.420778274536133e-06, "grad_norm": 1.177274465560913, "learning_rate": 2.61551504874947e-06, "loss": 0.5124, "mean_token_accuracy": 0.8399207592010498, "num_tokens": 23328148.0, "step": 618 }, { "epoch": 0.07874316244752576, "ewc_loss": 0.00860595703125, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 1.1296364068984985, "learning_rate": 2.619754133107249e-06, "loss": 0.5014, "mean_token_accuracy": 0.8406637907028198, "num_tokens": 23369441.0, "step": 619 }, { "epoch": 0.07887037272611627, "ewc_loss": 0.00836181640625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 7.212162017822266e-06, "grad_norm": 1.1913162469863892, "learning_rate": 2.623993217465028e-06, "loss": 0.5598, "mean_token_accuracy": 0.8220993280410767, "num_tokens": 23402977.0, "step": 620 }, { "epoch": 0.07899758300470679, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 1.1779556274414062, "learning_rate": 2.628232301822807e-06, "loss": 0.5362, "mean_token_accuracy": 0.8343168497085571, "num_tokens": 23441988.0, "step": 621 }, { "epoch": 0.07912479328329729, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 1.1554338932037354, "learning_rate": 2.6324713861805853e-06, "loss": 0.5651, "mean_token_accuracy": 0.8222590684890747, "num_tokens": 23480481.0, "step": 622 }, { "epoch": 0.0792520035618878, "ewc_loss": 0.0084228515625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 7.3015689849853516e-06, "grad_norm": 1.197065830230713, "learning_rate": 2.6367104705383637e-06, "loss": 0.5618, "mean_token_accuracy": 0.8210352063179016, "num_tokens": 23515188.0, "step": 623 }, { "epoch": 0.07937921384047832, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.1622905731201172e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 1.153143286705017, "learning_rate": 2.6409495548961426e-06, "loss": 0.5489, "mean_token_accuracy": 0.828877866268158, "num_tokens": 23557777.0, "step": 624 }, { "epoch": 0.07950642411906882, "ewc_loss": 0.008544921875, "ewc_loss_diag": 1.1682510375976562e-05, "ewc_loss_parallel": 7.361173629760742e-06, "grad_norm": 1.2052823305130005, "learning_rate": 2.645188639253921e-06, "loss": 0.5366, "mean_token_accuracy": 0.8352389335632324, "num_tokens": 23588357.0, "step": 625 }, { "epoch": 0.07963363439765933, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 1.1626046895980835, "learning_rate": 2.6494277236117e-06, "loss": 0.4991, "mean_token_accuracy": 0.8388766050338745, "num_tokens": 23628656.0, "step": 626 }, { "epoch": 0.07976084467624985, "ewc_loss": 0.00848388671875, "ewc_loss_diag": 1.1742115020751953e-05, "ewc_loss_parallel": 7.331371307373047e-06, "grad_norm": 1.1797059774398804, "learning_rate": 2.653666807969479e-06, "loss": 0.5976, "mean_token_accuracy": 0.8123133778572083, "num_tokens": 23666390.0, "step": 627 }, { "epoch": 0.07988805495484035, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.1801719665527344e-05, "ewc_loss_parallel": 7.62939453125e-06, "grad_norm": 1.17417573928833, "learning_rate": 2.6579058923272577e-06, "loss": 0.4568, "mean_token_accuracy": 0.8573623895645142, "num_tokens": 23699620.0, "step": 628 }, { "epoch": 0.08001526523343086, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 1.1707695722579956, "learning_rate": 2.662144976685036e-06, "loss": 0.5381, "mean_token_accuracy": 0.8315547704696655, "num_tokens": 23735883.0, "step": 629 }, { "epoch": 0.08014247551202137, "ewc_loss": 0.00885009765625, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 1.1706348657608032, "learning_rate": 2.666384061042815e-06, "loss": 0.498, "mean_token_accuracy": 0.8422462940216064, "num_tokens": 23769918.0, "step": 630 }, { "epoch": 0.08026968579061187, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 1.1404757499694824, "learning_rate": 2.670623145400594e-06, "loss": 0.5065, "mean_token_accuracy": 0.8396778106689453, "num_tokens": 23815402.0, "step": 631 }, { "epoch": 0.08039689606920239, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.1861324310302734e-05, "ewc_loss_parallel": 7.4803829193115234e-06, "grad_norm": 1.1344373226165771, "learning_rate": 2.6748622297583724e-06, "loss": 0.5032, "mean_token_accuracy": 0.8410877585411072, "num_tokens": 23858205.0, "step": 632 }, { "epoch": 0.0805241063477929, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 1.182618498802185, "learning_rate": 2.679101314116151e-06, "loss": 0.5172, "mean_token_accuracy": 0.8407157063484192, "num_tokens": 23893506.0, "step": 633 }, { "epoch": 0.08065131662638342, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.187508225440979, "learning_rate": 2.6833403984739298e-06, "loss": 0.5085, "mean_token_accuracy": 0.8358972072601318, "num_tokens": 23929457.0, "step": 634 }, { "epoch": 0.08077852690497392, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 1.1347997188568115, "learning_rate": 2.6875794828317086e-06, "loss": 0.5033, "mean_token_accuracy": 0.8440133929252625, "num_tokens": 23974577.0, "step": 635 }, { "epoch": 0.08090573718356443, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.510185241699219e-06, "grad_norm": 1.219787359237671, "learning_rate": 2.6918185671894875e-06, "loss": 0.5797, "mean_token_accuracy": 0.8201043605804443, "num_tokens": 24007490.0, "step": 636 }, { "epoch": 0.08103294746215495, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.20144784450531, "learning_rate": 2.696057651547266e-06, "loss": 0.6182, "mean_token_accuracy": 0.8062742352485657, "num_tokens": 24048058.0, "step": 637 }, { "epoch": 0.08116015774074545, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.599592208862305e-06, "grad_norm": 1.1307666301727295, "learning_rate": 2.700296735905045e-06, "loss": 0.484, "mean_token_accuracy": 0.8451059460639954, "num_tokens": 24089776.0, "step": 638 }, { "epoch": 0.08128736801933596, "ewc_loss": 0.0086669921875, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.450580596923828e-06, "grad_norm": 1.1475027799606323, "learning_rate": 2.7045358202628238e-06, "loss": 0.5344, "mean_token_accuracy": 0.8273307681083679, "num_tokens": 24135730.0, "step": 639 }, { "epoch": 0.08141457829792648, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.2031067609786987, "learning_rate": 2.7087749046206026e-06, "loss": 0.4903, "mean_token_accuracy": 0.8446118235588074, "num_tokens": 24168513.0, "step": 640 }, { "epoch": 0.08154178857651698, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.1882332563400269, "learning_rate": 2.7130139889783807e-06, "loss": 0.5536, "mean_token_accuracy": 0.8256254196166992, "num_tokens": 24206555.0, "step": 641 }, { "epoch": 0.08166899885510749, "ewc_loss": 0.0089111328125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.68899917602539e-06, "grad_norm": 1.1303163766860962, "learning_rate": 2.7172530733361596e-06, "loss": 0.461, "mean_token_accuracy": 0.8531864881515503, "num_tokens": 24248919.0, "step": 642 }, { "epoch": 0.081796209133698, "ewc_loss": 0.00872802734375, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.539987564086914e-06, "grad_norm": 1.1342512369155884, "learning_rate": 2.7214921576939385e-06, "loss": 0.4577, "mean_token_accuracy": 0.8553253412246704, "num_tokens": 24290493.0, "step": 643 }, { "epoch": 0.0819234194122885, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.1767148971557617, "learning_rate": 2.725731242051717e-06, "loss": 0.5131, "mean_token_accuracy": 0.8402631282806396, "num_tokens": 24327968.0, "step": 644 }, { "epoch": 0.08205062969087902, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.1670875549316406, "learning_rate": 2.729970326409496e-06, "loss": 0.5452, "mean_token_accuracy": 0.8298041820526123, "num_tokens": 24370019.0, "step": 645 }, { "epoch": 0.08217783996946953, "ewc_loss": 0.009033203125, "ewc_loss_diag": 1.1920928955078125e-05, "ewc_loss_parallel": 7.808208465576172e-06, "grad_norm": 1.2035064697265625, "learning_rate": 2.7342094107672747e-06, "loss": 0.5154, "mean_token_accuracy": 0.8396509885787964, "num_tokens": 24403911.0, "step": 646 }, { "epoch": 0.08230505024806005, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.1866463422775269, "learning_rate": 2.7384484951250536e-06, "loss": 0.534, "mean_token_accuracy": 0.8340234756469727, "num_tokens": 24445039.0, "step": 647 }, { "epoch": 0.08243226052665055, "ewc_loss": 0.00897216796875, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 7.748603820800781e-06, "grad_norm": 1.218745470046997, "learning_rate": 2.7426875794828316e-06, "loss": 0.5336, "mean_token_accuracy": 0.8286937475204468, "num_tokens": 24479543.0, "step": 648 }, { "epoch": 0.08255947080524106, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.2040138244628906e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.169982671737671, "learning_rate": 2.7469266638406105e-06, "loss": 0.4931, "mean_token_accuracy": 0.8415706157684326, "num_tokens": 24517483.0, "step": 649 }, { "epoch": 0.08268668108383158, "ewc_loss": 0.0087890625, "ewc_loss_diag": 1.2099742889404297e-05, "ewc_loss_parallel": 7.569789886474609e-06, "grad_norm": 1.1700633764266968, "learning_rate": 2.7511657481983894e-06, "loss": 0.5253, "mean_token_accuracy": 0.8314145803451538, "num_tokens": 24557702.0, "step": 650 }, { "epoch": 0.08281389136242208, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.2159347534179688e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.1861493587493896, "learning_rate": 2.7554048325561683e-06, "loss": 0.528, "mean_token_accuracy": 0.8349096775054932, "num_tokens": 24595145.0, "step": 651 }, { "epoch": 0.0829411016410126, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.2159347534179688e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.175512671470642, "learning_rate": 2.7596439169139467e-06, "loss": 0.5175, "mean_token_accuracy": 0.8369349241256714, "num_tokens": 24631564.0, "step": 652 }, { "epoch": 0.08306831191960311, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.1976511478424072, "learning_rate": 2.7638830012717256e-06, "loss": 0.5718, "mean_token_accuracy": 0.8212578892707825, "num_tokens": 24668563.0, "step": 653 }, { "epoch": 0.08319552219819361, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.2049272060394287, "learning_rate": 2.7681220856295045e-06, "loss": 0.5573, "mean_token_accuracy": 0.8234294652938843, "num_tokens": 24703662.0, "step": 654 }, { "epoch": 0.08332273247678412, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.1908708810806274, "learning_rate": 2.7723611699872834e-06, "loss": 0.5113, "mean_token_accuracy": 0.8416879773139954, "num_tokens": 24739504.0, "step": 655 }, { "epoch": 0.08344994275537464, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.1628366708755493, "learning_rate": 2.7766002543450614e-06, "loss": 0.5202, "mean_token_accuracy": 0.8322871923446655, "num_tokens": 24781884.0, "step": 656 }, { "epoch": 0.08357715303396514, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.2218952178955078e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.1697498559951782, "learning_rate": 2.7808393387028403e-06, "loss": 0.5742, "mean_token_accuracy": 0.8189064264297485, "num_tokens": 24823964.0, "step": 657 }, { "epoch": 0.08370436331255565, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.201711654663086, "learning_rate": 2.785078423060619e-06, "loss": 0.5171, "mean_token_accuracy": 0.8362910747528076, "num_tokens": 24859552.0, "step": 658 }, { "epoch": 0.08383157359114617, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 1.2093408107757568, "learning_rate": 2.789317507418398e-06, "loss": 0.5361, "mean_token_accuracy": 0.8299517631530762, "num_tokens": 24897580.0, "step": 659 }, { "epoch": 0.08395878386973668, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 1.1774731874465942, "learning_rate": 2.7935565917761765e-06, "loss": 0.4842, "mean_token_accuracy": 0.8435621857643127, "num_tokens": 24934616.0, "step": 660 }, { "epoch": 0.08408599414832718, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.20969557762146, "learning_rate": 2.7977956761339554e-06, "loss": 0.556, "mean_token_accuracy": 0.8245049118995667, "num_tokens": 24972149.0, "step": 661 }, { "epoch": 0.0842132044269177, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 1.1860958337783813, "learning_rate": 2.8020347604917343e-06, "loss": 0.4381, "mean_token_accuracy": 0.8616396188735962, "num_tokens": 25008682.0, "step": 662 }, { "epoch": 0.08434041470550821, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.156179428100586, "learning_rate": 2.8062738448495123e-06, "loss": 0.4637, "mean_token_accuracy": 0.8504425287246704, "num_tokens": 25048454.0, "step": 663 }, { "epoch": 0.08446762498409871, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.233816146850586e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.2147777080535889, "learning_rate": 2.8105129292072912e-06, "loss": 0.5404, "mean_token_accuracy": 0.8237242698669434, "num_tokens": 25086772.0, "step": 664 }, { "epoch": 0.08459483526268922, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 1.196308970451355, "learning_rate": 2.81475201356507e-06, "loss": 0.4896, "mean_token_accuracy": 0.8418370485305786, "num_tokens": 25122215.0, "step": 665 }, { "epoch": 0.08472204554127974, "ewc_loss": 0.00909423828125, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.1908154487609863, "learning_rate": 2.818991097922849e-06, "loss": 0.5322, "mean_token_accuracy": 0.8352447748184204, "num_tokens": 25159966.0, "step": 666 }, { "epoch": 0.08484925581987024, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.239776611328125e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 1.2229547500610352, "learning_rate": 2.8232301822806274e-06, "loss": 0.4865, "mean_token_accuracy": 0.8474874496459961, "num_tokens": 25191929.0, "step": 667 }, { "epoch": 0.08497646609846075, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 1.1894649267196655, "learning_rate": 2.8274692666384063e-06, "loss": 0.5076, "mean_token_accuracy": 0.8361181616783142, "num_tokens": 25233385.0, "step": 668 }, { "epoch": 0.08510367637705127, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 7.867813110351562e-06, "grad_norm": 1.1689069271087646, "learning_rate": 2.8317083509961852e-06, "loss": 0.5272, "mean_token_accuracy": 0.8345345258712769, "num_tokens": 25276943.0, "step": 669 }, { "epoch": 0.08523088665564178, "ewc_loss": 0.0091552734375, "ewc_loss_diag": 1.2516975402832031e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 1.2053673267364502, "learning_rate": 2.835947435353964e-06, "loss": 0.5052, "mean_token_accuracy": 0.840142548084259, "num_tokens": 25316552.0, "step": 670 }, { "epoch": 0.08535809693423228, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.2576580047607422e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 1.223228931427002, "learning_rate": 2.840186519711742e-06, "loss": 0.5592, "mean_token_accuracy": 0.8295422792434692, "num_tokens": 25357517.0, "step": 671 }, { "epoch": 0.0854853072128228, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 1.226057529449463, "learning_rate": 2.844425604069521e-06, "loss": 0.55, "mean_token_accuracy": 0.8273457288742065, "num_tokens": 25392693.0, "step": 672 }, { "epoch": 0.08561251749141331, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 1.2233909368515015, "learning_rate": 2.8486646884273e-06, "loss": 0.5133, "mean_token_accuracy": 0.8371680974960327, "num_tokens": 25430623.0, "step": 673 }, { "epoch": 0.08573972777000381, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 1.1951559782028198, "learning_rate": 2.852903772785079e-06, "loss": 0.4525, "mean_token_accuracy": 0.8557946681976318, "num_tokens": 25465043.0, "step": 674 }, { "epoch": 0.08586693804859433, "ewc_loss": 0.00921630859375, "ewc_loss_diag": 1.2636184692382812e-05, "ewc_loss_parallel": 7.927417755126953e-06, "grad_norm": 1.1751927137374878, "learning_rate": 2.8571428571428573e-06, "loss": 0.4756, "mean_token_accuracy": 0.8499056100845337, "num_tokens": 25503483.0, "step": 675 }, { "epoch": 0.08599414832718484, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 1.1785917282104492, "learning_rate": 2.861381941500636e-06, "loss": 0.528, "mean_token_accuracy": 0.8331913948059082, "num_tokens": 25547079.0, "step": 676 }, { "epoch": 0.08612135860577534, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 1.2479437589645386, "learning_rate": 2.865621025858415e-06, "loss": 0.5639, "mean_token_accuracy": 0.8211978673934937, "num_tokens": 25586494.0, "step": 677 }, { "epoch": 0.08624856888436586, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.2695789337158203e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 1.2434278726577759, "learning_rate": 2.869860110216193e-06, "loss": 0.5636, "mean_token_accuracy": 0.8208489418029785, "num_tokens": 25618998.0, "step": 678 }, { "epoch": 0.08637577916295637, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 1.1667646169662476, "learning_rate": 2.874099194573972e-06, "loss": 0.5047, "mean_token_accuracy": 0.8401842713356018, "num_tokens": 25663147.0, "step": 679 }, { "epoch": 0.08650298944154687, "ewc_loss": 0.00927734375, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 7.987022399902344e-06, "grad_norm": 1.2281146049499512, "learning_rate": 2.878338278931751e-06, "loss": 0.553, "mean_token_accuracy": 0.8310701251029968, "num_tokens": 25696830.0, "step": 680 }, { "epoch": 0.08663019972013739, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.210174322128296, "learning_rate": 2.8825773632895297e-06, "loss": 0.4819, "mean_token_accuracy": 0.8507944345474243, "num_tokens": 25732828.0, "step": 681 }, { "epoch": 0.0867574099987279, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 1.1750787496566772, "learning_rate": 2.886816447647308e-06, "loss": 0.4715, "mean_token_accuracy": 0.8483611345291138, "num_tokens": 25773545.0, "step": 682 }, { "epoch": 0.08688462027731841, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.2755393981933594e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 1.1888285875320435, "learning_rate": 2.891055532005087e-06, "loss": 0.5102, "mean_token_accuracy": 0.8388862609863281, "num_tokens": 25814792.0, "step": 683 }, { "epoch": 0.08701183055590891, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.2814998626708984e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.2133592367172241, "learning_rate": 2.895294616362866e-06, "loss": 0.525, "mean_token_accuracy": 0.8320326805114746, "num_tokens": 25850713.0, "step": 684 }, { "epoch": 0.08713904083449943, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.2814998626708984e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 1.2060015201568604, "learning_rate": 2.899533700720645e-06, "loss": 0.4952, "mean_token_accuracy": 0.8388988971710205, "num_tokens": 25888320.0, "step": 685 }, { "epoch": 0.08726625111308994, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.2814998626708984e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 1.2076114416122437, "learning_rate": 2.9037727850784233e-06, "loss": 0.538, "mean_token_accuracy": 0.8294309973716736, "num_tokens": 25924941.0, "step": 686 }, { "epoch": 0.08739346139168044, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.2874603271484375e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 1.211406946182251, "learning_rate": 2.9080118694362018e-06, "loss": 0.5037, "mean_token_accuracy": 0.8406850099563599, "num_tokens": 25964051.0, "step": 687 }, { "epoch": 0.08752067167027096, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 1.1835768222808838, "learning_rate": 2.9122509537939806e-06, "loss": 0.5067, "mean_token_accuracy": 0.8400417566299438, "num_tokens": 26007300.0, "step": 688 }, { "epoch": 0.08764788194886147, "ewc_loss": 0.00933837890625, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 1.2117279767990112, "learning_rate": 2.9164900381517595e-06, "loss": 0.5634, "mean_token_accuracy": 0.8185178637504578, "num_tokens": 26045833.0, "step": 689 }, { "epoch": 0.08777509222745197, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.2934207916259766e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 1.190977692604065, "learning_rate": 2.920729122509538e-06, "loss": 0.556, "mean_token_accuracy": 0.8244274854660034, "num_tokens": 26092277.0, "step": 690 }, { "epoch": 0.08790230250604249, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 1.1993207931518555, "learning_rate": 2.924968206867317e-06, "loss": 0.5091, "mean_token_accuracy": 0.8367041945457458, "num_tokens": 26130082.0, "step": 691 }, { "epoch": 0.088029512784633, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.1999226808547974, "learning_rate": 2.9292072912250958e-06, "loss": 0.4848, "mean_token_accuracy": 0.847389280796051, "num_tokens": 26170087.0, "step": 692 }, { "epoch": 0.0881567230632235, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 1.2287768125534058, "learning_rate": 2.9334463755828746e-06, "loss": 0.5645, "mean_token_accuracy": 0.8214367628097534, "num_tokens": 26207645.0, "step": 693 }, { "epoch": 0.08828393334181402, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.212685227394104, "learning_rate": 2.937685459940653e-06, "loss": 0.4875, "mean_token_accuracy": 0.8420683145523071, "num_tokens": 26242417.0, "step": 694 }, { "epoch": 0.08841114362040453, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 1.252856969833374, "learning_rate": 2.941924544298432e-06, "loss": 0.5547, "mean_token_accuracy": 0.8260941505432129, "num_tokens": 26275427.0, "step": 695 }, { "epoch": 0.08853835389899505, "ewc_loss": 0.00946044921875, "ewc_loss_diag": 1.3053417205810547e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.23778235912323, "learning_rate": 2.9461636286562104e-06, "loss": 0.5075, "mean_token_accuracy": 0.8391612768173218, "num_tokens": 26308692.0, "step": 696 }, { "epoch": 0.08866556417758555, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 1.1758731603622437, "learning_rate": 2.950402713013989e-06, "loss": 0.5012, "mean_token_accuracy": 0.8418577909469604, "num_tokens": 26352155.0, "step": 697 }, { "epoch": 0.08879277445617606, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 8.106231689453125e-06, "grad_norm": 1.2094753980636597, "learning_rate": 2.954641797371768e-06, "loss": 0.5432, "mean_token_accuracy": 0.8294028043746948, "num_tokens": 26392410.0, "step": 698 }, { "epoch": 0.08891998473476657, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.2047511339187622, "learning_rate": 2.9588808817295467e-06, "loss": 0.4704, "mean_token_accuracy": 0.8494595885276794, "num_tokens": 26430120.0, "step": 699 }, { "epoch": 0.08904719501335707, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.2271630764007568, "learning_rate": 2.9631199660873256e-06, "loss": 0.5331, "mean_token_accuracy": 0.828900933265686, "num_tokens": 26466159.0, "step": 700 }, { "epoch": 0.08917440529194759, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.3113021850585938e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.1991945505142212, "learning_rate": 2.967359050445104e-06, "loss": 0.4537, "mean_token_accuracy": 0.8565458059310913, "num_tokens": 26502382.0, "step": 701 }, { "epoch": 0.0893016155705381, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.203832983970642, "learning_rate": 2.971598134802883e-06, "loss": 0.492, "mean_token_accuracy": 0.8445628881454468, "num_tokens": 26540911.0, "step": 702 }, { "epoch": 0.0894288258491286, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.2298178672790527, "learning_rate": 2.975837219160662e-06, "loss": 0.5535, "mean_token_accuracy": 0.8236944079399109, "num_tokens": 26580081.0, "step": 703 }, { "epoch": 0.08955603612771912, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.3172626495361328e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.2208657264709473, "learning_rate": 2.9800763035184403e-06, "loss": 0.5191, "mean_token_accuracy": 0.8353092670440674, "num_tokens": 26622033.0, "step": 704 }, { "epoch": 0.08968324640630963, "ewc_loss": 0.009521484375, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 8.165836334228516e-06, "grad_norm": 1.2313392162322998, "learning_rate": 2.9843153878762187e-06, "loss": 0.5193, "mean_token_accuracy": 0.835616409778595, "num_tokens": 26657042.0, "step": 705 }, { "epoch": 0.08981045668490013, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.2366173267364502, "learning_rate": 2.9885544722339976e-06, "loss": 0.5313, "mean_token_accuracy": 0.832592785358429, "num_tokens": 26696668.0, "step": 706 }, { "epoch": 0.08993766696349065, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.2302972078323364, "learning_rate": 2.9927935565917765e-06, "loss": 0.5336, "mean_token_accuracy": 0.8336303234100342, "num_tokens": 26736550.0, "step": 707 }, { "epoch": 0.09006487724208116, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.2262922525405884, "learning_rate": 2.9970326409495554e-06, "loss": 0.5402, "mean_token_accuracy": 0.8308316469192505, "num_tokens": 26776311.0, "step": 708 }, { "epoch": 0.09019208752067168, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.329183578491211e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.2186764478683472, "learning_rate": 3.001271725307334e-06, "loss": 0.4801, "mean_token_accuracy": 0.8471940755844116, "num_tokens": 26812094.0, "step": 709 }, { "epoch": 0.09031929779926218, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.2374284267425537, "learning_rate": 3.0055108096651127e-06, "loss": 0.515, "mean_token_accuracy": 0.8381826877593994, "num_tokens": 26846317.0, "step": 710 }, { "epoch": 0.09044650807785269, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.33514404296875e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.1976672410964966, "learning_rate": 3.0097498940228916e-06, "loss": 0.4559, "mean_token_accuracy": 0.8541309833526611, "num_tokens": 26885247.0, "step": 711 }, { "epoch": 0.0905737183564432, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.228241205215454, "learning_rate": 3.0139889783806705e-06, "loss": 0.5198, "mean_token_accuracy": 0.836531400680542, "num_tokens": 26925578.0, "step": 712 }, { "epoch": 0.0907009286350337, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.3470649719238281e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.2458374500274658, "learning_rate": 3.0182280627384485e-06, "loss": 0.5437, "mean_token_accuracy": 0.8313985466957092, "num_tokens": 26961924.0, "step": 713 }, { "epoch": 0.09082813891362422, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.3530254364013672e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.209775686264038, "learning_rate": 3.0224671470962274e-06, "loss": 0.4738, "mean_token_accuracy": 0.8472452759742737, "num_tokens": 26999433.0, "step": 714 }, { "epoch": 0.09095534919221474, "ewc_loss": 0.00958251953125, "ewc_loss_diag": 1.3530254364013672e-05, "ewc_loss_parallel": 8.225440979003906e-06, "grad_norm": 1.970465064048767, "learning_rate": 3.0267062314540063e-06, "loss": 0.4962, "mean_token_accuracy": 0.8417116403579712, "num_tokens": 27044427.0, "step": 715 }, { "epoch": 0.09108255947080524, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 1.3530254364013672e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 1.3709630966186523, "learning_rate": 3.0309453158117847e-06, "loss": 0.5062, "mean_token_accuracy": 0.839432954788208, "num_tokens": 27080645.0, "step": 716 }, { "epoch": 0.09120976974939575, "ewc_loss": 0.0093994140625, "ewc_loss_diag": 1.3530254364013672e-05, "ewc_loss_parallel": 8.046627044677734e-06, "grad_norm": 1.1800845861434937, "learning_rate": 3.0351844001695636e-06, "loss": 0.5701, "mean_token_accuracy": 0.821628212928772, "num_tokens": 27118213.0, "step": 717 }, { "epoch": 0.09133698002798626, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.3530254364013672e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 1.2678918838500977, "learning_rate": 3.0394234845273425e-06, "loss": 0.4882, "mean_token_accuracy": 0.8448212146759033, "num_tokens": 27156827.0, "step": 718 }, { "epoch": 0.09146419030657676, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.3530254364013672e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2694255113601685, "learning_rate": 3.0436625688851214e-06, "loss": 0.5219, "mean_token_accuracy": 0.8307325839996338, "num_tokens": 27192016.0, "step": 719 }, { "epoch": 0.09159140058516728, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.3530254364013672e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2781342267990112, "learning_rate": 3.0479016532428994e-06, "loss": 0.5667, "mean_token_accuracy": 0.8183512687683105, "num_tokens": 27226037.0, "step": 720 }, { "epoch": 0.0917186108637578, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2529947757720947, "learning_rate": 3.0521407376006783e-06, "loss": 0.4811, "mean_token_accuracy": 0.8460474014282227, "num_tokens": 27260399.0, "step": 721 }, { "epoch": 0.09184582114234831, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2380348443984985, "learning_rate": 3.056379821958457e-06, "loss": 0.4915, "mean_token_accuracy": 0.8450678586959839, "num_tokens": 27297293.0, "step": 722 }, { "epoch": 0.09197303142093881, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.2494080066680908, "learning_rate": 3.060618906316236e-06, "loss": 0.4873, "mean_token_accuracy": 0.8453719615936279, "num_tokens": 27333045.0, "step": 723 }, { "epoch": 0.09210024169952932, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2619532346725464, "learning_rate": 3.0648579906740146e-06, "loss": 0.4846, "mean_token_accuracy": 0.8438538312911987, "num_tokens": 27369044.0, "step": 724 }, { "epoch": 0.09222745197811984, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.3589859008789062e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.27133309841156, "learning_rate": 3.0690970750317934e-06, "loss": 0.5394, "mean_token_accuracy": 0.8289751410484314, "num_tokens": 27407486.0, "step": 725 }, { "epoch": 0.09235466225671034, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2693017721176147, "learning_rate": 3.0733361593895723e-06, "loss": 0.5038, "mean_token_accuracy": 0.8410525321960449, "num_tokens": 27441835.0, "step": 726 }, { "epoch": 0.09248187253530085, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.220874309539795, "learning_rate": 3.077575243747351e-06, "loss": 0.5171, "mean_token_accuracy": 0.8345664739608765, "num_tokens": 27483105.0, "step": 727 }, { "epoch": 0.09260908281389137, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 1.2297576665878296, "learning_rate": 3.0818143281051292e-06, "loss": 0.5356, "mean_token_accuracy": 0.8329700231552124, "num_tokens": 27528888.0, "step": 728 }, { "epoch": 0.09273629309248187, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.3649463653564453e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.253562092781067, "learning_rate": 3.086053412462908e-06, "loss": 0.5744, "mean_token_accuracy": 0.8192302584648132, "num_tokens": 27571763.0, "step": 729 }, { "epoch": 0.09286350337107238, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.3227721452713013, "learning_rate": 3.090292496820687e-06, "loss": 0.505, "mean_token_accuracy": 0.8423995971679688, "num_tokens": 27606980.0, "step": 730 }, { "epoch": 0.0929907136496629, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2989922761917114, "learning_rate": 3.094531581178466e-06, "loss": 0.4973, "mean_token_accuracy": 0.8414337635040283, "num_tokens": 27640969.0, "step": 731 }, { "epoch": 0.0931179239282534, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.2442837953567505, "learning_rate": 3.0987706655362444e-06, "loss": 0.6339, "mean_token_accuracy": 0.7998613715171814, "num_tokens": 27684360.0, "step": 732 }, { "epoch": 0.09324513420684391, "ewc_loss": 0.0096435546875, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 1.2352205514907837, "learning_rate": 3.1030097498940232e-06, "loss": 0.4749, "mean_token_accuracy": 0.8463436365127563, "num_tokens": 27723596.0, "step": 733 }, { "epoch": 0.09337234448543442, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 1.2665295600891113, "learning_rate": 3.107248834251802e-06, "loss": 0.5105, "mean_token_accuracy": 0.831173300743103, "num_tokens": 27759469.0, "step": 734 }, { "epoch": 0.09349955476402494, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.3768672943115234e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.2513344287872314, "learning_rate": 3.11148791860958e-06, "loss": 0.5028, "mean_token_accuracy": 0.8406086564064026, "num_tokens": 27795783.0, "step": 735 }, { "epoch": 0.09362676504261544, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.3828277587890625e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 1.2737442255020142, "learning_rate": 3.115727002967359e-06, "loss": 0.5125, "mean_token_accuracy": 0.8342862725257874, "num_tokens": 27826146.0, "step": 736 }, { "epoch": 0.09375397532120595, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.242524266242981, "learning_rate": 3.119966087325138e-06, "loss": 0.5157, "mean_token_accuracy": 0.8387608528137207, "num_tokens": 27865581.0, "step": 737 }, { "epoch": 0.09388118559979647, "ewc_loss": 0.00970458984375, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 8.285045623779297e-06, "grad_norm": 1.2274450063705444, "learning_rate": 3.124205171682917e-06, "loss": 0.5592, "mean_token_accuracy": 0.8255645036697388, "num_tokens": 27911209.0, "step": 738 }, { "epoch": 0.09400839587838697, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 1.2502907514572144, "learning_rate": 3.1284442560406953e-06, "loss": 0.5216, "mean_token_accuracy": 0.8335897922515869, "num_tokens": 27951639.0, "step": 739 }, { "epoch": 0.09413560615697748, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.3887882232666016e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.2282623052597046, "learning_rate": 3.132683340398474e-06, "loss": 0.4554, "mean_token_accuracy": 0.8517510294914246, "num_tokens": 27990808.0, "step": 740 }, { "epoch": 0.094262816435568, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.261570930480957, "learning_rate": 3.136922424756253e-06, "loss": 0.4787, "mean_token_accuracy": 0.8494352102279663, "num_tokens": 28025421.0, "step": 741 }, { "epoch": 0.0943900267141585, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2134149074554443, "learning_rate": 3.141161509114032e-06, "loss": 0.4499, "mean_token_accuracy": 0.8585299253463745, "num_tokens": 28068652.0, "step": 742 }, { "epoch": 0.09451723699274901, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 1.2454930543899536, "learning_rate": 3.14540059347181e-06, "loss": 0.4936, "mean_token_accuracy": 0.8400302529335022, "num_tokens": 28104852.0, "step": 743 }, { "epoch": 0.09464444727133953, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2560371160507202, "learning_rate": 3.149639677829589e-06, "loss": 0.4651, "mean_token_accuracy": 0.8504184484481812, "num_tokens": 28141646.0, "step": 744 }, { "epoch": 0.09477165754993004, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.4007091522216797e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.268226981163025, "learning_rate": 3.1538787621873677e-06, "loss": 0.5641, "mean_token_accuracy": 0.8271620273590088, "num_tokens": 28178951.0, "step": 745 }, { "epoch": 0.09489886782852054, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2615779638290405, "learning_rate": 3.1581178465451466e-06, "loss": 0.5788, "mean_token_accuracy": 0.8215314149856567, "num_tokens": 28221911.0, "step": 746 }, { "epoch": 0.09502607810711106, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.2949267625808716, "learning_rate": 3.162356930902925e-06, "loss": 0.4984, "mean_token_accuracy": 0.8425282835960388, "num_tokens": 28252828.0, "step": 747 }, { "epoch": 0.09515328838570157, "ewc_loss": 0.00994873046875, "ewc_loss_diag": 1.4066696166992188e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.242148995399475, "learning_rate": 3.166596015260704e-06, "loss": 0.4736, "mean_token_accuracy": 0.8472632765769958, "num_tokens": 28293459.0, "step": 748 }, { "epoch": 0.09528049866429207, "ewc_loss": 0.009765625, "ewc_loss_diag": 1.4185905456542969e-05, "ewc_loss_parallel": 8.344650268554688e-06, "grad_norm": 1.2432225942611694, "learning_rate": 3.170835099618483e-06, "loss": 0.5084, "mean_token_accuracy": 0.8381106853485107, "num_tokens": 28332775.0, "step": 749 }, { "epoch": 0.09540770894288259, "ewc_loss": 0.00982666015625, "ewc_loss_diag": 1.4185905456542969e-05, "ewc_loss_parallel": 8.404254913330078e-06, "grad_norm": 1.2459280490875244, "learning_rate": 3.1750741839762617e-06, "loss": 0.4856, "mean_token_accuracy": 0.8450037240982056, "num_tokens": 28370367.0, "step": 750 }, { "epoch": 0.0955349192214731, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.4185905456542969e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2751126289367676, "learning_rate": 3.1793132683340398e-06, "loss": 0.5181, "mean_token_accuracy": 0.837656557559967, "num_tokens": 28404913.0, "step": 751 }, { "epoch": 0.0956621295000636, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.4185905456542969e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2616523504257202, "learning_rate": 3.1835523526918187e-06, "loss": 0.4595, "mean_token_accuracy": 0.8521761894226074, "num_tokens": 28441057.0, "step": 752 }, { "epoch": 0.09578933977865411, "ewc_loss": 0.0098876953125, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 8.463859558105469e-06, "grad_norm": 1.2696340084075928, "learning_rate": 3.1877914370495975e-06, "loss": 0.5499, "mean_token_accuracy": 0.8245187997817993, "num_tokens": 28479691.0, "step": 753 }, { "epoch": 0.09591655005724463, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.430511474609375e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.2640011310577393, "learning_rate": 3.192030521407376e-06, "loss": 0.5439, "mean_token_accuracy": 0.829920768737793, "num_tokens": 28520469.0, "step": 754 }, { "epoch": 0.09604376033583513, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.436471939086914e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.2489378452301025, "learning_rate": 3.196269605765155e-06, "loss": 0.4847, "mean_token_accuracy": 0.8441454172134399, "num_tokens": 28560622.0, "step": 755 }, { "epoch": 0.09617097061442564, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.2814913988113403, "learning_rate": 3.2005086901229338e-06, "loss": 0.5236, "mean_token_accuracy": 0.8315761089324951, "num_tokens": 28594890.0, "step": 756 }, { "epoch": 0.09629818089301616, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.2607698440551758, "learning_rate": 3.2047477744807127e-06, "loss": 0.517, "mean_token_accuracy": 0.8356438279151917, "num_tokens": 28633112.0, "step": 757 }, { "epoch": 0.09642539117160667, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.2564539909362793, "learning_rate": 3.208986858838491e-06, "loss": 0.5441, "mean_token_accuracy": 0.8251797556877136, "num_tokens": 28672195.0, "step": 758 }, { "epoch": 0.09655260145019717, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.4483928680419922e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.2588720321655273, "learning_rate": 3.2132259431962696e-06, "loss": 0.4841, "mean_token_accuracy": 0.8482029438018799, "num_tokens": 28710315.0, "step": 759 }, { "epoch": 0.09667981172878769, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.4543533325195312e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.263900637626648, "learning_rate": 3.2174650275540485e-06, "loss": 0.5231, "mean_token_accuracy": 0.8387643098831177, "num_tokens": 28748713.0, "step": 760 }, { "epoch": 0.0968070220073782, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 1.2412927150726318, "learning_rate": 3.2217041119118274e-06, "loss": 0.4888, "mean_token_accuracy": 0.8435665369033813, "num_tokens": 28789832.0, "step": 761 }, { "epoch": 0.0969342322859687, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.288360357284546, "learning_rate": 3.225943196269606e-06, "loss": 0.5133, "mean_token_accuracy": 0.8383949995040894, "num_tokens": 28822781.0, "step": 762 }, { "epoch": 0.09706144256455922, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.2748335599899292, "learning_rate": 3.2301822806273847e-06, "loss": 0.4984, "mean_token_accuracy": 0.8370425701141357, "num_tokens": 28857734.0, "step": 763 }, { "epoch": 0.09718865284314973, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.2455928325653076, "learning_rate": 3.2344213649851636e-06, "loss": 0.5303, "mean_token_accuracy": 0.831741213798523, "num_tokens": 28901296.0, "step": 764 }, { "epoch": 0.09731586312174023, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.2591105699539185, "learning_rate": 3.2386604493429425e-06, "loss": 0.4969, "mean_token_accuracy": 0.8394895195960999, "num_tokens": 28942053.0, "step": 765 }, { "epoch": 0.09744307340033075, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.4603137969970703e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 1.2933741807937622, "learning_rate": 3.242899533700721e-06, "loss": 0.468, "mean_token_accuracy": 0.8459183573722839, "num_tokens": 28975320.0, "step": 766 }, { "epoch": 0.09757028367892126, "ewc_loss": 0.01019287109375, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.2479591369628906, "learning_rate": 3.2471386180585e-06, "loss": 0.5153, "mean_token_accuracy": 0.8301458358764648, "num_tokens": 29019941.0, "step": 767 }, { "epoch": 0.09769749395751176, "ewc_loss": 0.010009765625, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 8.52346420288086e-06, "grad_norm": 1.2483316659927368, "learning_rate": 3.2513777024162783e-06, "loss": 0.5449, "mean_token_accuracy": 0.8299270868301392, "num_tokens": 29062608.0, "step": 768 }, { "epoch": 0.09782470423610228, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.2599694728851318, "learning_rate": 3.255616786774057e-06, "loss": 0.4903, "mean_token_accuracy": 0.8450689315795898, "num_tokens": 29106423.0, "step": 769 }, { "epoch": 0.09795191451469279, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.2416666746139526, "learning_rate": 3.2598558711318356e-06, "loss": 0.4373, "mean_token_accuracy": 0.857963502407074, "num_tokens": 29148717.0, "step": 770 }, { "epoch": 0.0980791247932833, "ewc_loss": 0.01019287109375, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.254047155380249, "learning_rate": 3.2640949554896145e-06, "loss": 0.4841, "mean_token_accuracy": 0.8456830978393555, "num_tokens": 29189464.0, "step": 771 }, { "epoch": 0.0982063350718738, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.2704368829727173, "learning_rate": 3.2683340398473934e-06, "loss": 0.5139, "mean_token_accuracy": 0.8363516330718994, "num_tokens": 29228384.0, "step": 772 }, { "epoch": 0.09833354535046432, "ewc_loss": 0.01019287109375, "ewc_loss_diag": 1.4662742614746094e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.259834885597229, "learning_rate": 3.272573124205172e-06, "loss": 0.5027, "mean_token_accuracy": 0.8435094356536865, "num_tokens": 29270942.0, "step": 773 }, { "epoch": 0.09846075562905483, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.2444270849227905, "learning_rate": 3.2768122085629507e-06, "loss": 0.4791, "mean_token_accuracy": 0.8476018905639648, "num_tokens": 29313770.0, "step": 774 }, { "epoch": 0.09858796590764533, "ewc_loss": 0.01019287109375, "ewc_loss_diag": 1.4781951904296875e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.2882808446884155, "learning_rate": 3.2810512929207296e-06, "loss": 0.4768, "mean_token_accuracy": 0.8452421426773071, "num_tokens": 29347924.0, "step": 775 }, { "epoch": 0.09871517618623585, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.267836332321167, "learning_rate": 3.2852903772785085e-06, "loss": 0.5263, "mean_token_accuracy": 0.8314387798309326, "num_tokens": 29386899.0, "step": 776 }, { "epoch": 0.09884238646482636, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.3125088214874268, "learning_rate": 3.2895294616362865e-06, "loss": 0.4714, "mean_token_accuracy": 0.8523374795913696, "num_tokens": 29423349.0, "step": 777 }, { "epoch": 0.09896959674341686, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.26384699344635, "learning_rate": 3.2937685459940654e-06, "loss": 0.4799, "mean_token_accuracy": 0.8437989950180054, "num_tokens": 29461127.0, "step": 778 }, { "epoch": 0.09909680702200738, "ewc_loss": 0.01007080078125, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 8.58306884765625e-06, "grad_norm": 1.284680724143982, "learning_rate": 3.2980076303518443e-06, "loss": 0.5628, "mean_token_accuracy": 0.8258395195007324, "num_tokens": 29498335.0, "step": 779 }, { "epoch": 0.09922401730059789, "ewc_loss": 0.01019287109375, "ewc_loss_diag": 1.4901161193847656e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.263351321220398, "learning_rate": 3.302246714709623e-06, "loss": 0.4907, "mean_token_accuracy": 0.8423653244972229, "num_tokens": 29541174.0, "step": 780 }, { "epoch": 0.09935122757918839, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.2746292352676392, "learning_rate": 3.3064857990674017e-06, "loss": 0.5158, "mean_token_accuracy": 0.8370218873023987, "num_tokens": 29578684.0, "step": 781 }, { "epoch": 0.0994784378577789, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 1.296048879623413, "learning_rate": 3.3107248834251805e-06, "loss": 0.464, "mean_token_accuracy": 0.8514631390571594, "num_tokens": 29611236.0, "step": 782 }, { "epoch": 0.09960564813636942, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.2605981826782227, "learning_rate": 3.3149639677829594e-06, "loss": 0.4813, "mean_token_accuracy": 0.8471100330352783, "num_tokens": 29653769.0, "step": 783 }, { "epoch": 0.09973285841495994, "ewc_loss": 0.0101318359375, "ewc_loss_diag": 1.4960765838623047e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.2628264427185059, "learning_rate": 3.3192030521407383e-06, "loss": 0.494, "mean_token_accuracy": 0.8448405265808105, "num_tokens": 29695231.0, "step": 784 }, { "epoch": 0.09986006869355044, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.322786808013916, "learning_rate": 3.3234421364985163e-06, "loss": 0.4758, "mean_token_accuracy": 0.8493750095367432, "num_tokens": 29725755.0, "step": 785 }, { "epoch": 0.09998727897214095, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 1.5079975128173828e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 1.3114124536514282, "learning_rate": 3.3276812208562952e-06, "loss": 0.4951, "mean_token_accuracy": 0.8405319452285767, "num_tokens": 29760270.0, "step": 786 }, { "epoch": 0.10011448925073146, "ewc_loss": 0.01019287109375, "ewc_loss_diag": 1.5139579772949219e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.2950156927108765, "learning_rate": 3.331920305214074e-06, "loss": 0.5454, "mean_token_accuracy": 0.8331220149993896, "num_tokens": 29797292.0, "step": 787 }, { "epoch": 0.10024169952932196, "ewc_loss": 0.01019287109375, "ewc_loss_diag": 1.5139579772949219e-05, "ewc_loss_parallel": 8.64267349243164e-06, "grad_norm": 1.278576374053955, "learning_rate": 3.3361593895718526e-06, "loss": 0.469, "mean_token_accuracy": 0.8534190058708191, "num_tokens": 29833265.0, "step": 788 }, { "epoch": 0.10036890980791248, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.5139579772949219e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.2693620920181274, "learning_rate": 3.3403984739296315e-06, "loss": 0.5473, "mean_token_accuracy": 0.8264524936676025, "num_tokens": 29876399.0, "step": 789 }, { "epoch": 0.100496120086503, "ewc_loss": 0.01031494140625, "ewc_loss_diag": 1.5139579772949219e-05, "ewc_loss_parallel": 8.761882781982422e-06, "grad_norm": 1.313287377357483, "learning_rate": 3.3446375582874103e-06, "loss": 0.5088, "mean_token_accuracy": 0.839840829372406, "num_tokens": 29913883.0, "step": 790 }, { "epoch": 0.1006233303650935, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 1.3293129205703735, "learning_rate": 3.3488766426451892e-06, "loss": 0.5631, "mean_token_accuracy": 0.8281645774841309, "num_tokens": 29951691.0, "step": 791 }, { "epoch": 0.10075054064368401, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 1.2701119184494019, "learning_rate": 3.3531157270029673e-06, "loss": 0.5266, "mean_token_accuracy": 0.833797037601471, "num_tokens": 29992978.0, "step": 792 }, { "epoch": 0.10087775092227452, "ewc_loss": 0.01025390625, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 8.702278137207031e-06, "grad_norm": 1.2644258737564087, "learning_rate": 3.357354811360746e-06, "loss": 0.5155, "mean_token_accuracy": 0.8332322239875793, "num_tokens": 30035602.0, "step": 793 }, { "epoch": 0.10100496120086502, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 1.3116999864578247, "learning_rate": 3.361593895718525e-06, "loss": 0.5139, "mean_token_accuracy": 0.8330063819885254, "num_tokens": 30070063.0, "step": 794 }, { "epoch": 0.10113217147945554, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 1.3299624919891357, "learning_rate": 3.365832980076304e-06, "loss": 0.5529, "mean_token_accuracy": 0.822414219379425, "num_tokens": 30101636.0, "step": 795 }, { "epoch": 0.10125938175804605, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 1.293575644493103, "learning_rate": 3.3700720644340824e-06, "loss": 0.5118, "mean_token_accuracy": 0.8296682834625244, "num_tokens": 30137429.0, "step": 796 }, { "epoch": 0.10138659203663657, "ewc_loss": 0.0103759765625, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 8.821487426757812e-06, "grad_norm": 1.2814635038375854, "learning_rate": 3.3743111487918613e-06, "loss": 0.4892, "mean_token_accuracy": 0.8450050950050354, "num_tokens": 30176173.0, "step": 797 }, { "epoch": 0.10151380231522707, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 1.3196146488189697, "learning_rate": 3.37855023314964e-06, "loss": 0.4845, "mean_token_accuracy": 0.845284640789032, "num_tokens": 30206588.0, "step": 798 }, { "epoch": 0.10164101259381758, "ewc_loss": 0.01055908203125, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 1.3342173099517822, "learning_rate": 3.382789317507419e-06, "loss": 0.5143, "mean_token_accuracy": 0.8364171981811523, "num_tokens": 30237345.0, "step": 799 }, { "epoch": 0.1017682228724081, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 1.52587890625e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 1.3077741861343384, "learning_rate": 3.387028401865197e-06, "loss": 0.5265, "mean_token_accuracy": 0.8370194435119629, "num_tokens": 30274543.0, "step": 800 }, { "epoch": 0.1018954331509986, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 1.2852760553359985, "learning_rate": 3.391267486222976e-06, "loss": 0.5008, "mean_token_accuracy": 0.8374513983726501, "num_tokens": 30312687.0, "step": 801 }, { "epoch": 0.10202264342958911, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 8.881092071533203e-06, "grad_norm": 1.323516607284546, "learning_rate": 3.395506570580755e-06, "loss": 0.5377, "mean_token_accuracy": 0.8273700475692749, "num_tokens": 30343993.0, "step": 802 }, { "epoch": 0.10214985370817962, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 1.314963459968567, "learning_rate": 3.3997456549385337e-06, "loss": 0.4855, "mean_token_accuracy": 0.8452428579330444, "num_tokens": 30378998.0, "step": 803 }, { "epoch": 0.10227706398677013, "ewc_loss": 0.01055908203125, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 1.3311092853546143, "learning_rate": 3.403984739296312e-06, "loss": 0.5525, "mean_token_accuracy": 0.8249311447143555, "num_tokens": 30412205.0, "step": 804 }, { "epoch": 0.10240427426536064, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 1.2914761304855347, "learning_rate": 3.408223823654091e-06, "loss": 0.4733, "mean_token_accuracy": 0.8477053642272949, "num_tokens": 30448489.0, "step": 805 }, { "epoch": 0.10253148454395115, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 1.287339687347412, "learning_rate": 3.41246290801187e-06, "loss": 0.5266, "mean_token_accuracy": 0.83990478515625, "num_tokens": 30488741.0, "step": 806 }, { "epoch": 0.10265869482254165, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 1.2982617616653442, "learning_rate": 3.416701992369648e-06, "loss": 0.5245, "mean_token_accuracy": 0.8406932950019836, "num_tokens": 30535715.0, "step": 807 }, { "epoch": 0.10278590510113217, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 1.2980378866195679, "learning_rate": 3.420941076727427e-06, "loss": 0.4816, "mean_token_accuracy": 0.8484015464782715, "num_tokens": 30574992.0, "step": 808 }, { "epoch": 0.10291311537972268, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 1.3390769958496094, "learning_rate": 3.4251801610852058e-06, "loss": 0.5514, "mean_token_accuracy": 0.8212478160858154, "num_tokens": 30615384.0, "step": 809 }, { "epoch": 0.1030403256583132, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 1.3186914920806885, "learning_rate": 3.4294192454429847e-06, "loss": 0.501, "mean_token_accuracy": 0.8410192131996155, "num_tokens": 30651011.0, "step": 810 }, { "epoch": 0.1031675359369037, "ewc_loss": 0.01043701171875, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 8.940696716308594e-06, "grad_norm": 1.3070491552352905, "learning_rate": 3.433658329800763e-06, "loss": 0.485, "mean_token_accuracy": 0.8455362319946289, "num_tokens": 30687213.0, "step": 811 }, { "epoch": 0.10329474621549421, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 1.537799835205078e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 1.2956327199935913, "learning_rate": 3.437897414158542e-06, "loss": 0.5237, "mean_token_accuracy": 0.8349817991256714, "num_tokens": 30726950.0, "step": 812 }, { "epoch": 0.10342195649408473, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 1.2725803852081299, "learning_rate": 3.442136498516321e-06, "loss": 0.4595, "mean_token_accuracy": 0.8517712950706482, "num_tokens": 30769359.0, "step": 813 }, { "epoch": 0.10354916677267523, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 1.2981911897659302, "learning_rate": 3.4463755828740998e-06, "loss": 0.5259, "mean_token_accuracy": 0.8347594738006592, "num_tokens": 30812480.0, "step": 814 }, { "epoch": 0.10367637705126574, "ewc_loss": 0.0108642578125, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 1.32181715965271, "learning_rate": 3.450614667231878e-06, "loss": 0.467, "mean_token_accuracy": 0.8474792242050171, "num_tokens": 30847287.0, "step": 815 }, { "epoch": 0.10380358732985626, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 1.300216555595398, "learning_rate": 3.4548537515896567e-06, "loss": 0.5141, "mean_token_accuracy": 0.8372403979301453, "num_tokens": 30887050.0, "step": 816 }, { "epoch": 0.10393079760844676, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 1.2931381464004517, "learning_rate": 3.4590928359474356e-06, "loss": 0.5137, "mean_token_accuracy": 0.8376274108886719, "num_tokens": 30926543.0, "step": 817 }, { "epoch": 0.10405800788703727, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 9.119510650634766e-06, "grad_norm": 1.303383469581604, "learning_rate": 3.4633319203052145e-06, "loss": 0.5334, "mean_token_accuracy": 0.8302116394042969, "num_tokens": 30966439.0, "step": 818 }, { "epoch": 0.10418521816562779, "ewc_loss": 0.01068115234375, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 1.3259676694869995, "learning_rate": 3.467571004662993e-06, "loss": 0.4742, "mean_token_accuracy": 0.8489317297935486, "num_tokens": 30999860.0, "step": 819 }, { "epoch": 0.1043124284442183, "ewc_loss": 0.0107421875, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 1.329086422920227, "learning_rate": 3.471810089020772e-06, "loss": 0.5194, "mean_token_accuracy": 0.8363648056983948, "num_tokens": 31037040.0, "step": 820 }, { "epoch": 0.1044396387228088, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 9.298324584960938e-06, "grad_norm": 1.3568840026855469, "learning_rate": 3.4760491733785507e-06, "loss": 0.5178, "mean_token_accuracy": 0.8301608562469482, "num_tokens": 31071508.0, "step": 821 }, { "epoch": 0.10456684900139931, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 1.5497207641601562e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 1.3107893466949463, "learning_rate": 3.4802882577363296e-06, "loss": 0.5005, "mean_token_accuracy": 0.8356336951255798, "num_tokens": 31107358.0, "step": 822 }, { "epoch": 0.10469405927998983, "ewc_loss": 0.0106201171875, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 9.059906005859375e-06, "grad_norm": 1.351542592048645, "learning_rate": 3.4845273420941076e-06, "loss": 0.5116, "mean_token_accuracy": 0.836925208568573, "num_tokens": 31144986.0, "step": 823 }, { "epoch": 0.10482126955858033, "ewc_loss": 0.01104736328125, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 9.47713851928711e-06, "grad_norm": 1.3368370532989502, "learning_rate": 3.4887664264518865e-06, "loss": 0.4987, "mean_token_accuracy": 0.8358753323554993, "num_tokens": 31179709.0, "step": 824 }, { "epoch": 0.10494847983717084, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 1.2900140285491943, "learning_rate": 3.4930055108096654e-06, "loss": 0.5133, "mean_token_accuracy": 0.8386144042015076, "num_tokens": 31222242.0, "step": 825 }, { "epoch": 0.10507569011576136, "ewc_loss": 0.0107421875, "ewc_loss_diag": 1.5616416931152344e-05, "ewc_loss_parallel": 9.179115295410156e-06, "grad_norm": 1.411661982536316, "learning_rate": 3.497244595167444e-06, "loss": 0.4882, "mean_token_accuracy": 0.8423819541931152, "num_tokens": 31264793.0, "step": 826 }, { "epoch": 0.10520290039435186, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 1.3527817726135254, "learning_rate": 3.5014836795252227e-06, "loss": 0.5487, "mean_token_accuracy": 0.8294345736503601, "num_tokens": 31300254.0, "step": 827 }, { "epoch": 0.10533011067294237, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 1.2964844703674316, "learning_rate": 3.5057227638830016e-06, "loss": 0.4871, "mean_token_accuracy": 0.8443282842636108, "num_tokens": 31339786.0, "step": 828 }, { "epoch": 0.10545732095153289, "ewc_loss": 0.01080322265625, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 9.238719940185547e-06, "grad_norm": 1.3334389925003052, "learning_rate": 3.5099618482407805e-06, "loss": 0.525, "mean_token_accuracy": 0.8312010765075684, "num_tokens": 31375100.0, "step": 829 }, { "epoch": 0.10558453123012339, "ewc_loss": 0.010986328125, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 1.313286304473877, "learning_rate": 3.514200932598559e-06, "loss": 0.5778, "mean_token_accuracy": 0.8195362091064453, "num_tokens": 31419000.0, "step": 830 }, { "epoch": 0.1057117415087139, "ewc_loss": 0.01092529296875, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 9.357929229736328e-06, "grad_norm": 1.3107141256332397, "learning_rate": 3.518440016956338e-06, "loss": 0.5193, "mean_token_accuracy": 0.8365944027900696, "num_tokens": 31459287.0, "step": 831 }, { "epoch": 0.10583895178730442, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 1.5735626220703125e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 1.308147668838501, "learning_rate": 3.5226791013141163e-06, "loss": 0.5268, "mean_token_accuracy": 0.8381286859512329, "num_tokens": 31502617.0, "step": 832 }, { "epoch": 0.10596616206589493, "ewc_loss": 0.0111083984375, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 9.5367431640625e-06, "grad_norm": 1.3151878118515015, "learning_rate": 3.526918185671895e-06, "loss": 0.5416, "mean_token_accuracy": 0.828467607498169, "num_tokens": 31545159.0, "step": 833 }, { "epoch": 0.10609337234448543, "ewc_loss": 0.01123046875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 1.3644564151763916, "learning_rate": 3.5311572700296736e-06, "loss": 0.468, "mean_token_accuracy": 0.852319598197937, "num_tokens": 31579193.0, "step": 834 }, { "epoch": 0.10622058262307595, "ewc_loss": 0.01129150390625, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 1.3201370239257812, "learning_rate": 3.5353963543874525e-06, "loss": 0.4861, "mean_token_accuracy": 0.8472769260406494, "num_tokens": 31618995.0, "step": 835 }, { "epoch": 0.10634779290166646, "ewc_loss": 0.010986328125, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 9.417533874511719e-06, "grad_norm": 1.32060968875885, "learning_rate": 3.5396354387452314e-06, "loss": 0.5124, "mean_token_accuracy": 0.8373017311096191, "num_tokens": 31657024.0, "step": 836 }, { "epoch": 0.10647500318025696, "ewc_loss": 0.01123046875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 1.2918576002120972, "learning_rate": 3.5438745231030103e-06, "loss": 0.5032, "mean_token_accuracy": 0.840856671333313, "num_tokens": 31700521.0, "step": 837 }, { "epoch": 0.10660221345884748, "ewc_loss": 0.01123046875, "ewc_loss_diag": 1.5974044799804688e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 1.3204280138015747, "learning_rate": 3.5481136074607888e-06, "loss": 0.4717, "mean_token_accuracy": 0.8526427745819092, "num_tokens": 31735942.0, "step": 838 }, { "epoch": 0.10672942373743799, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 1.3282936811447144, "learning_rate": 3.5523526918185676e-06, "loss": 0.5003, "mean_token_accuracy": 0.843300998210907, "num_tokens": 31774092.0, "step": 839 }, { "epoch": 0.10685663401602849, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 1.609325408935547e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 1.3084447383880615, "learning_rate": 3.5565917761763465e-06, "loss": 0.4802, "mean_token_accuracy": 0.8477402925491333, "num_tokens": 31816292.0, "step": 840 }, { "epoch": 0.106983844294619, "ewc_loss": 0.011474609375, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 1.3577951192855835, "learning_rate": 3.560830860534125e-06, "loss": 0.5002, "mean_token_accuracy": 0.8425639867782593, "num_tokens": 31848641.0, "step": 841 }, { "epoch": 0.10711105457320952, "ewc_loss": 0.011474609375, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 1.3694555759429932, "learning_rate": 3.5650699448919035e-06, "loss": 0.5508, "mean_token_accuracy": 0.826094388961792, "num_tokens": 31881292.0, "step": 842 }, { "epoch": 0.10723826485180002, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 9.775161743164062e-06, "grad_norm": 1.3432918787002563, "learning_rate": 3.5693090292496823e-06, "loss": 0.5087, "mean_token_accuracy": 0.8417543768882751, "num_tokens": 31915810.0, "step": 843 }, { "epoch": 0.10736547513039053, "ewc_loss": 0.01123046875, "ewc_loss_diag": 1.621246337890625e-05, "ewc_loss_parallel": 9.655952453613281e-06, "grad_norm": 1.3143625259399414, "learning_rate": 3.5735481136074612e-06, "loss": 0.4918, "mean_token_accuracy": 0.8438736200332642, "num_tokens": 31953621.0, "step": 844 }, { "epoch": 0.10749268540898105, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 1.3442654609680176, "learning_rate": 3.5777871979652397e-06, "loss": 0.5331, "mean_token_accuracy": 0.8311963677406311, "num_tokens": 31988738.0, "step": 845 }, { "epoch": 0.10761989568757156, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 1.3102123737335205, "learning_rate": 3.5820262823230186e-06, "loss": 0.4775, "mean_token_accuracy": 0.8524748682975769, "num_tokens": 32029184.0, "step": 846 }, { "epoch": 0.10774710596616206, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 1.3547426462173462, "learning_rate": 3.5862653666807975e-06, "loss": 0.5579, "mean_token_accuracy": 0.8227068781852722, "num_tokens": 32064413.0, "step": 847 }, { "epoch": 0.10787431624475258, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 1.3321731090545654, "learning_rate": 3.5905044510385763e-06, "loss": 0.5229, "mean_token_accuracy": 0.838483452796936, "num_tokens": 32104043.0, "step": 848 }, { "epoch": 0.10800152652334309, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 1.3379570245742798, "learning_rate": 3.5947435353963544e-06, "loss": 0.5091, "mean_token_accuracy": 0.8407419323921204, "num_tokens": 32140956.0, "step": 849 }, { "epoch": 0.10812873680193359, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 1.3569414615631104, "learning_rate": 3.5989826197541333e-06, "loss": 0.5575, "mean_token_accuracy": 0.8257617950439453, "num_tokens": 32181368.0, "step": 850 }, { "epoch": 0.1082559470805241, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 1.3974028825759888, "learning_rate": 3.603221704111912e-06, "loss": 0.507, "mean_token_accuracy": 0.8353146910667419, "num_tokens": 32222714.0, "step": 851 }, { "epoch": 0.10838315735911462, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 1.3032705783843994, "learning_rate": 3.607460788469691e-06, "loss": 0.4729, "mean_token_accuracy": 0.8532172441482544, "num_tokens": 32267870.0, "step": 852 }, { "epoch": 0.10851036763770512, "ewc_loss": 0.0113525390625, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 9.715557098388672e-06, "grad_norm": 1.3046174049377441, "learning_rate": 3.6116998728274695e-06, "loss": 0.447, "mean_token_accuracy": 0.8559247851371765, "num_tokens": 32307408.0, "step": 853 }, { "epoch": 0.10863757791629564, "ewc_loss": 0.011474609375, "ewc_loss_diag": 1.6450881958007812e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 1.3684147596359253, "learning_rate": 3.6159389571852484e-06, "loss": 0.582, "mean_token_accuracy": 0.8232913017272949, "num_tokens": 32343503.0, "step": 854 }, { "epoch": 0.10876478819488615, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 1.3211133480072021, "learning_rate": 3.6201780415430273e-06, "loss": 0.5244, "mean_token_accuracy": 0.8350250720977783, "num_tokens": 32386375.0, "step": 855 }, { "epoch": 0.10889199847347665, "ewc_loss": 0.011474609375, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 9.834766387939453e-06, "grad_norm": 1.3421385288238525, "learning_rate": 3.624417125900806e-06, "loss": 0.5372, "mean_token_accuracy": 0.8311670422554016, "num_tokens": 32424370.0, "step": 856 }, { "epoch": 0.10901920875206716, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 1.3509305715560913, "learning_rate": 3.628656210258584e-06, "loss": 0.5449, "mean_token_accuracy": 0.8275027275085449, "num_tokens": 32461998.0, "step": 857 }, { "epoch": 0.10914641903065768, "ewc_loss": 0.01153564453125, "ewc_loss_diag": 1.6570091247558594e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 1.3215196132659912, "learning_rate": 3.632895294616363e-06, "loss": 0.4816, "mean_token_accuracy": 0.8480135202407837, "num_tokens": 32498124.0, "step": 858 }, { "epoch": 0.1092736293092482, "ewc_loss": 0.01171875, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 1.3402771949768066, "learning_rate": 3.637134378974142e-06, "loss": 0.4761, "mean_token_accuracy": 0.8473742008209229, "num_tokens": 32537735.0, "step": 859 }, { "epoch": 0.1094008395878387, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 1.3388774394989014, "learning_rate": 3.641373463331921e-06, "loss": 0.5199, "mean_token_accuracy": 0.8373444080352783, "num_tokens": 32576219.0, "step": 860 }, { "epoch": 0.10952804986642921, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 1.6808509826660156e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 1.3461658954620361, "learning_rate": 3.6456125476896993e-06, "loss": 0.482, "mean_token_accuracy": 0.8407504558563232, "num_tokens": 32612868.0, "step": 861 }, { "epoch": 0.10965526014501972, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 1.3158930540084839, "learning_rate": 3.649851632047478e-06, "loss": 0.477, "mean_token_accuracy": 0.8508945107460022, "num_tokens": 32655412.0, "step": 862 }, { "epoch": 0.10978247042361022, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 1.3282380104064941, "learning_rate": 3.654090716405257e-06, "loss": 0.4446, "mean_token_accuracy": 0.8589927554130554, "num_tokens": 32692579.0, "step": 863 }, { "epoch": 0.10990968070220074, "ewc_loss": 0.01165771484375, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 9.953975677490234e-06, "grad_norm": 1.3465863466262817, "learning_rate": 3.658329800763035e-06, "loss": 0.541, "mean_token_accuracy": 0.8248670101165771, "num_tokens": 32731642.0, "step": 864 }, { "epoch": 0.11003689098079125, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 1.3599293231964111, "learning_rate": 3.662568885120814e-06, "loss": 0.5075, "mean_token_accuracy": 0.8376457691192627, "num_tokens": 32765707.0, "step": 865 }, { "epoch": 0.11016410125938175, "ewc_loss": 0.01171875, "ewc_loss_diag": 1.6927719116210938e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 1.3580116033554077, "learning_rate": 3.666807969478593e-06, "loss": 0.5387, "mean_token_accuracy": 0.826538622379303, "num_tokens": 32804613.0, "step": 866 }, { "epoch": 0.11029131153797227, "ewc_loss": 0.0115966796875, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 9.894371032714844e-06, "grad_norm": 1.354400396347046, "learning_rate": 3.6710470538363718e-06, "loss": 0.5235, "mean_token_accuracy": 0.8323677182197571, "num_tokens": 32841198.0, "step": 867 }, { "epoch": 0.11041852181656278, "ewc_loss": 0.01171875, "ewc_loss_diag": 1.704692840576172e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 1.34287428855896, "learning_rate": 3.6752861381941502e-06, "loss": 0.4888, "mean_token_accuracy": 0.8440888524055481, "num_tokens": 32878241.0, "step": 868 }, { "epoch": 0.11054573209515328, "ewc_loss": 0.01171875, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 1.0013580322265625e-05, "grad_norm": 1.3635064363479614, "learning_rate": 3.679525222551929e-06, "loss": 0.5499, "mean_token_accuracy": 0.8283203840255737, "num_tokens": 32915510.0, "step": 869 }, { "epoch": 0.1106729423737438, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 1.332119345664978, "learning_rate": 3.683764306909708e-06, "loss": 0.4488, "mean_token_accuracy": 0.8560129404067993, "num_tokens": 32953371.0, "step": 870 }, { "epoch": 0.11080015265233431, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 1.324485421180725, "learning_rate": 3.688003391267487e-06, "loss": 0.5099, "mean_token_accuracy": 0.8422969579696655, "num_tokens": 32998167.0, "step": 871 }, { "epoch": 0.11092736293092482, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 1.309643030166626, "learning_rate": 3.692242475625265e-06, "loss": 0.4797, "mean_token_accuracy": 0.8460581302642822, "num_tokens": 33046660.0, "step": 872 }, { "epoch": 0.11105457320951533, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 1.3951760530471802, "learning_rate": 3.696481559983044e-06, "loss": 0.5226, "mean_token_accuracy": 0.8315543532371521, "num_tokens": 33080672.0, "step": 873 }, { "epoch": 0.11118178348810584, "ewc_loss": 0.01220703125, "ewc_loss_diag": 1.71661376953125e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 1.3645038604736328, "learning_rate": 3.7007206443408227e-06, "loss": 0.4614, "mean_token_accuracy": 0.8511563539505005, "num_tokens": 33117023.0, "step": 874 }, { "epoch": 0.11130899376669635, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 1.3480186462402344, "learning_rate": 3.7049597286986016e-06, "loss": 0.5782, "mean_token_accuracy": 0.8190353512763977, "num_tokens": 33154843.0, "step": 875 }, { "epoch": 0.11143620404528685, "ewc_loss": 0.01177978515625, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 1.0073184967041016e-05, "grad_norm": 1.3585373163223267, "learning_rate": 3.70919881305638e-06, "loss": 0.532, "mean_token_accuracy": 0.8335843086242676, "num_tokens": 33193500.0, "step": 876 }, { "epoch": 0.11156341432387737, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 1.3262146711349487, "learning_rate": 3.713437897414159e-06, "loss": 0.523, "mean_token_accuracy": 0.8350891470909119, "num_tokens": 33240528.0, "step": 877 }, { "epoch": 0.11169062460246788, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 1.3658162355422974, "learning_rate": 3.717676981771938e-06, "loss": 0.5298, "mean_token_accuracy": 0.8322978019714355, "num_tokens": 33276959.0, "step": 878 }, { "epoch": 0.11181783488105838, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 1.351500391960144, "learning_rate": 3.721916066129716e-06, "loss": 0.4428, "mean_token_accuracy": 0.8581768870353699, "num_tokens": 33313573.0, "step": 879 }, { "epoch": 0.1119450451596489, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 1.3858485221862793, "learning_rate": 3.7261551504874947e-06, "loss": 0.5152, "mean_token_accuracy": 0.8402137756347656, "num_tokens": 33347408.0, "step": 880 }, { "epoch": 0.11207225543823941, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 1.355040192604065, "learning_rate": 3.7303942348452736e-06, "loss": 0.4481, "mean_token_accuracy": 0.8543903827667236, "num_tokens": 33381894.0, "step": 881 }, { "epoch": 0.11219946571682991, "ewc_loss": 0.0118408203125, "ewc_loss_diag": 1.728534698486328e-05, "ewc_loss_parallel": 1.0132789611816406e-05, "grad_norm": 1.3442147970199585, "learning_rate": 3.7346333192030525e-06, "loss": 0.4629, "mean_token_accuracy": 0.8545274138450623, "num_tokens": 33421746.0, "step": 882 }, { "epoch": 0.11232667599542043, "ewc_loss": 0.011962890625, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 1.0192394256591797e-05, "grad_norm": 1.3351556062698364, "learning_rate": 3.738872403560831e-06, "loss": 0.4842, "mean_token_accuracy": 0.8504900932312012, "num_tokens": 33465203.0, "step": 883 }, { "epoch": 0.11245388627401094, "ewc_loss": 0.011962890625, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 1.3717905282974243, "learning_rate": 3.74311148791861e-06, "loss": 0.5233, "mean_token_accuracy": 0.8341196775436401, "num_tokens": 33500333.0, "step": 884 }, { "epoch": 0.11258109655260146, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 1.3707871437072754, "learning_rate": 3.7473505722763887e-06, "loss": 0.5425, "mean_token_accuracy": 0.8269789218902588, "num_tokens": 33535180.0, "step": 885 }, { "epoch": 0.11270830683119196, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 1.35958993434906, "learning_rate": 3.7515896566341676e-06, "loss": 0.4938, "mean_token_accuracy": 0.841623842716217, "num_tokens": 33571298.0, "step": 886 }, { "epoch": 0.11283551710978247, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 1.7404556274414062e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 1.3459153175354004, "learning_rate": 3.7558287409919456e-06, "loss": 0.5193, "mean_token_accuracy": 0.8340732455253601, "num_tokens": 33613693.0, "step": 887 }, { "epoch": 0.11296272738837299, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 1.3886414766311646, "learning_rate": 3.7600678253497245e-06, "loss": 0.4837, "mean_token_accuracy": 0.8482223749160767, "num_tokens": 33649760.0, "step": 888 }, { "epoch": 0.11308993766696349, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 1.3268260955810547, "learning_rate": 3.7643069097075034e-06, "loss": 0.4463, "mean_token_accuracy": 0.8582066297531128, "num_tokens": 33692441.0, "step": 889 }, { "epoch": 0.113217147945554, "ewc_loss": 0.0120849609375, "ewc_loss_diag": 1.7523765563964844e-05, "ewc_loss_parallel": 1.0311603546142578e-05, "grad_norm": 1.3947440385818481, "learning_rate": 3.7685459940652823e-06, "loss": 0.4977, "mean_token_accuracy": 0.8410748243331909, "num_tokens": 33723034.0, "step": 890 }, { "epoch": 0.11334435822414451, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 1.352367639541626, "learning_rate": 3.7727850784230608e-06, "loss": 0.5135, "mean_token_accuracy": 0.8403768539428711, "num_tokens": 33764102.0, "step": 891 }, { "epoch": 0.11347156850273502, "ewc_loss": 0.01202392578125, "ewc_loss_diag": 1.7642974853515625e-05, "ewc_loss_parallel": 1.0251998901367188e-05, "grad_norm": 1.3421895503997803, "learning_rate": 3.7770241627808396e-06, "loss": 0.4723, "mean_token_accuracy": 0.8497428894042969, "num_tokens": 33803293.0, "step": 892 }, { "epoch": 0.11359877878132553, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 1.3638386726379395, "learning_rate": 3.7812632471386185e-06, "loss": 0.4732, "mean_token_accuracy": 0.84932941198349, "num_tokens": 33841681.0, "step": 893 }, { "epoch": 0.11372598905991604, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 1.3715167045593262, "learning_rate": 3.7855023314963974e-06, "loss": 0.4973, "mean_token_accuracy": 0.8369892835617065, "num_tokens": 33877207.0, "step": 894 }, { "epoch": 0.11385319933850654, "ewc_loss": 0.01220703125, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 1.3571581840515137, "learning_rate": 3.7897414158541754e-06, "loss": 0.5114, "mean_token_accuracy": 0.8352063894271851, "num_tokens": 33917703.0, "step": 895 }, { "epoch": 0.11398040961709706, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 1.3635097742080688, "learning_rate": 3.7939805002119543e-06, "loss": 0.4992, "mean_token_accuracy": 0.8434224128723145, "num_tokens": 33957007.0, "step": 896 }, { "epoch": 0.11410761989568757, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 1.3509306907653809, "learning_rate": 3.7982195845697332e-06, "loss": 0.4333, "mean_token_accuracy": 0.8600578308105469, "num_tokens": 33992615.0, "step": 897 }, { "epoch": 0.11423483017427809, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 1.3875961303710938, "learning_rate": 3.8024586689275117e-06, "loss": 0.5521, "mean_token_accuracy": 0.8300163745880127, "num_tokens": 34032049.0, "step": 898 }, { "epoch": 0.11436204045286859, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 1.3651031255722046, "learning_rate": 3.8066977532852906e-06, "loss": 0.4827, "mean_token_accuracy": 0.8441749811172485, "num_tokens": 34065978.0, "step": 899 }, { "epoch": 0.1144892507314591, "ewc_loss": 0.01214599609375, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.0371208190917969e-05, "grad_norm": 1.354283332824707, "learning_rate": 3.8109368376430694e-06, "loss": 0.4614, "mean_token_accuracy": 0.8538236021995544, "num_tokens": 34103454.0, "step": 900 }, { "epoch": 0.11461646101004962, "ewc_loss": 0.01220703125, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.043081283569336e-05, "grad_norm": 1.3702950477600098, "learning_rate": 3.815175922000848e-06, "loss": 0.439, "mean_token_accuracy": 0.8553876876831055, "num_tokens": 34137970.0, "step": 901 }, { "epoch": 0.11474367128864012, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 1.360499382019043, "learning_rate": 3.819415006358627e-06, "loss": 0.5336, "mean_token_accuracy": 0.8341877460479736, "num_tokens": 34176997.0, "step": 902 }, { "epoch": 0.11487088156723063, "ewc_loss": 0.01226806640625, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.049041748046875e-05, "grad_norm": 1.3693342208862305, "learning_rate": 3.823654090716406e-06, "loss": 0.505, "mean_token_accuracy": 0.8370002508163452, "num_tokens": 34215790.0, "step": 903 }, { "epoch": 0.11499809184582115, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 1.380274772644043, "learning_rate": 3.8278931750741846e-06, "loss": 0.5155, "mean_token_accuracy": 0.8431515693664551, "num_tokens": 34255212.0, "step": 904 }, { "epoch": 0.11512530212441165, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 1.3913813829421997, "learning_rate": 3.8321322594319634e-06, "loss": 0.4505, "mean_token_accuracy": 0.8565421104431152, "num_tokens": 34288432.0, "step": 905 }, { "epoch": 0.11525251240300216, "ewc_loss": 0.012451171875, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 1.40833580493927, "learning_rate": 3.8363713437897415e-06, "loss": 0.5585, "mean_token_accuracy": 0.822259783744812, "num_tokens": 34328991.0, "step": 906 }, { "epoch": 0.11537972268159268, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 1.7762184143066406e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 1.3859184980392456, "learning_rate": 3.84061042814752e-06, "loss": 0.547, "mean_token_accuracy": 0.8311216235160828, "num_tokens": 34366712.0, "step": 907 }, { "epoch": 0.11550693296018319, "ewc_loss": 0.01239013671875, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 1.0609626770019531e-05, "grad_norm": 1.3686437606811523, "learning_rate": 3.844849512505299e-06, "loss": 0.5334, "mean_token_accuracy": 0.829645037651062, "num_tokens": 34405757.0, "step": 908 }, { "epoch": 0.11563414323877369, "ewc_loss": 0.012451171875, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 1.0669231414794922e-05, "grad_norm": 1.360024094581604, "learning_rate": 3.849088596863078e-06, "loss": 0.4881, "mean_token_accuracy": 0.8463917970657349, "num_tokens": 34448019.0, "step": 909 }, { "epoch": 0.1157613535173642, "ewc_loss": 0.01251220703125, "ewc_loss_diag": 1.7881393432617188e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 1.354811429977417, "learning_rate": 3.853327681220856e-06, "loss": 0.5062, "mean_token_accuracy": 0.840682327747345, "num_tokens": 34491237.0, "step": 910 }, { "epoch": 0.11588856379595472, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 1.800060272216797e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 1.4016849994659424, "learning_rate": 3.857566765578635e-06, "loss": 0.4743, "mean_token_accuracy": 0.8478103876113892, "num_tokens": 34522842.0, "step": 911 }, { "epoch": 0.11601577407454522, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 1.3685131072998047, "learning_rate": 3.861805849936414e-06, "loss": 0.4983, "mean_token_accuracy": 0.845157265663147, "num_tokens": 34560551.0, "step": 912 }, { "epoch": 0.11614298435313573, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 1.3744820356369019, "learning_rate": 3.866044934294193e-06, "loss": 0.5458, "mean_token_accuracy": 0.8337178826332092, "num_tokens": 34603064.0, "step": 913 }, { "epoch": 0.11627019463172625, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 1.3922921419143677, "learning_rate": 3.870284018651972e-06, "loss": 0.5226, "mean_token_accuracy": 0.8352277278900146, "num_tokens": 34640283.0, "step": 914 }, { "epoch": 0.11639740491031675, "ewc_loss": 0.0126953125, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 1.369200587272644, "learning_rate": 3.874523103009751e-06, "loss": 0.4721, "mean_token_accuracy": 0.8517681956291199, "num_tokens": 34679840.0, "step": 915 }, { "epoch": 0.11652461518890726, "ewc_loss": 0.01263427734375, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 1.0788440704345703e-05, "grad_norm": 1.3711744546890259, "learning_rate": 3.8787621873675295e-06, "loss": 0.4596, "mean_token_accuracy": 0.853367030620575, "num_tokens": 34717045.0, "step": 916 }, { "epoch": 0.11665182546749778, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 1.360048532485962, "learning_rate": 3.8830012717253075e-06, "loss": 0.4629, "mean_token_accuracy": 0.8538987636566162, "num_tokens": 34755849.0, "step": 917 }, { "epoch": 0.11677903574608828, "ewc_loss": 0.0126953125, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 1.0848045349121094e-05, "grad_norm": 1.3725743293762207, "learning_rate": 3.887240356083086e-06, "loss": 0.437, "mean_token_accuracy": 0.8589205741882324, "num_tokens": 34795202.0, "step": 918 }, { "epoch": 0.11690624602467879, "ewc_loss": 0.01287841796875, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 1.4249496459960938, "learning_rate": 3.891479440440865e-06, "loss": 0.4754, "mean_token_accuracy": 0.8504859209060669, "num_tokens": 34824279.0, "step": 919 }, { "epoch": 0.1170334563032693, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 1.4243212938308716, "learning_rate": 3.895718524798644e-06, "loss": 0.5168, "mean_token_accuracy": 0.8352369070053101, "num_tokens": 34857631.0, "step": 920 }, { "epoch": 0.11716066658185982, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 1.811981201171875e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 1.377838373184204, "learning_rate": 3.899957609156422e-06, "loss": 0.4875, "mean_token_accuracy": 0.8456580638885498, "num_tokens": 34896969.0, "step": 921 }, { "epoch": 0.11728787686045032, "ewc_loss": 0.0125732421875, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 1.0728836059570312e-05, "grad_norm": 1.366031289100647, "learning_rate": 3.904196693514201e-06, "loss": 0.5052, "mean_token_accuracy": 0.8379824161529541, "num_tokens": 34942177.0, "step": 922 }, { "epoch": 0.11741508713904084, "ewc_loss": 0.01275634765625, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 1.0907649993896484e-05, "grad_norm": 1.405049204826355, "learning_rate": 3.90843577787198e-06, "loss": 0.5745, "mean_token_accuracy": 0.8205872178077698, "num_tokens": 34980440.0, "step": 923 }, { "epoch": 0.11754229741763135, "ewc_loss": 0.01287841796875, "ewc_loss_diag": 1.823902130126953e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 1.4046845436096191, "learning_rate": 3.912674862229759e-06, "loss": 0.5253, "mean_token_accuracy": 0.8335855007171631, "num_tokens": 35015343.0, "step": 924 }, { "epoch": 0.11766950769622185, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 1.3893651962280273, "learning_rate": 3.916913946587537e-06, "loss": 0.5166, "mean_token_accuracy": 0.8361846208572388, "num_tokens": 35050265.0, "step": 925 }, { "epoch": 0.11779671797481236, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 1.3858212232589722, "learning_rate": 3.921153030945316e-06, "loss": 0.5398, "mean_token_accuracy": 0.8290827870368958, "num_tokens": 35089235.0, "step": 926 }, { "epoch": 0.11792392825340288, "ewc_loss": 0.0128173828125, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 1.0967254638671875e-05, "grad_norm": 1.4058932065963745, "learning_rate": 3.925392115303095e-06, "loss": 0.4881, "mean_token_accuracy": 0.8446733951568604, "num_tokens": 35122887.0, "step": 927 }, { "epoch": 0.11805113853199338, "ewc_loss": 0.01287841796875, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 1.3957598209381104, "learning_rate": 3.9296311996608736e-06, "loss": 0.4919, "mean_token_accuracy": 0.8417593240737915, "num_tokens": 35158686.0, "step": 928 }, { "epoch": 0.1181783488105839, "ewc_loss": 0.01287841796875, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 1.385637879371643, "learning_rate": 3.9338702840186524e-06, "loss": 0.489, "mean_token_accuracy": 0.8422490954399109, "num_tokens": 35195215.0, "step": 929 }, { "epoch": 0.11830555908917441, "ewc_loss": 0.01287841796875, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 1.374435544013977, "learning_rate": 3.938109368376431e-06, "loss": 0.4937, "mean_token_accuracy": 0.8445974588394165, "num_tokens": 35236499.0, "step": 930 }, { "epoch": 0.11843276936776491, "ewc_loss": 0.01287841796875, "ewc_loss_diag": 1.8358230590820312e-05, "ewc_loss_parallel": 1.1026859283447266e-05, "grad_norm": 1.3838781118392944, "learning_rate": 3.94234845273421e-06, "loss": 0.5464, "mean_token_accuracy": 0.8278488516807556, "num_tokens": 35278809.0, "step": 931 }, { "epoch": 0.11855997964635542, "ewc_loss": 0.01300048828125, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 1.3962337970733643, "learning_rate": 3.946587537091989e-06, "loss": 0.4929, "mean_token_accuracy": 0.8420787453651428, "num_tokens": 35318342.0, "step": 932 }, { "epoch": 0.11868718992494594, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 1.8477439880371094e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 1.376676082611084, "learning_rate": 3.950826621449767e-06, "loss": 0.4917, "mean_token_accuracy": 0.8434760570526123, "num_tokens": 35361802.0, "step": 933 }, { "epoch": 0.11881440020353645, "ewc_loss": 0.012939453125, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 1.1086463928222656e-05, "grad_norm": 1.3846954107284546, "learning_rate": 3.955065705807546e-06, "loss": 0.4639, "mean_token_accuracy": 0.8526504039764404, "num_tokens": 35401244.0, "step": 934 }, { "epoch": 0.11894161048212695, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 1.4172710180282593, "learning_rate": 3.959304790165325e-06, "loss": 0.5025, "mean_token_accuracy": 0.8434377908706665, "num_tokens": 35438557.0, "step": 935 }, { "epoch": 0.11906882076071747, "ewc_loss": 0.01318359375, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 1.4129527807235718, "learning_rate": 3.963543874523103e-06, "loss": 0.4912, "mean_token_accuracy": 0.8455817699432373, "num_tokens": 35471686.0, "step": 936 }, { "epoch": 0.11919603103930798, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 1.8596649169921875e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 1.372721791267395, "learning_rate": 3.967782958880882e-06, "loss": 0.5171, "mean_token_accuracy": 0.8357709646224976, "num_tokens": 35514288.0, "step": 937 }, { "epoch": 0.11932324131789848, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 1.4181323051452637, "learning_rate": 3.972022043238661e-06, "loss": 0.5253, "mean_token_accuracy": 0.8335182666778564, "num_tokens": 35547588.0, "step": 938 }, { "epoch": 0.119450451596489, "ewc_loss": 0.01318359375, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 1.3998469114303589, "learning_rate": 3.97626112759644e-06, "loss": 0.463, "mean_token_accuracy": 0.8539185523986816, "num_tokens": 35582099.0, "step": 939 }, { "epoch": 0.11957766187507951, "ewc_loss": 0.0130615234375, "ewc_loss_diag": 1.8835067749023438e-05, "ewc_loss_parallel": 1.1146068572998047e-05, "grad_norm": 1.3920996189117432, "learning_rate": 3.980500211954218e-06, "loss": 0.4422, "mean_token_accuracy": 0.8571932315826416, "num_tokens": 35616173.0, "step": 940 }, { "epoch": 0.11970487215367001, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 1.4118411540985107, "learning_rate": 3.9847392963119965e-06, "loss": 0.5408, "mean_token_accuracy": 0.8312903642654419, "num_tokens": 35654801.0, "step": 941 }, { "epoch": 0.11983208243226053, "ewc_loss": 0.01318359375, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 1.3959120512008667, "learning_rate": 3.988978380669775e-06, "loss": 0.4867, "mean_token_accuracy": 0.8482363820075989, "num_tokens": 35692219.0, "step": 942 }, { "epoch": 0.11995929271085104, "ewc_loss": 0.01318359375, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 1.408196210861206, "learning_rate": 3.993217465027554e-06, "loss": 0.5479, "mean_token_accuracy": 0.8293308019638062, "num_tokens": 35732886.0, "step": 943 }, { "epoch": 0.12008650298944154, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 1.410752534866333, "learning_rate": 3.997456549385333e-06, "loss": 0.436, "mean_token_accuracy": 0.8611568212509155, "num_tokens": 35766796.0, "step": 944 }, { "epoch": 0.12021371326803205, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 1.4071959257125854, "learning_rate": 4.001695633743112e-06, "loss": 0.503, "mean_token_accuracy": 0.8399708271026611, "num_tokens": 35803065.0, "step": 945 }, { "epoch": 0.12034092354662257, "ewc_loss": 0.01318359375, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 1.4214037656784058, "learning_rate": 4.005934718100891e-06, "loss": 0.4925, "mean_token_accuracy": 0.8423236608505249, "num_tokens": 35836490.0, "step": 946 }, { "epoch": 0.12046813382521308, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 1.379956603050232, "learning_rate": 4.01017380245867e-06, "loss": 0.4512, "mean_token_accuracy": 0.8575645685195923, "num_tokens": 35877239.0, "step": 947 }, { "epoch": 0.12059534410380358, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 1.408074975013733, "learning_rate": 4.014412886816448e-06, "loss": 0.5584, "mean_token_accuracy": 0.8255354166030884, "num_tokens": 35917779.0, "step": 948 }, { "epoch": 0.1207225543823941, "ewc_loss": 0.01318359375, "ewc_loss_diag": 1.895427703857422e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 1.4019783735275269, "learning_rate": 4.018651971174227e-06, "loss": 0.4848, "mean_token_accuracy": 0.84590744972229, "num_tokens": 35955083.0, "step": 949 }, { "epoch": 0.12084976466098461, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 1.3883206844329834, "learning_rate": 4.022891055532006e-06, "loss": 0.5149, "mean_token_accuracy": 0.8361190557479858, "num_tokens": 35999581.0, "step": 950 }, { "epoch": 0.12097697493957511, "ewc_loss": 0.01318359375, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 1.1265277862548828e-05, "grad_norm": 1.4051097631454468, "learning_rate": 4.0271301398897845e-06, "loss": 0.5565, "mean_token_accuracy": 0.8296644687652588, "num_tokens": 36040891.0, "step": 951 }, { "epoch": 0.12110418521816563, "ewc_loss": 0.01324462890625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 1.1324882507324219e-05, "grad_norm": 1.4273661375045776, "learning_rate": 4.0313692242475625e-06, "loss": 0.4809, "mean_token_accuracy": 0.8438090682029724, "num_tokens": 36074958.0, "step": 952 }, { "epoch": 0.12123139549675614, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 1.3945218324661255, "learning_rate": 4.0356083086053414e-06, "loss": 0.512, "mean_token_accuracy": 0.8358927965164185, "num_tokens": 36115211.0, "step": 953 }, { "epoch": 0.12135860577534664, "ewc_loss": 0.01312255859375, "ewc_loss_diag": 1.9073486328125e-05, "ewc_loss_parallel": 1.1205673217773438e-05, "grad_norm": 1.3989864587783813, "learning_rate": 4.03984739296312e-06, "loss": 0.5123, "mean_token_accuracy": 0.8400416374206543, "num_tokens": 36156876.0, "step": 954 }, { "epoch": 0.12148581605393716, "ewc_loss": 0.0133056640625, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 1.3925244808197021, "learning_rate": 4.044086477320898e-06, "loss": 0.4907, "mean_token_accuracy": 0.8420555591583252, "num_tokens": 36197277.0, "step": 955 }, { "epoch": 0.12161302633252767, "ewc_loss": 0.013427734375, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 1.1444091796875e-05, "grad_norm": 1.433118224143982, "learning_rate": 4.048325561678677e-06, "loss": 0.5235, "mean_token_accuracy": 0.8345068097114563, "num_tokens": 36232839.0, "step": 956 }, { "epoch": 0.12174023661111817, "ewc_loss": 0.013427734375, "ewc_loss_diag": 1.919269561767578e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 1.407810926437378, "learning_rate": 4.052564646036456e-06, "loss": 0.5138, "mean_token_accuracy": 0.8365641832351685, "num_tokens": 36273531.0, "step": 957 }, { "epoch": 0.12186744688970869, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 1.4246841669082642, "learning_rate": 4.056803730394235e-06, "loss": 0.527, "mean_token_accuracy": 0.8315622210502625, "num_tokens": 36309325.0, "step": 958 }, { "epoch": 0.1219946571682992, "ewc_loss": 0.01336669921875, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 1.138448715209961e-05, "grad_norm": 1.4251210689544678, "learning_rate": 4.061042814752014e-06, "loss": 0.5036, "mean_token_accuracy": 0.8409768342971802, "num_tokens": 36344457.0, "step": 959 }, { "epoch": 0.12212186744688971, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 1.4548017978668213, "learning_rate": 4.065281899109793e-06, "loss": 0.5077, "mean_token_accuracy": 0.8363434672355652, "num_tokens": 36376731.0, "step": 960 }, { "epoch": 0.12224907772548022, "ewc_loss": 0.013427734375, "ewc_loss_diag": 1.9431114196777344e-05, "ewc_loss_parallel": 1.150369644165039e-05, "grad_norm": 1.3964601755142212, "learning_rate": 4.069520983467572e-06, "loss": 0.4607, "mean_token_accuracy": 0.8539259433746338, "num_tokens": 36415529.0, "step": 961 }, { "epoch": 0.12237628800407073, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 1.4292423725128174, "learning_rate": 4.0737600678253505e-06, "loss": 0.4887, "mean_token_accuracy": 0.8428483009338379, "num_tokens": 36449522.0, "step": 962 }, { "epoch": 0.12250349828266124, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 1.4165472984313965, "learning_rate": 4.077999152183129e-06, "loss": 0.4392, "mean_token_accuracy": 0.8592789173126221, "num_tokens": 36483010.0, "step": 963 }, { "epoch": 0.12263070856125174, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 1.3923075199127197, "learning_rate": 4.0822382365409075e-06, "loss": 0.5098, "mean_token_accuracy": 0.8409545421600342, "num_tokens": 36527993.0, "step": 964 }, { "epoch": 0.12275791883984226, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 1.4085609912872314, "learning_rate": 4.086477320898686e-06, "loss": 0.4812, "mean_token_accuracy": 0.8459006547927856, "num_tokens": 36568479.0, "step": 965 }, { "epoch": 0.12288512911843277, "ewc_loss": 0.01348876953125, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 1.1563301086425781e-05, "grad_norm": 1.4202210903167725, "learning_rate": 4.090716405256465e-06, "loss": 0.5199, "mean_token_accuracy": 0.8361964821815491, "num_tokens": 36607196.0, "step": 966 }, { "epoch": 0.12301233939702327, "ewc_loss": 0.01361083984375, "ewc_loss_diag": 1.9550323486328125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 1.436519980430603, "learning_rate": 4.094955489614243e-06, "loss": 0.4861, "mean_token_accuracy": 0.8446158170700073, "num_tokens": 36640845.0, "step": 967 }, { "epoch": 0.12313954967561379, "ewc_loss": 0.013671875, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 1.4232616424560547, "learning_rate": 4.099194573972022e-06, "loss": 0.4853, "mean_token_accuracy": 0.8467597961425781, "num_tokens": 36682182.0, "step": 968 }, { "epoch": 0.1232667599542043, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 1.4148091077804565, "learning_rate": 4.103433658329801e-06, "loss": 0.5289, "mean_token_accuracy": 0.8344615697860718, "num_tokens": 36723629.0, "step": 969 }, { "epoch": 0.1233939702327948, "ewc_loss": 0.013671875, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 1.4072203636169434, "learning_rate": 4.10767274268758e-06, "loss": 0.512, "mean_token_accuracy": 0.836979866027832, "num_tokens": 36763560.0, "step": 970 }, { "epoch": 0.12352118051138532, "ewc_loss": 0.013671875, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 1.4365023374557495, "learning_rate": 4.111911827045359e-06, "loss": 0.5776, "mean_token_accuracy": 0.8146895170211792, "num_tokens": 36802267.0, "step": 971 }, { "epoch": 0.12364839078997583, "ewc_loss": 0.013916015625, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 1.4195301532745361, "learning_rate": 4.116150911403137e-06, "loss": 0.4669, "mean_token_accuracy": 0.8537601232528687, "num_tokens": 36840047.0, "step": 972 }, { "epoch": 0.12377560106856635, "ewc_loss": 0.0135498046875, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 1.1622905731201172e-05, "grad_norm": 1.4290883541107178, "learning_rate": 4.120389995760916e-06, "loss": 0.5295, "mean_token_accuracy": 0.83449387550354, "num_tokens": 36876788.0, "step": 973 }, { "epoch": 0.12390281134715685, "ewc_loss": 0.013671875, "ewc_loss_diag": 1.9788742065429688e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 1.4412955045700073, "learning_rate": 4.124629080118695e-06, "loss": 0.4785, "mean_token_accuracy": 0.8453257083892822, "num_tokens": 36909954.0, "step": 974 }, { "epoch": 0.12403002162574736, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 1.431863784790039, "learning_rate": 4.1288681644764735e-06, "loss": 0.5176, "mean_token_accuracy": 0.8350636959075928, "num_tokens": 36948371.0, "step": 975 }, { "epoch": 0.12415723190433788, "ewc_loss": 0.013671875, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 1.1682510375976562e-05, "grad_norm": 1.4461876153945923, "learning_rate": 4.133107248834252e-06, "loss": 0.527, "mean_token_accuracy": 0.832181453704834, "num_tokens": 36983304.0, "step": 976 }, { "epoch": 0.12428444218292838, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 1.4121724367141724, "learning_rate": 4.137346333192031e-06, "loss": 0.4478, "mean_token_accuracy": 0.8559030294418335, "num_tokens": 37019644.0, "step": 977 }, { "epoch": 0.12441165246151889, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 1.4268468618392944, "learning_rate": 4.141585417549809e-06, "loss": 0.518, "mean_token_accuracy": 0.8341137170791626, "num_tokens": 37056906.0, "step": 978 }, { "epoch": 0.1245388627401094, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 1.470933437347412, "learning_rate": 4.145824501907588e-06, "loss": 0.4743, "mean_token_accuracy": 0.8471240401268005, "num_tokens": 37090396.0, "step": 979 }, { "epoch": 0.1246660730186999, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 1.4227174520492554, "learning_rate": 4.150063586265367e-06, "loss": 0.5118, "mean_token_accuracy": 0.8391639590263367, "num_tokens": 37129246.0, "step": 980 }, { "epoch": 0.12479328329729042, "ewc_loss": 0.0137939453125, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 1.1801719665527344e-05, "grad_norm": 1.4213682413101196, "learning_rate": 4.154302670623146e-06, "loss": 0.4589, "mean_token_accuracy": 0.8532071709632874, "num_tokens": 37166344.0, "step": 981 }, { "epoch": 0.12492049357588093, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 1.4173396825790405, "learning_rate": 4.158541754980924e-06, "loss": 0.4436, "mean_token_accuracy": 0.8603799343109131, "num_tokens": 37202827.0, "step": 982 }, { "epoch": 0.12504770385447145, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 1.4292734861373901, "learning_rate": 4.162780839338703e-06, "loss": 0.509, "mean_token_accuracy": 0.8352732062339783, "num_tokens": 37239606.0, "step": 983 }, { "epoch": 0.12517491413306195, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.002716064453125e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 1.4420984983444214, "learning_rate": 4.167019923696482e-06, "loss": 0.5044, "mean_token_accuracy": 0.8371231555938721, "num_tokens": 37275154.0, "step": 984 }, { "epoch": 0.12530212441165245, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 1.4216535091400146, "learning_rate": 4.171259008054261e-06, "loss": 0.4601, "mean_token_accuracy": 0.8549834489822388, "num_tokens": 37311849.0, "step": 985 }, { "epoch": 0.12542933469024298, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.014636993408203e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 1.4234931468963623, "learning_rate": 4.1754980924120395e-06, "loss": 0.5127, "mean_token_accuracy": 0.8392579555511475, "num_tokens": 37353174.0, "step": 986 }, { "epoch": 0.12555654496883348, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 1.4253435134887695, "learning_rate": 4.1797371767698184e-06, "loss": 0.4682, "mean_token_accuracy": 0.8522673845291138, "num_tokens": 37392081.0, "step": 987 }, { "epoch": 0.12568375524742398, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 1.423330307006836, "learning_rate": 4.183976261127597e-06, "loss": 0.4662, "mean_token_accuracy": 0.851913571357727, "num_tokens": 37430639.0, "step": 988 }, { "epoch": 0.1258109655260145, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 1.460060477256775, "learning_rate": 4.188215345485375e-06, "loss": 0.5522, "mean_token_accuracy": 0.8240500688552856, "num_tokens": 37467126.0, "step": 989 }, { "epoch": 0.125938175804605, "ewc_loss": 0.01409912109375, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 1.4375606775283813, "learning_rate": 4.192454429843154e-06, "loss": 0.484, "mean_token_accuracy": 0.8477196097373962, "num_tokens": 37505657.0, "step": 990 }, { "epoch": 0.12606538608319554, "ewc_loss": 0.01397705078125, "ewc_loss_diag": 2.0265579223632812e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 1.419459342956543, "learning_rate": 4.196693514200933e-06, "loss": 0.4893, "mean_token_accuracy": 0.8459628820419312, "num_tokens": 37549262.0, "step": 991 }, { "epoch": 0.12619259636178604, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 1.4252500534057617, "learning_rate": 4.200932598558712e-06, "loss": 0.5022, "mean_token_accuracy": 0.8391332030296326, "num_tokens": 37589967.0, "step": 992 }, { "epoch": 0.12631980664037654, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 1.4194782972335815, "learning_rate": 4.20517168291649e-06, "loss": 0.5049, "mean_token_accuracy": 0.8416962623596191, "num_tokens": 37635686.0, "step": 993 }, { "epoch": 0.12644701691896706, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 1.4445990324020386, "learning_rate": 4.209410767274269e-06, "loss": 0.5456, "mean_token_accuracy": 0.8334853649139404, "num_tokens": 37673753.0, "step": 994 }, { "epoch": 0.12657422719755757, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.2159347534179688e-05, "grad_norm": 1.452283263206482, "learning_rate": 4.213649851632048e-06, "loss": 0.4548, "mean_token_accuracy": 0.8558390140533447, "num_tokens": 37706345.0, "step": 995 }, { "epoch": 0.12670143747614807, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 1.4326591491699219, "learning_rate": 4.217888935989827e-06, "loss": 0.5259, "mean_token_accuracy": 0.8313740491867065, "num_tokens": 37750356.0, "step": 996 }, { "epoch": 0.1268286477547386, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.1861324310302734e-05, "grad_norm": 1.444949746131897, "learning_rate": 4.222128020347605e-06, "loss": 0.5186, "mean_token_accuracy": 0.8363001942634583, "num_tokens": 37791466.0, "step": 997 }, { "epoch": 0.1269558580333291, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 1.436522126197815, "learning_rate": 4.226367104705384e-06, "loss": 0.5247, "mean_token_accuracy": 0.8331887722015381, "num_tokens": 37837257.0, "step": 998 }, { "epoch": 0.1270830683119196, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.2040138244628906e-05, "grad_norm": 1.434924840927124, "learning_rate": 4.2306061890631625e-06, "loss": 0.5126, "mean_token_accuracy": 0.838456928730011, "num_tokens": 37880223.0, "step": 999 }, { "epoch": 0.12721027859051012, "ewc_loss": 0.013916015625, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.1920928955078125e-05, "grad_norm": 1.4743484258651733, "learning_rate": 4.234845273420941e-06, "loss": 0.5918, "mean_token_accuracy": 0.8209471702575684, "num_tokens": 37917662.0, "step": 1000 }, { "epoch": 0.12733748886910062, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 1.4228075742721558, "learning_rate": 4.23908435777872e-06, "loss": 0.518, "mean_token_accuracy": 0.8332815170288086, "num_tokens": 37961985.0, "step": 1001 }, { "epoch": 0.12746469914769112, "ewc_loss": 0.0140380859375, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.1980533599853516e-05, "grad_norm": 1.4481511116027832, "learning_rate": 4.243323442136499e-06, "loss": 0.5063, "mean_token_accuracy": 0.842710554599762, "num_tokens": 37998960.0, "step": 1002 }, { "epoch": 0.12759190942628165, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 1.444823145866394, "learning_rate": 4.247562526494278e-06, "loss": 0.4796, "mean_token_accuracy": 0.8491517901420593, "num_tokens": 38040920.0, "step": 1003 }, { "epoch": 0.12771911970487215, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.0384788513183594e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 1.490411639213562, "learning_rate": 4.251801610852057e-06, "loss": 0.5261, "mean_token_accuracy": 0.8344782590866089, "num_tokens": 38071814.0, "step": 1004 }, { "epoch": 0.12784632998346265, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 1.442611575126648, "learning_rate": 4.256040695209835e-06, "loss": 0.5441, "mean_token_accuracy": 0.8306417465209961, "num_tokens": 38112784.0, "step": 1005 }, { "epoch": 0.12797354026205318, "ewc_loss": 0.01416015625, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 1.2099742889404297e-05, "grad_norm": 1.4450631141662598, "learning_rate": 4.260279779567614e-06, "loss": 0.5596, "mean_token_accuracy": 0.8281978368759155, "num_tokens": 38152140.0, "step": 1006 }, { "epoch": 0.12810075054064368, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 1.4387140274047852, "learning_rate": 4.264518863925393e-06, "loss": 0.4636, "mean_token_accuracy": 0.8525104522705078, "num_tokens": 38188293.0, "step": 1007 }, { "epoch": 0.12822796081923418, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.0503997802734375e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 1.4542909860610962, "learning_rate": 4.268757948283171e-06, "loss": 0.517, "mean_token_accuracy": 0.8367630243301392, "num_tokens": 38229136.0, "step": 1008 }, { "epoch": 0.1283551710978247, "ewc_loss": 0.0142822265625, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 1.2218952178955078e-05, "grad_norm": 1.4368896484375, "learning_rate": 4.27299703264095e-06, "loss": 0.5344, "mean_token_accuracy": 0.8327143788337708, "num_tokens": 38273330.0, "step": 1009 }, { "epoch": 0.1284823813764152, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 1.4453779458999634, "learning_rate": 4.2772361169987285e-06, "loss": 0.4979, "mean_token_accuracy": 0.8417810201644897, "num_tokens": 38315066.0, "step": 1010 }, { "epoch": 0.1286095916550057, "ewc_loss": 0.01434326171875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 1.4613192081451416, "learning_rate": 4.281475201356507e-06, "loss": 0.5034, "mean_token_accuracy": 0.8407787084579468, "num_tokens": 38353430.0, "step": 1011 }, { "epoch": 0.12873680193359624, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 1.4565118551254272, "learning_rate": 4.2857142857142855e-06, "loss": 0.5358, "mean_token_accuracy": 0.8334738612174988, "num_tokens": 38391919.0, "step": 1012 }, { "epoch": 0.12886401221218674, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 1.4461572170257568, "learning_rate": 4.289953370072064e-06, "loss": 0.4782, "mean_token_accuracy": 0.8499376177787781, "num_tokens": 38430104.0, "step": 1013 }, { "epoch": 0.12899122249077727, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 1.4587194919586182, "learning_rate": 4.294192454429843e-06, "loss": 0.5246, "mean_token_accuracy": 0.83592689037323, "num_tokens": 38468775.0, "step": 1014 }, { "epoch": 0.12911843276936777, "ewc_loss": 0.0145263671875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 1.245737075805664e-05, "grad_norm": 1.4201738834381104, "learning_rate": 4.298431538787622e-06, "loss": 0.4121, "mean_token_accuracy": 0.8699671030044556, "num_tokens": 38510330.0, "step": 1015 }, { "epoch": 0.12924564304795827, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 1.233816146850586e-05, "grad_norm": 1.4828522205352783, "learning_rate": 4.302670623145401e-06, "loss": 0.5524, "mean_token_accuracy": 0.8270365595817566, "num_tokens": 38543840.0, "step": 1016 }, { "epoch": 0.1293728533265488, "ewc_loss": 0.01483154296875, "ewc_loss_diag": 2.0623207092285156e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 1.4676413536071777, "learning_rate": 4.30690970750318e-06, "loss": 0.506, "mean_token_accuracy": 0.842329204082489, "num_tokens": 38581670.0, "step": 1017 }, { "epoch": 0.1295000636051393, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 1.4889214038848877, "learning_rate": 4.311148791860959e-06, "loss": 0.4936, "mean_token_accuracy": 0.8403717279434204, "num_tokens": 38612642.0, "step": 1018 }, { "epoch": 0.1296272738837298, "ewc_loss": 0.01495361328125, "ewc_loss_diag": 2.0742416381835938e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 1.4642289876937866, "learning_rate": 4.315387876218738e-06, "loss": 0.5323, "mean_token_accuracy": 0.8364929556846619, "num_tokens": 38647579.0, "step": 1019 }, { "epoch": 0.12975448416232033, "ewc_loss": 0.01446533203125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 1.239776611328125e-05, "grad_norm": 1.4737260341644287, "learning_rate": 4.319626960576516e-06, "loss": 0.5313, "mean_token_accuracy": 0.8296335339546204, "num_tokens": 38683114.0, "step": 1020 }, { "epoch": 0.12988169444091083, "ewc_loss": 0.01470947265625, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 1.4570086002349854, "learning_rate": 4.3238660449342946e-06, "loss": 0.4755, "mean_token_accuracy": 0.8471318483352661, "num_tokens": 38722990.0, "step": 1021 }, { "epoch": 0.13000890471950133, "ewc_loss": 0.0146484375, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 1.2576580047607422e-05, "grad_norm": 1.4449931383132935, "learning_rate": 4.3281051292920735e-06, "loss": 0.5014, "mean_token_accuracy": 0.8412937521934509, "num_tokens": 38765041.0, "step": 1022 }, { "epoch": 0.13013611499809186, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 2.086162567138672e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 1.4495952129364014, "learning_rate": 4.332344213649852e-06, "loss": 0.4666, "mean_token_accuracy": 0.850371241569519, "num_tokens": 38802241.0, "step": 1023 }, { "epoch": 0.13026332527668236, "ewc_loss": 0.014892578125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 1.2755393981933594e-05, "grad_norm": 1.4779350757598877, "learning_rate": 4.33658329800763e-06, "loss": 0.4941, "mean_token_accuracy": 0.8456079363822937, "num_tokens": 38839444.0, "step": 1024 }, { "epoch": 0.13039053555527286, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 1.4427748918533325, "learning_rate": 4.340822382365409e-06, "loss": 0.4766, "mean_token_accuracy": 0.8489140272140503, "num_tokens": 38882102.0, "step": 1025 }, { "epoch": 0.13051774583386339, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 1.2636184692382812e-05, "grad_norm": 1.4442758560180664, "learning_rate": 4.345061466723188e-06, "loss": 0.4658, "mean_token_accuracy": 0.8538074493408203, "num_tokens": 38925855.0, "step": 1026 }, { "epoch": 0.13064495611245389, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 2.09808349609375e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 1.4803067445755005, "learning_rate": 4.349300551080966e-06, "loss": 0.5228, "mean_token_accuracy": 0.8362942337989807, "num_tokens": 38962293.0, "step": 1027 }, { "epoch": 0.1307721663910444, "ewc_loss": 0.014892578125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 1.2814998626708984e-05, "grad_norm": 1.4700263738632202, "learning_rate": 4.353539635438745e-06, "loss": 0.4901, "mean_token_accuracy": 0.842902421951294, "num_tokens": 38999000.0, "step": 1028 }, { "epoch": 0.13089937666963491, "ewc_loss": 0.0147705078125, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 1.2695789337158203e-05, "grad_norm": 1.5931390523910522, "learning_rate": 4.357778719796524e-06, "loss": 0.5513, "mean_token_accuracy": 0.823727011680603, "num_tokens": 39032819.0, "step": 1029 }, { "epoch": 0.13102658694822542, "ewc_loss": 0.0155029296875, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 1.4593685865402222, "learning_rate": 4.362017804154303e-06, "loss": 0.4742, "mean_token_accuracy": 0.8514619469642639, "num_tokens": 39070247.0, "step": 1030 }, { "epoch": 0.13115379722681592, "ewc_loss": 0.014404296875, "ewc_loss_diag": 2.110004425048828e-05, "ewc_loss_parallel": 1.2278556823730469e-05, "grad_norm": 1.4662376642227173, "learning_rate": 4.366256888512082e-06, "loss": 0.4703, "mean_token_accuracy": 0.8523164391517639, "num_tokens": 39105310.0, "step": 1031 }, { "epoch": 0.13128100750540644, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 1.4631253480911255, "learning_rate": 4.370495972869861e-06, "loss": 0.502, "mean_token_accuracy": 0.8403651714324951, "num_tokens": 39144218.0, "step": 1032 }, { "epoch": 0.13140821778399694, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 1.4953619241714478, "learning_rate": 4.3747350572276395e-06, "loss": 0.4922, "mean_token_accuracy": 0.8430382609367371, "num_tokens": 39177248.0, "step": 1033 }, { "epoch": 0.13153542806258745, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 2.1219253540039062e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 1.4619615077972412, "learning_rate": 4.378974141585418e-06, "loss": 0.5016, "mean_token_accuracy": 0.8418292999267578, "num_tokens": 39214230.0, "step": 1034 }, { "epoch": 0.13166263834117797, "ewc_loss": 0.0150146484375, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 1.2874603271484375e-05, "grad_norm": 1.4604604244232178, "learning_rate": 4.383213225943196e-06, "loss": 0.4935, "mean_token_accuracy": 0.8413135409355164, "num_tokens": 39252189.0, "step": 1035 }, { "epoch": 0.13178984861976847, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 1.4675596952438354, "learning_rate": 4.387452310300975e-06, "loss": 0.4433, "mean_token_accuracy": 0.859804093837738, "num_tokens": 39288058.0, "step": 1036 }, { "epoch": 0.13191705889835897, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 1.4720326662063599, "learning_rate": 4.391691394658754e-06, "loss": 0.4518, "mean_token_accuracy": 0.8538029193878174, "num_tokens": 39322157.0, "step": 1037 }, { "epoch": 0.1320442691769495, "ewc_loss": 0.01513671875, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 1.2993812561035156e-05, "grad_norm": 1.476894736289978, "learning_rate": 4.395930479016533e-06, "loss": 0.4679, "mean_token_accuracy": 0.8491065502166748, "num_tokens": 39357678.0, "step": 1038 }, { "epoch": 0.13217147945554, "ewc_loss": 0.01519775390625, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 1.3053417205810547e-05, "grad_norm": 1.4554014205932617, "learning_rate": 4.400169563374311e-06, "loss": 0.4888, "mean_token_accuracy": 0.844788670539856, "num_tokens": 39398909.0, "step": 1039 }, { "epoch": 0.13229868973413053, "ewc_loss": 0.0152587890625, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 1.5183333158493042, "learning_rate": 4.40440864773209e-06, "loss": 0.5653, "mean_token_accuracy": 0.8244225978851318, "num_tokens": 39437934.0, "step": 1040 }, { "epoch": 0.13242590001272103, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 2.1457672119140625e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 1.4801592826843262, "learning_rate": 4.408647732089869e-06, "loss": 0.486, "mean_token_accuracy": 0.8427016139030457, "num_tokens": 39471772.0, "step": 1041 }, { "epoch": 0.13255311029131153, "ewc_loss": 0.01507568359375, "ewc_loss_diag": 2.1576881408691406e-05, "ewc_loss_parallel": 1.2934207916259766e-05, "grad_norm": 1.476773738861084, "learning_rate": 4.412886816447648e-06, "loss": 0.5008, "mean_token_accuracy": 0.8459926843643188, "num_tokens": 39512398.0, "step": 1042 }, { "epoch": 0.13268032056990206, "ewc_loss": 0.015380859375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 1.472691297531128, "learning_rate": 4.417125900805427e-06, "loss": 0.4943, "mean_token_accuracy": 0.8445405960083008, "num_tokens": 39554735.0, "step": 1043 }, { "epoch": 0.13280753084849256, "ewc_loss": 0.015380859375, "ewc_loss_diag": 2.1696090698242188e-05, "ewc_loss_parallel": 1.3172626495361328e-05, "grad_norm": 1.4793983697891235, "learning_rate": 4.4213649851632055e-06, "loss": 0.5436, "mean_token_accuracy": 0.8331575989723206, "num_tokens": 39595397.0, "step": 1044 }, { "epoch": 0.13293474112708306, "ewc_loss": 0.015625, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 1.4844019412994385, "learning_rate": 4.4256040695209836e-06, "loss": 0.528, "mean_token_accuracy": 0.8346663117408752, "num_tokens": 39632474.0, "step": 1045 }, { "epoch": 0.1330619514056736, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 1.4973549842834473, "learning_rate": 4.4298431538787625e-06, "loss": 0.5139, "mean_token_accuracy": 0.83356773853302, "num_tokens": 39666408.0, "step": 1046 }, { "epoch": 0.1331891616842641, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 1.4766883850097656, "learning_rate": 4.434082238236541e-06, "loss": 0.5499, "mean_token_accuracy": 0.8265719413757324, "num_tokens": 39709033.0, "step": 1047 }, { "epoch": 0.1333163719628546, "ewc_loss": 0.01544189453125, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 1.3232231140136719e-05, "grad_norm": 1.4737915992736816, "learning_rate": 4.43832132259432e-06, "loss": 0.5211, "mean_token_accuracy": 0.8379151225090027, "num_tokens": 39749732.0, "step": 1048 }, { "epoch": 0.13344358224144512, "ewc_loss": 0.015625, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 1.4921221733093262, "learning_rate": 4.442560406952099e-06, "loss": 0.5299, "mean_token_accuracy": 0.8301708102226257, "num_tokens": 39788125.0, "step": 1049 }, { "epoch": 0.13357079252003562, "ewc_loss": 0.015625, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 1.466200590133667, "learning_rate": 4.446799491309877e-06, "loss": 0.472, "mean_token_accuracy": 0.8524012565612793, "num_tokens": 39830125.0, "step": 1050 }, { "epoch": 0.13369800279862612, "ewc_loss": 0.015625, "ewc_loss_diag": 2.181529998779297e-05, "ewc_loss_parallel": 1.341104507446289e-05, "grad_norm": 1.482890248298645, "learning_rate": 4.451038575667656e-06, "loss": 0.528, "mean_token_accuracy": 0.8315531611442566, "num_tokens": 39868236.0, "step": 1051 }, { "epoch": 0.13382521307721665, "ewc_loss": 0.015869140625, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 1.4587478637695312, "learning_rate": 4.455277660025435e-06, "loss": 0.4959, "mean_token_accuracy": 0.8423044681549072, "num_tokens": 39913254.0, "step": 1052 }, { "epoch": 0.13395242335580715, "ewc_loss": 0.01556396484375, "ewc_loss_diag": 2.193450927734375e-05, "ewc_loss_parallel": 1.33514404296875e-05, "grad_norm": 1.5734933614730835, "learning_rate": 4.459516744383214e-06, "loss": 0.5055, "mean_token_accuracy": 0.8391901254653931, "num_tokens": 39948242.0, "step": 1053 }, { "epoch": 0.13407963363439765, "ewc_loss": 0.01611328125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 1.4954754114151, "learning_rate": 4.463755828740992e-06, "loss": 0.5216, "mean_token_accuracy": 0.8357042074203491, "num_tokens": 39989196.0, "step": 1054 }, { "epoch": 0.13420684391298818, "ewc_loss": 0.01531982421875, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 1.3113021850585938e-05, "grad_norm": 1.4873188734054565, "learning_rate": 4.467994913098771e-06, "loss": 0.5204, "mean_token_accuracy": 0.843632698059082, "num_tokens": 40028662.0, "step": 1055 }, { "epoch": 0.13433405419157868, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 1.4884772300720215, "learning_rate": 4.47223399745655e-06, "loss": 0.521, "mean_token_accuracy": 0.8361245393753052, "num_tokens": 40066178.0, "step": 1056 }, { "epoch": 0.13446126447016918, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 1.3470649719238281e-05, "grad_norm": 1.4711012840270996, "learning_rate": 4.4764730818143285e-06, "loss": 0.5126, "mean_token_accuracy": 0.83695387840271, "num_tokens": 40106908.0, "step": 1057 }, { "epoch": 0.1345884747487597, "ewc_loss": 0.015869140625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 1.4860475063323975, "learning_rate": 4.480712166172107e-06, "loss": 0.4755, "mean_token_accuracy": 0.8500763177871704, "num_tokens": 40146872.0, "step": 1058 }, { "epoch": 0.1347156850273502, "ewc_loss": 0.015869140625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 1.487589716911316, "learning_rate": 4.484951250529886e-06, "loss": 0.5055, "mean_token_accuracy": 0.840183675289154, "num_tokens": 40186717.0, "step": 1059 }, { "epoch": 0.1348428953059407, "ewc_loss": 0.015869140625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 1.4894297122955322, "learning_rate": 4.489190334887665e-06, "loss": 0.5438, "mean_token_accuracy": 0.833533763885498, "num_tokens": 40227119.0, "step": 1060 }, { "epoch": 0.13497010558453124, "ewc_loss": 0.015869140625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 1.5016002655029297, "learning_rate": 4.493429419245444e-06, "loss": 0.5216, "mean_token_accuracy": 0.8366162180900574, "num_tokens": 40262302.0, "step": 1061 }, { "epoch": 0.13509731586312174, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 1.4967163801193237, "learning_rate": 4.497668503603222e-06, "loss": 0.4686, "mean_token_accuracy": 0.85190349817276, "num_tokens": 40296783.0, "step": 1062 }, { "epoch": 0.13522452614171224, "ewc_loss": 0.015869140625, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 1.5059245824813843, "learning_rate": 4.501907587961001e-06, "loss": 0.531, "mean_token_accuracy": 0.8302266001701355, "num_tokens": 40331323.0, "step": 1063 }, { "epoch": 0.13535173642030277, "ewc_loss": 0.01611328125, "ewc_loss_diag": 2.205371856689453e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 1.4852874279022217, "learning_rate": 4.50614667231878e-06, "loss": 0.5026, "mean_token_accuracy": 0.8408657908439636, "num_tokens": 40369965.0, "step": 1064 }, { "epoch": 0.13547894669889327, "ewc_loss": 0.0157470703125, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 1.3530254364013672e-05, "grad_norm": 1.4781633615493774, "learning_rate": 4.510385756676558e-06, "loss": 0.5518, "mean_token_accuracy": 0.8303608894348145, "num_tokens": 40410820.0, "step": 1065 }, { "epoch": 0.1356061569774838, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 1.4902665615081787, "learning_rate": 4.514624841034337e-06, "loss": 0.4566, "mean_token_accuracy": 0.8549810647964478, "num_tokens": 40446942.0, "step": 1066 }, { "epoch": 0.1357333672560743, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 1.488554835319519, "learning_rate": 4.518863925392116e-06, "loss": 0.4738, "mean_token_accuracy": 0.8504232168197632, "num_tokens": 40484684.0, "step": 1067 }, { "epoch": 0.1358605775346648, "ewc_loss": 0.01611328125, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 1.4866304397583008, "learning_rate": 4.5231030097498945e-06, "loss": 0.4533, "mean_token_accuracy": 0.8531062006950378, "num_tokens": 40521164.0, "step": 1068 }, { "epoch": 0.13598778781325532, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 2.2292137145996094e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 1.4709348678588867, "learning_rate": 4.5273420941076726e-06, "loss": 0.4891, "mean_token_accuracy": 0.8434244990348816, "num_tokens": 40566192.0, "step": 1069 }, { "epoch": 0.13611499809184582, "ewc_loss": 0.015869140625, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 1.3589859008789062e-05, "grad_norm": 1.4968091249465942, "learning_rate": 4.5315811784654514e-06, "loss": 0.5237, "mean_token_accuracy": 0.8351877927780151, "num_tokens": 40605534.0, "step": 1070 }, { "epoch": 0.13624220837043632, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 2.2411346435546875e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 1.486167311668396, "learning_rate": 4.53582026282323e-06, "loss": 0.4775, "mean_token_accuracy": 0.8469669818878174, "num_tokens": 40645478.0, "step": 1071 }, { "epoch": 0.13636941864902685, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 2.2530555725097656e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 1.510646104812622, "learning_rate": 4.540059347181009e-06, "loss": 0.5622, "mean_token_accuracy": 0.822885274887085, "num_tokens": 40684591.0, "step": 1072 }, { "epoch": 0.13649662892761735, "ewc_loss": 0.01611328125, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 1.4883524179458618, "learning_rate": 4.544298431538788e-06, "loss": 0.4968, "mean_token_accuracy": 0.8423787951469421, "num_tokens": 40724437.0, "step": 1073 }, { "epoch": 0.13662383920620785, "ewc_loss": 0.01611328125, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 1.3768672943115234e-05, "grad_norm": 1.5237051248550415, "learning_rate": 4.548537515896567e-06, "loss": 0.5413, "mean_token_accuracy": 0.8291454315185547, "num_tokens": 40758324.0, "step": 1074 }, { "epoch": 0.13675104948479838, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 1.3887882232666016e-05, "grad_norm": 1.5170347690582275, "learning_rate": 4.552776600254346e-06, "loss": 0.5252, "mean_token_accuracy": 0.834648609161377, "num_tokens": 40794182.0, "step": 1075 }, { "epoch": 0.13687825976338888, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 1.3709068298339844e-05, "grad_norm": 1.5190268754959106, "learning_rate": 4.557015684612125e-06, "loss": 0.4743, "mean_token_accuracy": 0.8511145114898682, "num_tokens": 40832293.0, "step": 1076 }, { "epoch": 0.13700547004197938, "ewc_loss": 0.01611328125, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 1.4847171306610107, "learning_rate": 4.561254768969903e-06, "loss": 0.4699, "mean_token_accuracy": 0.8495993614196777, "num_tokens": 40871078.0, "step": 1077 }, { "epoch": 0.1371326803205699, "ewc_loss": 0.0159912109375, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 1.3649463653564453e-05, "grad_norm": 1.4896118640899658, "learning_rate": 4.565493853327682e-06, "loss": 0.4807, "mean_token_accuracy": 0.8491861820220947, "num_tokens": 40909061.0, "step": 1078 }, { "epoch": 0.1372598905991604, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 1.5143778324127197, "learning_rate": 4.5697329376854606e-06, "loss": 0.5189, "mean_token_accuracy": 0.8415837287902832, "num_tokens": 40945623.0, "step": 1079 }, { "epoch": 0.1373871008777509, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 1.512041687965393, "learning_rate": 4.573972022043239e-06, "loss": 0.5836, "mean_token_accuracy": 0.8166306614875793, "num_tokens": 40985009.0, "step": 1080 }, { "epoch": 0.13751431115634144, "ewc_loss": 0.01611328125, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 1.491444706916809, "learning_rate": 4.5782111064010175e-06, "loss": 0.5333, "mean_token_accuracy": 0.8337358832359314, "num_tokens": 41027377.0, "step": 1081 }, { "epoch": 0.13764152143493194, "ewc_loss": 0.01611328125, "ewc_loss_diag": 2.276897430419922e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 1.5172842741012573, "learning_rate": 4.582450190758796e-06, "loss": 0.537, "mean_token_accuracy": 0.8311878442764282, "num_tokens": 41068816.0, "step": 1082 }, { "epoch": 0.13776873171352244, "ewc_loss": 0.016357421875, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 1.4946209192276, "learning_rate": 4.586689275116575e-06, "loss": 0.4577, "mean_token_accuracy": 0.8547859191894531, "num_tokens": 41106603.0, "step": 1083 }, { "epoch": 0.13789594199211297, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 1.5114326477050781, "learning_rate": 4.590928359474353e-06, "loss": 0.5449, "mean_token_accuracy": 0.82912278175354, "num_tokens": 41142862.0, "step": 1084 }, { "epoch": 0.13802315227070347, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 2.288818359375e-05, "ewc_loss_parallel": 1.3947486877441406e-05, "grad_norm": 1.4971270561218262, "learning_rate": 4.595167443832132e-06, "loss": 0.4962, "mean_token_accuracy": 0.8447563052177429, "num_tokens": 41182761.0, "step": 1085 }, { "epoch": 0.13815036254929397, "ewc_loss": 0.01611328125, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 1.3828277587890625e-05, "grad_norm": 1.5047303438186646, "learning_rate": 4.599406528189911e-06, "loss": 0.4467, "mean_token_accuracy": 0.8587777018547058, "num_tokens": 41221318.0, "step": 1086 }, { "epoch": 0.1382775728278845, "ewc_loss": 0.016357421875, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 1.5011301040649414, "learning_rate": 4.60364561254769e-06, "loss": 0.4861, "mean_token_accuracy": 0.8427484035491943, "num_tokens": 41257748.0, "step": 1087 }, { "epoch": 0.138404783106475, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 1.4997055530548096, "learning_rate": 4.607884696905469e-06, "loss": 0.4902, "mean_token_accuracy": 0.8444265723228455, "num_tokens": 41299287.0, "step": 1088 }, { "epoch": 0.1385319933850655, "ewc_loss": 0.016357421875, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 1.5039730072021484, "learning_rate": 4.612123781263248e-06, "loss": 0.5315, "mean_token_accuracy": 0.831812858581543, "num_tokens": 41344600.0, "step": 1089 }, { "epoch": 0.13865920366365603, "ewc_loss": 0.016357421875, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 1.4066696166992188e-05, "grad_norm": 1.52652108669281, "learning_rate": 4.616362865621027e-06, "loss": 0.4241, "mean_token_accuracy": 0.8617154359817505, "num_tokens": 41380681.0, "step": 1090 }, { "epoch": 0.13878641394224653, "ewc_loss": 0.016357421875, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 1.5151070356369019, "learning_rate": 4.6206019499788055e-06, "loss": 0.5193, "mean_token_accuracy": 0.8395202159881592, "num_tokens": 41418751.0, "step": 1091 }, { "epoch": 0.13891362422083706, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 1.5308558940887451, "learning_rate": 4.6248410343365835e-06, "loss": 0.5661, "mean_token_accuracy": 0.8204004764556885, "num_tokens": 41455617.0, "step": 1092 }, { "epoch": 0.13904083449942756, "ewc_loss": 0.016357421875, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 1.49245285987854, "learning_rate": 4.629080118694362e-06, "loss": 0.4773, "mean_token_accuracy": 0.8496794700622559, "num_tokens": 41494480.0, "step": 1093 }, { "epoch": 0.13916804477801806, "ewc_loss": 0.0162353515625, "ewc_loss_diag": 2.300739288330078e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 1.5195385217666626, "learning_rate": 4.633319203052141e-06, "loss": 0.4924, "mean_token_accuracy": 0.8467170000076294, "num_tokens": 41529294.0, "step": 1094 }, { "epoch": 0.13929525505660859, "ewc_loss": 0.0166015625, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 1.4883111715316772, "learning_rate": 4.63755828740992e-06, "loss": 0.469, "mean_token_accuracy": 0.850963830947876, "num_tokens": 41570787.0, "step": 1095 }, { "epoch": 0.1394224653351991, "ewc_loss": 0.016357421875, "ewc_loss_diag": 2.3245811462402344e-05, "ewc_loss_parallel": 1.4007091522216797e-05, "grad_norm": 1.5188032388687134, "learning_rate": 4.641797371767698e-06, "loss": 0.4822, "mean_token_accuracy": 0.8503375053405762, "num_tokens": 41606596.0, "step": 1096 }, { "epoch": 0.1395496756137896, "ewc_loss": 0.0166015625, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 1.503671646118164, "learning_rate": 4.646036456125477e-06, "loss": 0.5036, "mean_token_accuracy": 0.8386135697364807, "num_tokens": 41645317.0, "step": 1097 }, { "epoch": 0.13967688589238011, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 1.5231199264526367, "learning_rate": 4.650275540483256e-06, "loss": 0.4702, "mean_token_accuracy": 0.8473799228668213, "num_tokens": 41680424.0, "step": 1098 }, { "epoch": 0.13980409617097062, "ewc_loss": 0.0166015625, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.430511474609375e-05, "grad_norm": 1.5358458757400513, "learning_rate": 4.654514624841035e-06, "loss": 0.4971, "mean_token_accuracy": 0.8438484072685242, "num_tokens": 41712492.0, "step": 1099 }, { "epoch": 0.13993130644956112, "ewc_loss": 0.0166015625, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 1.528000831604004, "learning_rate": 4.658753709198813e-06, "loss": 0.4991, "mean_token_accuracy": 0.8438798785209656, "num_tokens": 41747666.0, "step": 1100 }, { "epoch": 0.14005851672815164, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 2.3365020751953125e-05, "ewc_loss_parallel": 1.4185905456542969e-05, "grad_norm": 1.5160092115402222, "learning_rate": 4.662992793556592e-06, "loss": 0.4791, "mean_token_accuracy": 0.8487036228179932, "num_tokens": 41784638.0, "step": 1101 }, { "epoch": 0.14018572700674214, "ewc_loss": 0.0166015625, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 1.509200930595398, "learning_rate": 4.667231877914371e-06, "loss": 0.5657, "mean_token_accuracy": 0.8230956792831421, "num_tokens": 41826547.0, "step": 1102 }, { "epoch": 0.14031293728533265, "ewc_loss": 0.0166015625, "ewc_loss_diag": 2.3603439331054688e-05, "ewc_loss_parallel": 1.424551010131836e-05, "grad_norm": 1.5373340845108032, "learning_rate": 4.6714709622721496e-06, "loss": 0.5126, "mean_token_accuracy": 0.8374274373054504, "num_tokens": 41859526.0, "step": 1103 }, { "epoch": 0.14044014756392317, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 1.5044770240783691, "learning_rate": 4.6757100466299284e-06, "loss": 0.5144, "mean_token_accuracy": 0.8393574953079224, "num_tokens": 41901177.0, "step": 1104 }, { "epoch": 0.14056735784251367, "ewc_loss": 0.0164794921875, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 1.4126300811767578e-05, "grad_norm": 1.5159900188446045, "learning_rate": 4.679949130987707e-06, "loss": 0.5254, "mean_token_accuracy": 0.8323707580566406, "num_tokens": 41941579.0, "step": 1105 }, { "epoch": 0.14069456812110417, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 1.5325846672058105, "learning_rate": 4.684188215345486e-06, "loss": 0.4556, "mean_token_accuracy": 0.8546829223632812, "num_tokens": 41975315.0, "step": 1106 }, { "epoch": 0.1408217783996947, "ewc_loss": 0.016845703125, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 1.5153590440750122, "learning_rate": 4.688427299703264e-06, "loss": 0.4461, "mean_token_accuracy": 0.853481650352478, "num_tokens": 42011950.0, "step": 1107 }, { "epoch": 0.1409489886782852, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 1.5166282653808594, "learning_rate": 4.692666384061043e-06, "loss": 0.5085, "mean_token_accuracy": 0.8393697738647461, "num_tokens": 42051764.0, "step": 1108 }, { "epoch": 0.1410761989568757, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 1.5098214149475098, "learning_rate": 4.696905468418822e-06, "loss": 0.523, "mean_token_accuracy": 0.8324705362319946, "num_tokens": 42094218.0, "step": 1109 }, { "epoch": 0.14120340923546623, "ewc_loss": 0.0167236328125, "ewc_loss_diag": 2.372264862060547e-05, "ewc_loss_parallel": 1.436471939086914e-05, "grad_norm": 1.5389727354049683, "learning_rate": 4.701144552776601e-06, "loss": 0.4982, "mean_token_accuracy": 0.8424153923988342, "num_tokens": 42129120.0, "step": 1110 }, { "epoch": 0.14133061951405673, "ewc_loss": 0.016845703125, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 1.5219440460205078, "learning_rate": 4.705383637134379e-06, "loss": 0.515, "mean_token_accuracy": 0.8379234075546265, "num_tokens": 42167437.0, "step": 1111 }, { "epoch": 0.14145782979264723, "ewc_loss": 0.016845703125, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 1.530646800994873, "learning_rate": 4.709622721492158e-06, "loss": 0.4864, "mean_token_accuracy": 0.8481684923171997, "num_tokens": 42203143.0, "step": 1112 }, { "epoch": 0.14158504007123776, "ewc_loss": 0.016845703125, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 1.535536289215088, "learning_rate": 4.713861805849937e-06, "loss": 0.4833, "mean_token_accuracy": 0.8473042249679565, "num_tokens": 42237024.0, "step": 1113 }, { "epoch": 0.14171225034982826, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 1.5461832284927368, "learning_rate": 4.718100890207716e-06, "loss": 0.5596, "mean_token_accuracy": 0.8253930807113647, "num_tokens": 42272847.0, "step": 1114 }, { "epoch": 0.1418394606284188, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 2.396106719970703e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 1.5467917919158936, "learning_rate": 4.7223399745654945e-06, "loss": 0.5139, "mean_token_accuracy": 0.8368402123451233, "num_tokens": 42309335.0, "step": 1115 }, { "epoch": 0.1419666709070093, "ewc_loss": 0.016845703125, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.4424324035644531e-05, "grad_norm": 1.5333136320114136, "learning_rate": 4.726579058923273e-06, "loss": 0.5091, "mean_token_accuracy": 0.842237651348114, "num_tokens": 42346349.0, "step": 1116 }, { "epoch": 0.1420938811855998, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 1.5291330814361572, "learning_rate": 4.730818143281051e-06, "loss": 0.5372, "mean_token_accuracy": 0.8331693410873413, "num_tokens": 42389310.0, "step": 1117 }, { "epoch": 0.14222109146419032, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 1.5263872146606445, "learning_rate": 4.73505722763883e-06, "loss": 0.4767, "mean_token_accuracy": 0.849065899848938, "num_tokens": 42427808.0, "step": 1118 }, { "epoch": 0.14234830174278082, "ewc_loss": 0.016845703125, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.4483928680419922e-05, "grad_norm": 1.537113070487976, "learning_rate": 4.739296311996609e-06, "loss": 0.5007, "mean_token_accuracy": 0.8407637476921082, "num_tokens": 42465531.0, "step": 1119 }, { "epoch": 0.14247551202137132, "ewc_loss": 0.01708984375, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 1.5177973508834839, "learning_rate": 4.743535396354388e-06, "loss": 0.4955, "mean_token_accuracy": 0.8466349840164185, "num_tokens": 42507613.0, "step": 1120 }, { "epoch": 0.14260272229996185, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 2.4080276489257812e-05, "ewc_loss_parallel": 1.4543533325195312e-05, "grad_norm": 1.551796555519104, "learning_rate": 4.747774480712167e-06, "loss": 0.5192, "mean_token_accuracy": 0.8394530415534973, "num_tokens": 42542876.0, "step": 1121 }, { "epoch": 0.14272993257855235, "ewc_loss": 0.01708984375, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 1.5504482984542847, "learning_rate": 4.752013565069945e-06, "loss": 0.5358, "mean_token_accuracy": 0.8321757316589355, "num_tokens": 42577587.0, "step": 1122 }, { "epoch": 0.14285714285714285, "ewc_loss": 0.0169677734375, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.4603137969970703e-05, "grad_norm": 1.5425405502319336, "learning_rate": 4.756252649427724e-06, "loss": 0.5041, "mean_token_accuracy": 0.8410888314247131, "num_tokens": 42612733.0, "step": 1123 }, { "epoch": 0.14298435313573338, "ewc_loss": 0.01708984375, "ewc_loss_diag": 2.4199485778808594e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 1.535237193107605, "learning_rate": 4.760491733785503e-06, "loss": 0.449, "mean_token_accuracy": 0.8568610548973083, "num_tokens": 42646831.0, "step": 1124 }, { "epoch": 0.14311156341432388, "ewc_loss": 0.01708984375, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 1.563309907913208, "learning_rate": 4.764730818143282e-06, "loss": 0.4984, "mean_token_accuracy": 0.8362541198730469, "num_tokens": 42681351.0, "step": 1125 }, { "epoch": 0.14323877369291438, "ewc_loss": 0.01708984375, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 1.5261722803115845, "learning_rate": 4.76896990250106e-06, "loss": 0.5489, "mean_token_accuracy": 0.8273475170135498, "num_tokens": 42723491.0, "step": 1126 }, { "epoch": 0.1433659839715049, "ewc_loss": 0.01708984375, "ewc_loss_diag": 2.4318695068359375e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 1.534261703491211, "learning_rate": 4.7732089868588386e-06, "loss": 0.5088, "mean_token_accuracy": 0.8387540578842163, "num_tokens": 42763349.0, "step": 1127 }, { "epoch": 0.1434931942500954, "ewc_loss": 0.01708984375, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 1.531523585319519, "learning_rate": 4.7774480712166174e-06, "loss": 0.5219, "mean_token_accuracy": 0.8347687721252441, "num_tokens": 42803977.0, "step": 1128 }, { "epoch": 0.1436204045286859, "ewc_loss": 0.01708984375, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 1.5241422653198242, "learning_rate": 4.781687155574396e-06, "loss": 0.5245, "mean_token_accuracy": 0.833876371383667, "num_tokens": 42847313.0, "step": 1129 }, { "epoch": 0.14374761480727644, "ewc_loss": 0.01708984375, "ewc_loss_diag": 2.4437904357910156e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 1.5232576131820679, "learning_rate": 4.785926239932175e-06, "loss": 0.4715, "mean_token_accuracy": 0.8494797945022583, "num_tokens": 42887468.0, "step": 1130 }, { "epoch": 0.14387482508586694, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 1.5412436723709106, "learning_rate": 4.790165324289954e-06, "loss": 0.5003, "mean_token_accuracy": 0.843720018863678, "num_tokens": 42925502.0, "step": 1131 }, { "epoch": 0.14400203536445744, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 2.4557113647460938e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 1.5504941940307617, "learning_rate": 4.794404408647733e-06, "loss": 0.5022, "mean_token_accuracy": 0.8440049290657043, "num_tokens": 42961930.0, "step": 1132 }, { "epoch": 0.14412924564304797, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 1.5182737112045288, "learning_rate": 4.798643493005512e-06, "loss": 0.4481, "mean_token_accuracy": 0.8572700619697571, "num_tokens": 43002121.0, "step": 1133 }, { "epoch": 0.14425645592163847, "ewc_loss": 0.01708984375, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 1.562644600868225, "learning_rate": 4.80288257736329e-06, "loss": 0.5459, "mean_token_accuracy": 0.8302308320999146, "num_tokens": 43039343.0, "step": 1134 }, { "epoch": 0.14438366620022897, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 1.5364516973495483, "learning_rate": 4.807121661721069e-06, "loss": 0.4482, "mean_token_accuracy": 0.8585540056228638, "num_tokens": 43075035.0, "step": 1135 }, { "epoch": 0.1445108764788195, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 1.5361956357955933, "learning_rate": 4.811360746078848e-06, "loss": 0.5017, "mean_token_accuracy": 0.8439870476722717, "num_tokens": 43114350.0, "step": 1136 }, { "epoch": 0.14463808675741, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 1.5745015144348145, "learning_rate": 4.815599830436626e-06, "loss": 0.5007, "mean_token_accuracy": 0.8419033288955688, "num_tokens": 43146445.0, "step": 1137 }, { "epoch": 0.1447652970360005, "ewc_loss": 0.01708984375, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 1.5432407855987549, "learning_rate": 4.819838914794405e-06, "loss": 0.4851, "mean_token_accuracy": 0.8471301794052124, "num_tokens": 43182216.0, "step": 1138 }, { "epoch": 0.14489250731459102, "ewc_loss": 0.01708984375, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.4662742614746094e-05, "grad_norm": 1.55614173412323, "learning_rate": 4.8240779991521835e-06, "loss": 0.4929, "mean_token_accuracy": 0.842573344707489, "num_tokens": 43216014.0, "step": 1139 }, { "epoch": 0.14501971759318152, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 2.467632293701172e-05, "ewc_loss_parallel": 1.4722347259521484e-05, "grad_norm": 1.5367995500564575, "learning_rate": 4.828317083509962e-06, "loss": 0.4999, "mean_token_accuracy": 0.8416165709495544, "num_tokens": 43252909.0, "step": 1140 }, { "epoch": 0.14514692787177205, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 1.5540990829467773, "learning_rate": 4.83255616786774e-06, "loss": 0.4776, "mean_token_accuracy": 0.8522166609764099, "num_tokens": 43285653.0, "step": 1141 }, { "epoch": 0.14527413815036255, "ewc_loss": 0.017333984375, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 1.5459381341934204, "learning_rate": 4.836795252225519e-06, "loss": 0.4741, "mean_token_accuracy": 0.8526079654693604, "num_tokens": 43322087.0, "step": 1142 }, { "epoch": 0.14540134842895305, "ewc_loss": 0.017333984375, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 1.5588980913162231, "learning_rate": 4.841034336583298e-06, "loss": 0.4853, "mean_token_accuracy": 0.8494484424591064, "num_tokens": 43360526.0, "step": 1143 }, { "epoch": 0.14552855870754358, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 1.531551480293274, "learning_rate": 4.845273420941077e-06, "loss": 0.4644, "mean_token_accuracy": 0.8548346161842346, "num_tokens": 43400267.0, "step": 1144 }, { "epoch": 0.14565576898613408, "ewc_loss": 0.017333984375, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 1.5904616117477417, "learning_rate": 4.849512505298856e-06, "loss": 0.5038, "mean_token_accuracy": 0.8377398252487183, "num_tokens": 43430411.0, "step": 1145 }, { "epoch": 0.14578297926472458, "ewc_loss": 0.017333984375, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 1.5358388423919678, "learning_rate": 4.853751589656635e-06, "loss": 0.4891, "mean_token_accuracy": 0.847194492816925, "num_tokens": 43472199.0, "step": 1146 }, { "epoch": 0.1459101895433151, "ewc_loss": 0.0172119140625, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.4781951904296875e-05, "grad_norm": 1.5437345504760742, "learning_rate": 4.857990674014414e-06, "loss": 0.5125, "mean_token_accuracy": 0.8384243249893188, "num_tokens": 43515896.0, "step": 1147 }, { "epoch": 0.1460373998219056, "ewc_loss": 0.017333984375, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 1.5453801155090332, "learning_rate": 4.862229758372193e-06, "loss": 0.5339, "mean_token_accuracy": 0.8292283415794373, "num_tokens": 43556088.0, "step": 1148 }, { "epoch": 0.1461646101004961, "ewc_loss": 0.017333984375, "ewc_loss_diag": 2.47955322265625e-05, "ewc_loss_parallel": 1.4841556549072266e-05, "grad_norm": 1.5537495613098145, "learning_rate": 4.866468842729971e-06, "loss": 0.5387, "mean_token_accuracy": 0.8288581371307373, "num_tokens": 43596623.0, "step": 1149 }, { "epoch": 0.14629182037908664, "ewc_loss": 0.017333984375, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 1.5708235502243042, "learning_rate": 4.8707079270877495e-06, "loss": 0.5114, "mean_token_accuracy": 0.8397350311279297, "num_tokens": 43634757.0, "step": 1150 }, { "epoch": 0.14641903065767714, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 1.5547924041748047, "learning_rate": 4.874947011445528e-06, "loss": 0.5174, "mean_token_accuracy": 0.8383362889289856, "num_tokens": 43671979.0, "step": 1151 }, { "epoch": 0.14654624093626764, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 1.5317283868789673, "learning_rate": 4.879186095803307e-06, "loss": 0.4417, "mean_token_accuracy": 0.86025470495224, "num_tokens": 43711898.0, "step": 1152 }, { "epoch": 0.14667345121485817, "ewc_loss": 0.017333984375, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.4901161193847656e-05, "grad_norm": 1.5676348209381104, "learning_rate": 4.883425180161085e-06, "loss": 0.4731, "mean_token_accuracy": 0.8510869741439819, "num_tokens": 43748394.0, "step": 1153 }, { "epoch": 0.14680066149344867, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 1.5645968914031982, "learning_rate": 4.887664264518864e-06, "loss": 0.5237, "mean_token_accuracy": 0.8417947292327881, "num_tokens": 43784730.0, "step": 1154 }, { "epoch": 0.14692787177203917, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 2.491474151611328e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 1.56792151927948, "learning_rate": 4.891903348876643e-06, "loss": 0.5134, "mean_token_accuracy": 0.842976987361908, "num_tokens": 43823993.0, "step": 1155 }, { "epoch": 0.1470550820506297, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 1.576785683631897, "learning_rate": 4.896142433234421e-06, "loss": 0.5476, "mean_token_accuracy": 0.8343408107757568, "num_tokens": 43861673.0, "step": 1156 }, { "epoch": 0.1471822923292202, "ewc_loss": 0.0174560546875, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.4960765838623047e-05, "grad_norm": 1.569378137588501, "learning_rate": 4.9003815175922e-06, "loss": 0.5431, "mean_token_accuracy": 0.8346606492996216, "num_tokens": 43900123.0, "step": 1157 }, { "epoch": 0.1473095026078107, "ewc_loss": 0.017578125, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 1.5589712858200073, "learning_rate": 4.904620601949979e-06, "loss": 0.4986, "mean_token_accuracy": 0.8426511883735657, "num_tokens": 43939371.0, "step": 1158 }, { "epoch": 0.14743671288640123, "ewc_loss": 0.017578125, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 1.58677339553833, "learning_rate": 4.908859686307758e-06, "loss": 0.532, "mean_token_accuracy": 0.8354775905609131, "num_tokens": 43972960.0, "step": 1159 }, { "epoch": 0.14756392316499173, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 1.5724389553070068, "learning_rate": 4.913098770665537e-06, "loss": 0.5356, "mean_token_accuracy": 0.8361324071884155, "num_tokens": 44009919.0, "step": 1160 }, { "epoch": 0.14769113344358223, "ewc_loss": 0.017578125, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.5020370483398438e-05, "grad_norm": 1.5629726648330688, "learning_rate": 4.9173378550233155e-06, "loss": 0.4825, "mean_token_accuracy": 0.8440468907356262, "num_tokens": 44050491.0, "step": 1161 }, { "epoch": 0.14781834372217276, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 1.5613692998886108, "learning_rate": 4.9215769393810944e-06, "loss": 0.501, "mean_token_accuracy": 0.846783459186554, "num_tokens": 44087999.0, "step": 1162 }, { "epoch": 0.14794555400076326, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 1.5709339380264282, "learning_rate": 4.925816023738873e-06, "loss": 0.5512, "mean_token_accuracy": 0.8291313648223877, "num_tokens": 44126014.0, "step": 1163 }, { "epoch": 0.14807276427935376, "ewc_loss": 0.017822265625, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.52587890625e-05, "grad_norm": 1.5967087745666504, "learning_rate": 4.930055108096651e-06, "loss": 0.4401, "mean_token_accuracy": 0.8609241843223572, "num_tokens": 44154713.0, "step": 1164 }, { "epoch": 0.1481999745579443, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 1.5972682237625122, "learning_rate": 4.93429419245443e-06, "loss": 0.5657, "mean_token_accuracy": 0.8245494961738586, "num_tokens": 44189624.0, "step": 1165 }, { "epoch": 0.1483271848365348, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 2.5153160095214844e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 1.5743831396102905, "learning_rate": 4.938533276812209e-06, "loss": 0.5069, "mean_token_accuracy": 0.8411064147949219, "num_tokens": 44225010.0, "step": 1166 }, { "epoch": 0.14845439511512531, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.5139579772949219e-05, "grad_norm": 1.555474877357483, "learning_rate": 4.942772361169988e-06, "loss": 0.4205, "mean_token_accuracy": 0.8620691299438477, "num_tokens": 44264987.0, "step": 1167 }, { "epoch": 0.14858160539371582, "ewc_loss": 0.0177001953125, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.519918441772461e-05, "grad_norm": 1.5544525384902954, "learning_rate": 4.947011445527766e-06, "loss": 0.4962, "mean_token_accuracy": 0.8457098007202148, "num_tokens": 44306705.0, "step": 1168 }, { "epoch": 0.14870881567230632, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 1.5505709648132324, "learning_rate": 4.951250529885545e-06, "loss": 0.4526, "mean_token_accuracy": 0.8574248552322388, "num_tokens": 44347227.0, "step": 1169 }, { "epoch": 0.14883602595089684, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 2.5391578674316406e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 1.6997978687286377, "learning_rate": 4.955489614243324e-06, "loss": 0.5875, "mean_token_accuracy": 0.8132768869400024, "num_tokens": 44385273.0, "step": 1170 }, { "epoch": 0.14896323622948734, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.5592219829559326, "learning_rate": 4.959728698601103e-06, "loss": 0.5067, "mean_token_accuracy": 0.8372440338134766, "num_tokens": 44426177.0, "step": 1171 }, { "epoch": 0.14909044650807785, "ewc_loss": 0.01806640625, "ewc_loss_diag": 2.5510787963867188e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 1.59145987033844, "learning_rate": 4.963967782958881e-06, "loss": 0.4828, "mean_token_accuracy": 0.8476411700248718, "num_tokens": 44468609.0, "step": 1172 }, { "epoch": 0.14921765678666837, "ewc_loss": 0.0179443359375, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 1.5595320463180542, "learning_rate": 4.96820686731666e-06, "loss": 0.4826, "mean_token_accuracy": 0.8458737730979919, "num_tokens": 44507212.0, "step": 1173 }, { "epoch": 0.14934486706525887, "ewc_loss": 0.018310546875, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.6036568880081177, "learning_rate": 4.9724459516744385e-06, "loss": 0.4671, "mean_token_accuracy": 0.8511021137237549, "num_tokens": 44538551.0, "step": 1174 }, { "epoch": 0.14947207734384937, "ewc_loss": 0.01806640625, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 1.5860623121261597, "learning_rate": 4.976685036032217e-06, "loss": 0.558, "mean_token_accuracy": 0.8240644931793213, "num_tokens": 44577105.0, "step": 1175 }, { "epoch": 0.1495992876224399, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.5623712539672852, "learning_rate": 4.980924120389996e-06, "loss": 0.4368, "mean_token_accuracy": 0.8618285655975342, "num_tokens": 44615548.0, "step": 1176 }, { "epoch": 0.1497264979010304, "ewc_loss": 0.01806640625, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 1.563297152519226, "learning_rate": 4.985163204747775e-06, "loss": 0.4441, "mean_token_accuracy": 0.8594728708267212, "num_tokens": 44654447.0, "step": 1177 }, { "epoch": 0.1498537081796209, "ewc_loss": 0.018310546875, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.56571626663208, "learning_rate": 4.989402289105554e-06, "loss": 0.498, "mean_token_accuracy": 0.8468854427337646, "num_tokens": 44697995.0, "step": 1178 }, { "epoch": 0.14998091845821143, "ewc_loss": 0.018310546875, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.600645899772644, "learning_rate": 4.993641373463332e-06, "loss": 0.5585, "mean_token_accuracy": 0.8247949481010437, "num_tokens": 44734376.0, "step": 1179 }, { "epoch": 0.15010812873680193, "ewc_loss": 0.018310546875, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.601150393486023, "learning_rate": 4.997880457821111e-06, "loss": 0.5067, "mean_token_accuracy": 0.8395043015480042, "num_tokens": 44772745.0, "step": 1180 }, { "epoch": 0.15023533901539243, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.5529991388320923, "learning_rate": 5.002119542178889e-06, "loss": 0.495, "mean_token_accuracy": 0.845145583152771, "num_tokens": 44816530.0, "step": 1181 }, { "epoch": 0.15036254929398296, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.5985231399536133, "learning_rate": 5.006358626536668e-06, "loss": 0.4556, "mean_token_accuracy": 0.8543047308921814, "num_tokens": 44853428.0, "step": 1182 }, { "epoch": 0.15048975957257346, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.562999725341797e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.5643038749694824, "learning_rate": 5.010597710894447e-06, "loss": 0.4771, "mean_token_accuracy": 0.8479836583137512, "num_tokens": 44891118.0, "step": 1183 }, { "epoch": 0.15061696985116396, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.586841583251953e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.58515465259552, "learning_rate": 5.014836795252226e-06, "loss": 0.4675, "mean_token_accuracy": 0.8514009118080139, "num_tokens": 44929143.0, "step": 1184 }, { "epoch": 0.1507441801297545, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.586841583251953e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.5837565660476685, "learning_rate": 5.0190758796100045e-06, "loss": 0.4733, "mean_token_accuracy": 0.8525803685188293, "num_tokens": 44966361.0, "step": 1185 }, { "epoch": 0.150871390408345, "ewc_loss": 0.018310546875, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.5726083517074585, "learning_rate": 5.0233149639677834e-06, "loss": 0.5094, "mean_token_accuracy": 0.8376481533050537, "num_tokens": 45007993.0, "step": 1186 }, { "epoch": 0.1509986006869355, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.6145623922348022, "learning_rate": 5.027554048325562e-06, "loss": 0.546, "mean_token_accuracy": 0.8273730874061584, "num_tokens": 45040337.0, "step": 1187 }, { "epoch": 0.15112581096552602, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.5832456350326538, "learning_rate": 5.031793132683341e-06, "loss": 0.5553, "mean_token_accuracy": 0.8260395526885986, "num_tokens": 45080730.0, "step": 1188 }, { "epoch": 0.15125302124411652, "ewc_loss": 0.018310546875, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.6012543439865112, "learning_rate": 5.036032217041119e-06, "loss": 0.5145, "mean_token_accuracy": 0.8356505632400513, "num_tokens": 45116748.0, "step": 1189 }, { "epoch": 0.15138023152270705, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.5987625122070312e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.5686390399932861, "learning_rate": 5.040271301398898e-06, "loss": 0.5044, "mean_token_accuracy": 0.8369455337524414, "num_tokens": 45158596.0, "step": 1190 }, { "epoch": 0.15150744180129755, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.5686781406402588, "learning_rate": 5.044510385756677e-06, "loss": 0.4792, "mean_token_accuracy": 0.8461698889732361, "num_tokens": 45202433.0, "step": 1191 }, { "epoch": 0.15163465207988805, "ewc_loss": 0.018310546875, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.68284010887146, "learning_rate": 5.048749470114456e-06, "loss": 0.4884, "mean_token_accuracy": 0.8475385904312134, "num_tokens": 45239202.0, "step": 1192 }, { "epoch": 0.15176186235847858, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.6106834411621094e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.620033621788025, "learning_rate": 5.052988554472235e-06, "loss": 0.5046, "mean_token_accuracy": 0.8406984210014343, "num_tokens": 45270150.0, "step": 1193 }, { "epoch": 0.15188907263706908, "ewc_loss": 0.0181884765625, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.5497207641601562e-05, "grad_norm": 1.5848249197006226, "learning_rate": 5.057227638830014e-06, "loss": 0.5033, "mean_token_accuracy": 0.8431003093719482, "num_tokens": 45306394.0, "step": 1194 }, { "epoch": 0.15201628291565958, "ewc_loss": 0.01806640625, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.537799835205078e-05, "grad_norm": 1.5862537622451782, "learning_rate": 5.0614667231877925e-06, "loss": 0.5204, "mean_token_accuracy": 0.8377147912979126, "num_tokens": 45344945.0, "step": 1195 }, { "epoch": 0.1521434931942501, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.60270094871521, "learning_rate": 5.06570580754557e-06, "loss": 0.4734, "mean_token_accuracy": 0.8526666164398193, "num_tokens": 45378179.0, "step": 1196 }, { "epoch": 0.1522707034728406, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.5644540786743164, "learning_rate": 5.069944891903349e-06, "loss": 0.4886, "mean_token_accuracy": 0.8441506624221802, "num_tokens": 45421194.0, "step": 1197 }, { "epoch": 0.1523979137514311, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.5969300270080566, "learning_rate": 5.0741839762611275e-06, "loss": 0.4991, "mean_token_accuracy": 0.8410332202911377, "num_tokens": 45458226.0, "step": 1198 }, { "epoch": 0.15252512403002164, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.5791313648223877, "learning_rate": 5.078423060618906e-06, "loss": 0.5007, "mean_token_accuracy": 0.8427417874336243, "num_tokens": 45500007.0, "step": 1199 }, { "epoch": 0.15265233430861214, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.5999499559402466, "learning_rate": 5.082662144976685e-06, "loss": 0.4933, "mean_token_accuracy": 0.8400799036026001, "num_tokens": 45536605.0, "step": 1200 }, { "epoch": 0.15277954458720264, "ewc_loss": 0.018310546875, "ewc_loss_diag": 2.6345252990722656e-05, "ewc_loss_parallel": 1.5616416931152344e-05, "grad_norm": 1.590977668762207, "learning_rate": 5.086901229334464e-06, "loss": 0.4892, "mean_token_accuracy": 0.8494160175323486, "num_tokens": 45572137.0, "step": 1201 }, { "epoch": 0.15290675486579317, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.5932248830795288, "learning_rate": 5.091140313692243e-06, "loss": 0.4945, "mean_token_accuracy": 0.8428972959518433, "num_tokens": 45608659.0, "step": 1202 }, { "epoch": 0.15303396514438367, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.5934314727783203, "learning_rate": 5.095379398050022e-06, "loss": 0.4805, "mean_token_accuracy": 0.8471852540969849, "num_tokens": 45647002.0, "step": 1203 }, { "epoch": 0.15316117542297417, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.6196540594100952, "learning_rate": 5.0996184824078e-06, "loss": 0.5068, "mean_token_accuracy": 0.8377121090888977, "num_tokens": 45680194.0, "step": 1204 }, { "epoch": 0.1532883857015647, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.6464462280273438e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.5821001529693604, "learning_rate": 5.103857566765579e-06, "loss": 0.4794, "mean_token_accuracy": 0.8482893705368042, "num_tokens": 45719629.0, "step": 1205 }, { "epoch": 0.1534155959801552, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.5975720882415771, "learning_rate": 5.108096651123358e-06, "loss": 0.4493, "mean_token_accuracy": 0.8586728572845459, "num_tokens": 45754630.0, "step": 1206 }, { "epoch": 0.1535428062587457, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.658367156982422e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.587974190711975, "learning_rate": 5.112335735481137e-06, "loss": 0.4854, "mean_token_accuracy": 0.8447222709655762, "num_tokens": 45791904.0, "step": 1207 }, { "epoch": 0.15367001653733622, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.5778554677963257, "learning_rate": 5.1165748198389155e-06, "loss": 0.4048, "mean_token_accuracy": 0.8734982013702393, "num_tokens": 45830553.0, "step": 1208 }, { "epoch": 0.15379722681592672, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.6087393760681152, "learning_rate": 5.120813904196694e-06, "loss": 0.4797, "mean_token_accuracy": 0.8505428433418274, "num_tokens": 45865957.0, "step": 1209 }, { "epoch": 0.15392443709451722, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6225160360336304, "learning_rate": 5.125052988554473e-06, "loss": 0.5399, "mean_token_accuracy": 0.831747829914093, "num_tokens": 45898993.0, "step": 1210 }, { "epoch": 0.15405164737310775, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6059824228286743, "learning_rate": 5.1292920729122505e-06, "loss": 0.4702, "mean_token_accuracy": 0.8517925143241882, "num_tokens": 45932843.0, "step": 1211 }, { "epoch": 0.15417885765169825, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.6702880859375e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.580520510673523, "learning_rate": 5.133531157270029e-06, "loss": 0.4689, "mean_token_accuracy": 0.8547075390815735, "num_tokens": 45973247.0, "step": 1212 }, { "epoch": 0.15430606793028875, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.682209014892578e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.5969152450561523, "learning_rate": 5.137770241627808e-06, "loss": 0.4297, "mean_token_accuracy": 0.8633956909179688, "num_tokens": 46008641.0, "step": 1213 }, { "epoch": 0.15443327820887928, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.682209014892578e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.596693992614746, "learning_rate": 5.142009325985587e-06, "loss": 0.461, "mean_token_accuracy": 0.8578697443008423, "num_tokens": 46046789.0, "step": 1214 }, { "epoch": 0.15456048848746978, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.5950827598571777, "learning_rate": 5.146248410343366e-06, "loss": 0.4998, "mean_token_accuracy": 0.844120979309082, "num_tokens": 46086904.0, "step": 1215 }, { "epoch": 0.1546876987660603, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.5793946981430054, "learning_rate": 5.150487494701145e-06, "loss": 0.4509, "mean_token_accuracy": 0.8599727153778076, "num_tokens": 46129743.0, "step": 1216 }, { "epoch": 0.1548149090446508, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6118645668029785, "learning_rate": 5.154726579058924e-06, "loss": 0.4937, "mean_token_accuracy": 0.8439838886260986, "num_tokens": 46164902.0, "step": 1217 }, { "epoch": 0.1549421193232413, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.5735996961593628, "learning_rate": 5.158965663416703e-06, "loss": 0.4906, "mean_token_accuracy": 0.8475438356399536, "num_tokens": 46209689.0, "step": 1218 }, { "epoch": 0.15506932960183184, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.587083339691162, "learning_rate": 5.163204747774481e-06, "loss": 0.4414, "mean_token_accuracy": 0.8590492010116577, "num_tokens": 46251365.0, "step": 1219 }, { "epoch": 0.15519653988042234, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.5945744514465332, "learning_rate": 5.1674438321322596e-06, "loss": 0.4932, "mean_token_accuracy": 0.84273362159729, "num_tokens": 46290961.0, "step": 1220 }, { "epoch": 0.15532375015901284, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6084667444229126, "learning_rate": 5.1716829164900385e-06, "loss": 0.4711, "mean_token_accuracy": 0.8521993160247803, "num_tokens": 46326317.0, "step": 1221 }, { "epoch": 0.15545096043760337, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6018089056015015, "learning_rate": 5.175922000847817e-06, "loss": 0.4409, "mean_token_accuracy": 0.8600209951400757, "num_tokens": 46361018.0, "step": 1222 }, { "epoch": 0.15557817071619387, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.6068077087402344, "learning_rate": 5.180161085205596e-06, "loss": 0.4711, "mean_token_accuracy": 0.8492047190666199, "num_tokens": 46397344.0, "step": 1223 }, { "epoch": 0.15570538099478437, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6043193340301514, "learning_rate": 5.184400169563375e-06, "loss": 0.5309, "mean_token_accuracy": 0.8339524865150452, "num_tokens": 46437257.0, "step": 1224 }, { "epoch": 0.1558325912733749, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6028169393539429, "learning_rate": 5.188639253921154e-06, "loss": 0.5058, "mean_token_accuracy": 0.8410242795944214, "num_tokens": 46477844.0, "step": 1225 }, { "epoch": 0.1559598015519654, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6012753248214722, "learning_rate": 5.192878338278933e-06, "loss": 0.4703, "mean_token_accuracy": 0.8504382967948914, "num_tokens": 46514556.0, "step": 1226 }, { "epoch": 0.1560870118305559, "ewc_loss": 0.0184326171875, "ewc_loss_diag": 2.7060508728027344e-05, "ewc_loss_parallel": 1.5735626220703125e-05, "grad_norm": 1.5897654294967651, "learning_rate": 5.19711742263671e-06, "loss": 0.5446, "mean_token_accuracy": 0.830008864402771, "num_tokens": 46559274.0, "step": 1227 }, { "epoch": 0.15621422210914643, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6297446489334106, "learning_rate": 5.201356506994489e-06, "loss": 0.4745, "mean_token_accuracy": 0.8466691374778748, "num_tokens": 46591104.0, "step": 1228 }, { "epoch": 0.15634143238773693, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6065292358398438, "learning_rate": 5.205595591352268e-06, "loss": 0.5629, "mean_token_accuracy": 0.8226664662361145, "num_tokens": 46631296.0, "step": 1229 }, { "epoch": 0.15646864266632743, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.5998809337615967, "learning_rate": 5.209834675710047e-06, "loss": 0.5004, "mean_token_accuracy": 0.8436782360076904, "num_tokens": 46673079.0, "step": 1230 }, { "epoch": 0.15659585294491796, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6283708810806274, "learning_rate": 5.214073760067826e-06, "loss": 0.563, "mean_token_accuracy": 0.8247209787368774, "num_tokens": 46706820.0, "step": 1231 }, { "epoch": 0.15672306322350846, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6072393655776978, "learning_rate": 5.2183128444256045e-06, "loss": 0.4915, "mean_token_accuracy": 0.847425103187561, "num_tokens": 46742466.0, "step": 1232 }, { "epoch": 0.15685027350209896, "ewc_loss": 0.0185546875, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.6126770973205566, "learning_rate": 5.222551928783383e-06, "loss": 0.4698, "mean_token_accuracy": 0.8499574065208435, "num_tokens": 46778882.0, "step": 1233 }, { "epoch": 0.1569774837806895, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 2.7298927307128906e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 1.611132025718689, "learning_rate": 5.226791013141161e-06, "loss": 0.5295, "mean_token_accuracy": 0.8372312784194946, "num_tokens": 46821509.0, "step": 1234 }, { "epoch": 0.15710469405928, "ewc_loss": 0.018798828125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 1.6034321784973145, "learning_rate": 5.23103009749894e-06, "loss": 0.4759, "mean_token_accuracy": 0.8493584394454956, "num_tokens": 46862247.0, "step": 1235 }, { "epoch": 0.1572319043378705, "ewc_loss": 0.018798828125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 1.6006605625152588, "learning_rate": 5.235269181856719e-06, "loss": 0.4986, "mean_token_accuracy": 0.8440709710121155, "num_tokens": 46906645.0, "step": 1236 }, { "epoch": 0.15735911461646102, "ewc_loss": 0.018798828125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 1.6532334089279175, "learning_rate": 5.239508266214498e-06, "loss": 0.5453, "mean_token_accuracy": 0.8259559869766235, "num_tokens": 46940492.0, "step": 1237 }, { "epoch": 0.15748632489505152, "ewc_loss": 0.018798828125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 1.6289641857147217, "learning_rate": 5.243747350572277e-06, "loss": 0.5047, "mean_token_accuracy": 0.8431610465049744, "num_tokens": 46976871.0, "step": 1238 }, { "epoch": 0.15761353517364202, "ewc_loss": 0.018798828125, "ewc_loss_diag": 2.753734588623047e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 1.6235626935958862, "learning_rate": 5.247986434930056e-06, "loss": 0.4971, "mean_token_accuracy": 0.8459481000900269, "num_tokens": 47013689.0, "step": 1239 }, { "epoch": 0.15774074545223254, "ewc_loss": 0.0186767578125, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.5854835510253906e-05, "grad_norm": 1.627434492111206, "learning_rate": 5.252225519287835e-06, "loss": 0.5566, "mean_token_accuracy": 0.8255046010017395, "num_tokens": 47050995.0, "step": 1240 }, { "epoch": 0.15786795573082305, "ewc_loss": 0.018798828125, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.5974044799804688e-05, "grad_norm": 1.6105711460113525, "learning_rate": 5.256464603645614e-06, "loss": 0.4847, "mean_token_accuracy": 0.8505081534385681, "num_tokens": 47088571.0, "step": 1241 }, { "epoch": 0.15799516600941357, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 2.777576446533203e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 1.609445333480835, "learning_rate": 5.260703688003392e-06, "loss": 0.4983, "mean_token_accuracy": 0.8441494703292847, "num_tokens": 47130547.0, "step": 1242 }, { "epoch": 0.15812237628800407, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 1.6345804929733276, "learning_rate": 5.2649427723611705e-06, "loss": 0.497, "mean_token_accuracy": 0.843102216720581, "num_tokens": 47162724.0, "step": 1243 }, { "epoch": 0.15824958656659457, "ewc_loss": 0.01904296875, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 1.64100182056427, "learning_rate": 5.269181856718949e-06, "loss": 0.5386, "mean_token_accuracy": 0.8313906192779541, "num_tokens": 47200081.0, "step": 1244 }, { "epoch": 0.1583767968451851, "ewc_loss": 0.01904296875, "ewc_loss_diag": 2.7894973754882812e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 1.6301774978637695, "learning_rate": 5.2734209410767275e-06, "loss": 0.4774, "mean_token_accuracy": 0.852325439453125, "num_tokens": 47235074.0, "step": 1245 }, { "epoch": 0.1585040071237756, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 1.608119249343872, "learning_rate": 5.277660025434506e-06, "loss": 0.4368, "mean_token_accuracy": 0.8592485189437866, "num_tokens": 47274258.0, "step": 1246 }, { "epoch": 0.1586312174023661, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 1.6144357919692993, "learning_rate": 5.281899109792285e-06, "loss": 0.4629, "mean_token_accuracy": 0.853227972984314, "num_tokens": 47314232.0, "step": 1247 }, { "epoch": 0.15875842768095663, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 1.6124986410140991, "learning_rate": 5.286138194150064e-06, "loss": 0.5078, "mean_token_accuracy": 0.8418501615524292, "num_tokens": 47355245.0, "step": 1248 }, { "epoch": 0.15888563795954713, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 1.6400437355041504, "learning_rate": 5.290377278507842e-06, "loss": 0.542, "mean_token_accuracy": 0.8351743221282959, "num_tokens": 47394356.0, "step": 1249 }, { "epoch": 0.15901284823813763, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 1.6178539991378784, "learning_rate": 5.294616362865621e-06, "loss": 0.4804, "mean_token_accuracy": 0.8495687246322632, "num_tokens": 47435220.0, "step": 1250 }, { "epoch": 0.15914005851672816, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6377158164978027, "learning_rate": 5.2988554472234e-06, "loss": 0.5571, "mean_token_accuracy": 0.8338509202003479, "num_tokens": 47477294.0, "step": 1251 }, { "epoch": 0.15926726879531866, "ewc_loss": 0.01904296875, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 1.648079514503479, "learning_rate": 5.303094531581179e-06, "loss": 0.5136, "mean_token_accuracy": 0.8381355404853821, "num_tokens": 47512216.0, "step": 1252 }, { "epoch": 0.15939447907390916, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6052254438400269, "learning_rate": 5.307333615938958e-06, "loss": 0.4327, "mean_token_accuracy": 0.8652970790863037, "num_tokens": 47550075.0, "step": 1253 }, { "epoch": 0.1595216893524997, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 1.6702067852020264, "learning_rate": 5.3115727002967366e-06, "loss": 0.5178, "mean_token_accuracy": 0.8374489545822144, "num_tokens": 47581595.0, "step": 1254 }, { "epoch": 0.1596488996310902, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6145557165145874, "learning_rate": 5.3158117846545155e-06, "loss": 0.4208, "mean_token_accuracy": 0.8630187511444092, "num_tokens": 47617336.0, "step": 1255 }, { "epoch": 0.1597761099096807, "ewc_loss": 0.0189208984375, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.609325408935547e-05, "grad_norm": 1.6219587326049805, "learning_rate": 5.320050869012294e-06, "loss": 0.4952, "mean_token_accuracy": 0.8460540771484375, "num_tokens": 47662701.0, "step": 1256 }, { "epoch": 0.15990332018827122, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6187248229980469, "learning_rate": 5.324289953370072e-06, "loss": 0.5444, "mean_token_accuracy": 0.8321320414543152, "num_tokens": 47703335.0, "step": 1257 }, { "epoch": 0.16003053046686172, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6194977760314941, "learning_rate": 5.328529037727851e-06, "loss": 0.4854, "mean_token_accuracy": 0.8468279838562012, "num_tokens": 47746980.0, "step": 1258 }, { "epoch": 0.16015774074545222, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.8014183044433594e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6454598903656006, "learning_rate": 5.33276812208563e-06, "loss": 0.4821, "mean_token_accuracy": 0.8465065956115723, "num_tokens": 47784606.0, "step": 1259 }, { "epoch": 0.16028495102404275, "ewc_loss": 0.01904296875, "ewc_loss_diag": 2.8252601623535156e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 1.6429113149642944, "learning_rate": 5.337007206443409e-06, "loss": 0.5181, "mean_token_accuracy": 0.8385933041572571, "num_tokens": 47821298.0, "step": 1260 }, { "epoch": 0.16041216130263325, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.8252601623535156e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6183834075927734, "learning_rate": 5.341246290801188e-06, "loss": 0.4812, "mean_token_accuracy": 0.8508982062339783, "num_tokens": 47863102.0, "step": 1261 }, { "epoch": 0.16053937158122375, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.8252601623535156e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6214611530303955, "learning_rate": 5.345485375158966e-06, "loss": 0.4652, "mean_token_accuracy": 0.852972686290741, "num_tokens": 47903124.0, "step": 1262 }, { "epoch": 0.16066658185981428, "ewc_loss": 0.01904296875, "ewc_loss_diag": 2.8252601623535156e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 1.640592336654663, "learning_rate": 5.349724459516745e-06, "loss": 0.4621, "mean_token_accuracy": 0.8514301776885986, "num_tokens": 47939441.0, "step": 1263 }, { "epoch": 0.16079379213840478, "ewc_loss": 0.01904296875, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 1.6284162998199463, "learning_rate": 5.353963543874523e-06, "loss": 0.5287, "mean_token_accuracy": 0.8362661600112915, "num_tokens": 47981236.0, "step": 1264 }, { "epoch": 0.1609210024169953, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6619864702224731, "learning_rate": 5.358202628232302e-06, "loss": 0.4955, "mean_token_accuracy": 0.8463562726974487, "num_tokens": 48011431.0, "step": 1265 }, { "epoch": 0.1610482126955858, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6213098764419556, "learning_rate": 5.362441712590081e-06, "loss": 0.483, "mean_token_accuracy": 0.8484897613525391, "num_tokens": 48051003.0, "step": 1266 }, { "epoch": 0.1611754229741763, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6391184329986572, "learning_rate": 5.3666807969478595e-06, "loss": 0.5281, "mean_token_accuracy": 0.835518479347229, "num_tokens": 48091239.0, "step": 1267 }, { "epoch": 0.16130263325276684, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.8371810913085938e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.657637596130371, "learning_rate": 5.370919881305638e-06, "loss": 0.5154, "mean_token_accuracy": 0.8426939249038696, "num_tokens": 48126021.0, "step": 1268 }, { "epoch": 0.16142984353135734, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6490904092788696, "learning_rate": 5.375158965663417e-06, "loss": 0.4989, "mean_token_accuracy": 0.845827043056488, "num_tokens": 48160459.0, "step": 1269 }, { "epoch": 0.16155705380994784, "ewc_loss": 0.01904296875, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 1.6283457279205322, "learning_rate": 5.379398050021196e-06, "loss": 0.4866, "mean_token_accuracy": 0.8480267524719238, "num_tokens": 48200199.0, "step": 1270 }, { "epoch": 0.16168426408853837, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6518481969833374, "learning_rate": 5.383637134378975e-06, "loss": 0.4699, "mean_token_accuracy": 0.8496382236480713, "num_tokens": 48233926.0, "step": 1271 }, { "epoch": 0.16181147436712887, "ewc_loss": 0.019287109375, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.6438579559326172, "learning_rate": 5.387876218736753e-06, "loss": 0.5634, "mean_token_accuracy": 0.8214775323867798, "num_tokens": 48279667.0, "step": 1272 }, { "epoch": 0.16193868464571937, "ewc_loss": 0.01904296875, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.621246337890625e-05, "grad_norm": 1.6251130104064941, "learning_rate": 5.392115303094532e-06, "loss": 0.4516, "mean_token_accuracy": 0.857835054397583, "num_tokens": 48319185.0, "step": 1273 }, { "epoch": 0.1620658949243099, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6367422342300415, "learning_rate": 5.396354387452311e-06, "loss": 0.4395, "mean_token_accuracy": 0.8630150556564331, "num_tokens": 48357250.0, "step": 1274 }, { "epoch": 0.1621931052029004, "ewc_loss": 0.0191650390625, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6486340761184692, "learning_rate": 5.40059347181009e-06, "loss": 0.523, "mean_token_accuracy": 0.8373291492462158, "num_tokens": 48396577.0, "step": 1275 }, { "epoch": 0.1623203154814909, "ewc_loss": 0.019287109375, "ewc_loss_diag": 2.849102020263672e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.6724940538406372, "learning_rate": 5.404832556167869e-06, "loss": 0.5203, "mean_token_accuracy": 0.8350095748901367, "num_tokens": 48431360.0, "step": 1276 }, { "epoch": 0.16244752576008142, "ewc_loss": 0.019287109375, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.6426254510879517, "learning_rate": 5.4090716405256475e-06, "loss": 0.4861, "mean_token_accuracy": 0.8496263027191162, "num_tokens": 48471640.0, "step": 1277 }, { "epoch": 0.16257473603867192, "ewc_loss": 0.019287109375, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.6597460508346558, "learning_rate": 5.413310724883426e-06, "loss": 0.5427, "mean_token_accuracy": 0.8241763114929199, "num_tokens": 48509870.0, "step": 1278 }, { "epoch": 0.16270194631726242, "ewc_loss": 0.019287109375, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6322513818740845, "learning_rate": 5.417549809241205e-06, "loss": 0.5195, "mean_token_accuracy": 0.8407003283500671, "num_tokens": 48554174.0, "step": 1279 }, { "epoch": 0.16282915659585295, "ewc_loss": 0.019287109375, "ewc_loss_diag": 2.872943878173828e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6839978694915771, "learning_rate": 5.4217888935989825e-06, "loss": 0.5252, "mean_token_accuracy": 0.8369780778884888, "num_tokens": 48588590.0, "step": 1280 }, { "epoch": 0.16295636687444345, "ewc_loss": 0.019287109375, "ewc_loss_diag": 2.8848648071289062e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6271847486495972, "learning_rate": 5.426027977956761e-06, "loss": 0.544, "mean_token_accuracy": 0.8290537595748901, "num_tokens": 48633567.0, "step": 1281 }, { "epoch": 0.16308357715303395, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.6769732236862183, "learning_rate": 5.43026706231454e-06, "loss": 0.5639, "mean_token_accuracy": 0.8250603675842285, "num_tokens": 48673764.0, "step": 1282 }, { "epoch": 0.16321078743162448, "ewc_loss": 0.019287109375, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.633167266845703e-05, "grad_norm": 1.6228394508361816, "learning_rate": 5.434506146672319e-06, "loss": 0.5402, "mean_token_accuracy": 0.8319728970527649, "num_tokens": 48722531.0, "step": 1283 }, { "epoch": 0.16333799771021498, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.683709979057312, "learning_rate": 5.438745231030098e-06, "loss": 0.531, "mean_token_accuracy": 0.8316679000854492, "num_tokens": 48758273.0, "step": 1284 }, { "epoch": 0.16346520798880548, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.647214651107788, "learning_rate": 5.442984315387877e-06, "loss": 0.4246, "mean_token_accuracy": 0.8639450073242188, "num_tokens": 48793016.0, "step": 1285 }, { "epoch": 0.163592418267396, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 2.8967857360839844e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.661664605140686, "learning_rate": 5.447223399745656e-06, "loss": 0.4984, "mean_token_accuracy": 0.8460710644721985, "num_tokens": 48832204.0, "step": 1286 }, { "epoch": 0.1637196285459865, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.647408366203308, "learning_rate": 5.451462484103434e-06, "loss": 0.4708, "mean_token_accuracy": 0.853265643119812, "num_tokens": 48867199.0, "step": 1287 }, { "epoch": 0.163846838824577, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.6685478687286377, "learning_rate": 5.455701568461213e-06, "loss": 0.4884, "mean_token_accuracy": 0.8483403325080872, "num_tokens": 48901423.0, "step": 1288 }, { "epoch": 0.16397404910316754, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6650645732879639, "learning_rate": 5.459940652818992e-06, "loss": 0.5698, "mean_token_accuracy": 0.8229713439941406, "num_tokens": 48941883.0, "step": 1289 }, { "epoch": 0.16410125938175804, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 2.9087066650390625e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.6924811601638794, "learning_rate": 5.4641797371767705e-06, "loss": 0.5002, "mean_token_accuracy": 0.84232097864151, "num_tokens": 48971764.0, "step": 1290 }, { "epoch": 0.16422846966034857, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.660749912261963, "learning_rate": 5.468418821534549e-06, "loss": 0.4419, "mean_token_accuracy": 0.8594071865081787, "num_tokens": 49003687.0, "step": 1291 }, { "epoch": 0.16435567993893907, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.66446852684021, "learning_rate": 5.472657905892328e-06, "loss": 0.4495, "mean_token_accuracy": 0.8583062291145325, "num_tokens": 49036896.0, "step": 1292 }, { "epoch": 0.16448289021752957, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6625728607177734, "learning_rate": 5.476896990250107e-06, "loss": 0.566, "mean_token_accuracy": 0.8243362903594971, "num_tokens": 49076549.0, "step": 1293 }, { "epoch": 0.1646101004961201, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6462106704711914, "learning_rate": 5.481136074607886e-06, "loss": 0.451, "mean_token_accuracy": 0.8532317876815796, "num_tokens": 49114507.0, "step": 1294 }, { "epoch": 0.1647373107747106, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6422215700149536, "learning_rate": 5.485375158965663e-06, "loss": 0.4615, "mean_token_accuracy": 0.8565714359283447, "num_tokens": 49159794.0, "step": 1295 }, { "epoch": 0.1648645210533011, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6748241186141968, "learning_rate": 5.489614243323442e-06, "loss": 0.5015, "mean_token_accuracy": 0.8423364162445068, "num_tokens": 49198766.0, "step": 1296 }, { "epoch": 0.16499173133189163, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.651498556137085, "learning_rate": 5.493853327681221e-06, "loss": 0.446, "mean_token_accuracy": 0.8593400716781616, "num_tokens": 49236481.0, "step": 1297 }, { "epoch": 0.16511894161048213, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.9206275939941406e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6941500902175903, "learning_rate": 5.498092412039e-06, "loss": 0.5687, "mean_token_accuracy": 0.8267033696174622, "num_tokens": 49271730.0, "step": 1298 }, { "epoch": 0.16524615188907263, "ewc_loss": 0.0194091796875, "ewc_loss_diag": 2.944469451904297e-05, "ewc_loss_parallel": 1.6450881958007812e-05, "grad_norm": 1.6244994401931763, "learning_rate": 5.502331496396779e-06, "loss": 0.4593, "mean_token_accuracy": 0.8533410429954529, "num_tokens": 49316208.0, "step": 1299 }, { "epoch": 0.16537336216766316, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.944469451904297e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6803187131881714, "learning_rate": 5.506570580754558e-06, "loss": 0.4974, "mean_token_accuracy": 0.8438679575920105, "num_tokens": 49355343.0, "step": 1300 }, { "epoch": 0.16550057244625366, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.944469451904297e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6561201810836792, "learning_rate": 5.5108096651123365e-06, "loss": 0.4731, "mean_token_accuracy": 0.8526935577392578, "num_tokens": 49394055.0, "step": 1301 }, { "epoch": 0.16562778272484416, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.944469451904297e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6881837844848633, "learning_rate": 5.5150487494701146e-06, "loss": 0.5012, "mean_token_accuracy": 0.8446856141090393, "num_tokens": 49432466.0, "step": 1302 }, { "epoch": 0.1657549930034347, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6527924537658691, "learning_rate": 5.5192878338278934e-06, "loss": 0.4834, "mean_token_accuracy": 0.846640944480896, "num_tokens": 49470746.0, "step": 1303 }, { "epoch": 0.1658822032820252, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.7095906734466553, "learning_rate": 5.523526918185672e-06, "loss": 0.5282, "mean_token_accuracy": 0.8328372240066528, "num_tokens": 49502879.0, "step": 1304 }, { "epoch": 0.1660094135606157, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.66500723361969, "learning_rate": 5.527766002543451e-06, "loss": 0.4461, "mean_token_accuracy": 0.8557876348495483, "num_tokens": 49534875.0, "step": 1305 }, { "epoch": 0.16613662383920622, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6716338396072388, "learning_rate": 5.53200508690123e-06, "loss": 0.5142, "mean_token_accuracy": 0.8422826528549194, "num_tokens": 49573537.0, "step": 1306 }, { "epoch": 0.16626383411779672, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.956390380859375e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6479047536849976, "learning_rate": 5.536244171259009e-06, "loss": 0.4789, "mean_token_accuracy": 0.8507094383239746, "num_tokens": 49615190.0, "step": 1307 }, { "epoch": 0.16639104439638722, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.712141990661621, "learning_rate": 5.540483255616788e-06, "loss": 0.5777, "mean_token_accuracy": 0.8203630447387695, "num_tokens": 49649788.0, "step": 1308 }, { "epoch": 0.16651825467497774, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6652958393096924, "learning_rate": 5.544722339974567e-06, "loss": 0.4874, "mean_token_accuracy": 0.8471136689186096, "num_tokens": 49683728.0, "step": 1309 }, { "epoch": 0.16664546495356825, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6745949983596802, "learning_rate": 5.548961424332344e-06, "loss": 0.4805, "mean_token_accuracy": 0.843834638595581, "num_tokens": 49719331.0, "step": 1310 }, { "epoch": 0.16677267523215875, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6433194875717163, "learning_rate": 5.553200508690123e-06, "loss": 0.439, "mean_token_accuracy": 0.8609565496444702, "num_tokens": 49760158.0, "step": 1311 }, { "epoch": 0.16689988551074927, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.968311309814453e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6980804204940796, "learning_rate": 5.557439593047902e-06, "loss": 0.5399, "mean_token_accuracy": 0.8303058743476868, "num_tokens": 49795285.0, "step": 1312 }, { "epoch": 0.16702709578933977, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6508864164352417, "learning_rate": 5.561678677405681e-06, "loss": 0.495, "mean_token_accuracy": 0.8451648354530334, "num_tokens": 49839071.0, "step": 1313 }, { "epoch": 0.16715430606793028, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6990866661071777, "learning_rate": 5.5659177617634595e-06, "loss": 0.4734, "mean_token_accuracy": 0.8530100584030151, "num_tokens": 49870063.0, "step": 1314 }, { "epoch": 0.1672815163465208, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6600592136383057, "learning_rate": 5.570156846121238e-06, "loss": 0.5086, "mean_token_accuracy": 0.8394951820373535, "num_tokens": 49908232.0, "step": 1315 }, { "epoch": 0.1674087266251113, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6952165365219116, "learning_rate": 5.574395930479017e-06, "loss": 0.5321, "mean_token_accuracy": 0.8349676132202148, "num_tokens": 49944676.0, "step": 1316 }, { "epoch": 0.16753593690370183, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6751840114593506, "learning_rate": 5.578635014836796e-06, "loss": 0.4756, "mean_token_accuracy": 0.8525747656822205, "num_tokens": 49978472.0, "step": 1317 }, { "epoch": 0.16766314718229233, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.7094395160675049, "learning_rate": 5.582874099194574e-06, "loss": 0.526, "mean_token_accuracy": 0.8369725942611694, "num_tokens": 50010564.0, "step": 1318 }, { "epoch": 0.16779035746088283, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6620349884033203, "learning_rate": 5.587113183552353e-06, "loss": 0.5144, "mean_token_accuracy": 0.8389883041381836, "num_tokens": 50049538.0, "step": 1319 }, { "epoch": 0.16791756773947336, "ewc_loss": 0.01953125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6570091247558594e-05, "grad_norm": 1.6644936800003052, "learning_rate": 5.591352267910132e-06, "loss": 0.4706, "mean_token_accuracy": 0.8524597883224487, "num_tokens": 50090779.0, "step": 1320 }, { "epoch": 0.16804477801806386, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6906744241714478, "learning_rate": 5.595591352267911e-06, "loss": 0.5181, "mean_token_accuracy": 0.8402678966522217, "num_tokens": 50129527.0, "step": 1321 }, { "epoch": 0.16817198829665436, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6845499277114868, "learning_rate": 5.59983043662569e-06, "loss": 0.4811, "mean_token_accuracy": 0.8485367298126221, "num_tokens": 50167950.0, "step": 1322 }, { "epoch": 0.1682991985752449, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6846227645874023, "learning_rate": 5.604069520983469e-06, "loss": 0.4838, "mean_token_accuracy": 0.8489406108856201, "num_tokens": 50205079.0, "step": 1323 }, { "epoch": 0.1684264088538354, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6989803314208984, "learning_rate": 5.6083086053412475e-06, "loss": 0.4741, "mean_token_accuracy": 0.8502455353736877, "num_tokens": 50237909.0, "step": 1324 }, { "epoch": 0.1685536191324259, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.695347785949707, "learning_rate": 5.612547689699025e-06, "loss": 0.5362, "mean_token_accuracy": 0.836542546749115, "num_tokens": 50272825.0, "step": 1325 }, { "epoch": 0.16868082941101642, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6725939512252808, "learning_rate": 5.6167867740568036e-06, "loss": 0.5202, "mean_token_accuracy": 0.8407699465751648, "num_tokens": 50312964.0, "step": 1326 }, { "epoch": 0.16880803968960692, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6720542907714844, "learning_rate": 5.6210258584145824e-06, "loss": 0.521, "mean_token_accuracy": 0.836010754108429, "num_tokens": 50355912.0, "step": 1327 }, { "epoch": 0.16893524996819742, "ewc_loss": 0.0196533203125, "ewc_loss_diag": 2.9802322387695312e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6741013526916504, "learning_rate": 5.625264942772361e-06, "loss": 0.5049, "mean_token_accuracy": 0.8452506065368652, "num_tokens": 50395875.0, "step": 1328 }, { "epoch": 0.16906246024678795, "ewc_loss": 0.019775390625, "ewc_loss_diag": 2.9921531677246094e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.690816044807434, "learning_rate": 5.62950402713014e-06, "loss": 0.5573, "mean_token_accuracy": 0.8339118361473083, "num_tokens": 50435628.0, "step": 1329 }, { "epoch": 0.16918967052537845, "ewc_loss": 0.019775390625, "ewc_loss_diag": 2.9921531677246094e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6844905614852905, "learning_rate": 5.633743111487919e-06, "loss": 0.5605, "mean_token_accuracy": 0.8222934007644653, "num_tokens": 50474812.0, "step": 1330 }, { "epoch": 0.16931688080396895, "ewc_loss": 0.019775390625, "ewc_loss_diag": 3.0040740966796875e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.684831976890564, "learning_rate": 5.637982195845698e-06, "loss": 0.4406, "mean_token_accuracy": 0.8641060590744019, "num_tokens": 50510058.0, "step": 1331 }, { "epoch": 0.16944409108255948, "ewc_loss": 0.019775390625, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6801633834838867, "learning_rate": 5.642221280203477e-06, "loss": 0.4421, "mean_token_accuracy": 0.8614194393157959, "num_tokens": 50546282.0, "step": 1332 }, { "epoch": 0.16957130136114998, "ewc_loss": 0.019775390625, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.7245407104492188, "learning_rate": 5.646460364561255e-06, "loss": 0.4571, "mean_token_accuracy": 0.8544268608093262, "num_tokens": 50577457.0, "step": 1333 }, { "epoch": 0.16969851163974048, "ewc_loss": 0.019775390625, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6868398189544678, "learning_rate": 5.650699448919034e-06, "loss": 0.5558, "mean_token_accuracy": 0.8275737762451172, "num_tokens": 50619754.0, "step": 1334 }, { "epoch": 0.169825721918331, "ewc_loss": 0.019775390625, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.700883388519287, "learning_rate": 5.654938533276813e-06, "loss": 0.4914, "mean_token_accuracy": 0.8453806042671204, "num_tokens": 50660674.0, "step": 1335 }, { "epoch": 0.1699529321969215, "ewc_loss": 0.019775390625, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.679634690284729, "learning_rate": 5.6591776176345916e-06, "loss": 0.4845, "mean_token_accuracy": 0.8474934697151184, "num_tokens": 50698404.0, "step": 1336 }, { "epoch": 0.170080142475512, "ewc_loss": 0.019775390625, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6882728338241577, "learning_rate": 5.6634167019923704e-06, "loss": 0.4927, "mean_token_accuracy": 0.8445403575897217, "num_tokens": 50736329.0, "step": 1337 }, { "epoch": 0.17020735275410254, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 1.7088596820831299, "learning_rate": 5.667655786350149e-06, "loss": 0.5505, "mean_token_accuracy": 0.8223439455032349, "num_tokens": 50770065.0, "step": 1338 }, { "epoch": 0.17033456303269304, "ewc_loss": 0.019775390625, "ewc_loss_diag": 3.0159950256347656e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6810508966445923, "learning_rate": 5.671894870707928e-06, "loss": 0.4865, "mean_token_accuracy": 0.8507851958274841, "num_tokens": 50811603.0, "step": 1339 }, { "epoch": 0.17046177331128357, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 1.6835639476776123, "learning_rate": 5.676133955065705e-06, "loss": 0.4278, "mean_token_accuracy": 0.8625553846359253, "num_tokens": 50846141.0, "step": 1340 }, { "epoch": 0.17058898358987407, "ewc_loss": 0.019775390625, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6875311136245728, "learning_rate": 5.680373039423484e-06, "loss": 0.4754, "mean_token_accuracy": 0.8517091274261475, "num_tokens": 50885511.0, "step": 1341 }, { "epoch": 0.17071619386846457, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 1.6986907720565796, "learning_rate": 5.684612123781263e-06, "loss": 0.5195, "mean_token_accuracy": 0.8350821733474731, "num_tokens": 50926374.0, "step": 1342 }, { "epoch": 0.1708434041470551, "ewc_loss": 0.019775390625, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6908740997314453, "learning_rate": 5.688851208139042e-06, "loss": 0.5322, "mean_token_accuracy": 0.8299057483673096, "num_tokens": 50965453.0, "step": 1343 }, { "epoch": 0.1709706144256456, "ewc_loss": 0.019775390625, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.6689300537109375e-05, "grad_norm": 1.6842085123062134, "learning_rate": 5.693090292496821e-06, "loss": 0.5232, "mean_token_accuracy": 0.8411328792572021, "num_tokens": 51007921.0, "step": 1344 }, { "epoch": 0.1710978247042361, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 3.0279159545898438e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 1.6885610818862915, "learning_rate": 5.6973293768546e-06, "loss": 0.5, "mean_token_accuracy": 0.8428177833557129, "num_tokens": 51046982.0, "step": 1345 }, { "epoch": 0.17122503498282662, "ewc_loss": 0.02001953125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 1.6900007724761963, "learning_rate": 5.701568461212379e-06, "loss": 0.4375, "mean_token_accuracy": 0.8628242611885071, "num_tokens": 51083192.0, "step": 1346 }, { "epoch": 0.17135224526141712, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 1.678774118423462, "learning_rate": 5.705807545570158e-06, "loss": 0.4477, "mean_token_accuracy": 0.8586374521255493, "num_tokens": 51125258.0, "step": 1347 }, { "epoch": 0.17147945554000762, "ewc_loss": 0.02001953125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 1.691710114479065, "learning_rate": 5.710046629927936e-06, "loss": 0.4835, "mean_token_accuracy": 0.8477705121040344, "num_tokens": 51164036.0, "step": 1348 }, { "epoch": 0.17160666581859815, "ewc_loss": 0.02001953125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 1.6904585361480713, "learning_rate": 5.7142857142857145e-06, "loss": 0.4494, "mean_token_accuracy": 0.8570862412452698, "num_tokens": 51201288.0, "step": 1349 }, { "epoch": 0.17173387609718865, "ewc_loss": 0.02001953125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 1.7073471546173096, "learning_rate": 5.718524798643493e-06, "loss": 0.5181, "mean_token_accuracy": 0.8380144834518433, "num_tokens": 51235931.0, "step": 1350 }, { "epoch": 0.17186108637577915, "ewc_loss": 0.02001953125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 1.6849321126937866, "learning_rate": 5.722763883001272e-06, "loss": 0.5248, "mean_token_accuracy": 0.8375887870788574, "num_tokens": 51277188.0, "step": 1351 }, { "epoch": 0.17198829665436968, "ewc_loss": 0.02001953125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 1.7102303504943848, "learning_rate": 5.727002967359051e-06, "loss": 0.4921, "mean_token_accuracy": 0.8441325426101685, "num_tokens": 51313829.0, "step": 1352 }, { "epoch": 0.17211550693296018, "ewc_loss": 0.0198974609375, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.6808509826660156e-05, "grad_norm": 1.6944700479507446, "learning_rate": 5.73124205171683e-06, "loss": 0.5541, "mean_token_accuracy": 0.8241232633590698, "num_tokens": 51355678.0, "step": 1353 }, { "epoch": 0.17224271721155068, "ewc_loss": 0.02001953125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 1.705105185508728, "learning_rate": 5.735481136074609e-06, "loss": 0.4651, "mean_token_accuracy": 0.8530746698379517, "num_tokens": 51390125.0, "step": 1354 }, { "epoch": 0.1723699274901412, "ewc_loss": 0.02001953125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 1.6928600072860718, "learning_rate": 5.739720220432386e-06, "loss": 0.4648, "mean_token_accuracy": 0.8506718873977661, "num_tokens": 51426384.0, "step": 1355 }, { "epoch": 0.1724971377687317, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 1.708668828010559, "learning_rate": 5.743959304790165e-06, "loss": 0.5598, "mean_token_accuracy": 0.8302851915359497, "num_tokens": 51467087.0, "step": 1356 }, { "epoch": 0.1726243480473222, "ewc_loss": 0.02001953125, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 1.6966092586517334, "learning_rate": 5.748198389147944e-06, "loss": 0.5153, "mean_token_accuracy": 0.8396689891815186, "num_tokens": 51505126.0, "step": 1357 }, { "epoch": 0.17275155832591274, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 3.039836883544922e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 1.718886375427246, "learning_rate": 5.752437473505723e-06, "loss": 0.5381, "mean_token_accuracy": 0.8317369222640991, "num_tokens": 51546623.0, "step": 1358 }, { "epoch": 0.17287876860450324, "ewc_loss": 0.02001953125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 1.7224410772323608, "learning_rate": 5.756676557863502e-06, "loss": 0.5086, "mean_token_accuracy": 0.8397935032844543, "num_tokens": 51583175.0, "step": 1359 }, { "epoch": 0.17300597888309374, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 1.6858782768249512, "learning_rate": 5.7609156422212805e-06, "loss": 0.4657, "mean_token_accuracy": 0.8520914912223816, "num_tokens": 51621454.0, "step": 1360 }, { "epoch": 0.17313318916168427, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 1.7115558385849, "learning_rate": 5.7651547265790594e-06, "loss": 0.4889, "mean_token_accuracy": 0.8471634984016418, "num_tokens": 51659482.0, "step": 1361 }, { "epoch": 0.17326039944027477, "ewc_loss": 0.02001953125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 1.7075698375701904, "learning_rate": 5.769393810936838e-06, "loss": 0.4917, "mean_token_accuracy": 0.8464776277542114, "num_tokens": 51695478.0, "step": 1362 }, { "epoch": 0.17338760971886527, "ewc_loss": 0.020263671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 1.7050987482070923, "learning_rate": 5.773632895294616e-06, "loss": 0.5373, "mean_token_accuracy": 0.8292027711868286, "num_tokens": 51736436.0, "step": 1363 }, { "epoch": 0.1735148199974558, "ewc_loss": 0.02001953125, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.6927719116210938e-05, "grad_norm": 1.7063801288604736, "learning_rate": 5.777871979652395e-06, "loss": 0.4689, "mean_token_accuracy": 0.8519752025604248, "num_tokens": 51772502.0, "step": 1364 }, { "epoch": 0.1736420302760463, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 1.6954689025878906, "learning_rate": 5.782111064010174e-06, "loss": 0.5029, "mean_token_accuracy": 0.8409241437911987, "num_tokens": 51816343.0, "step": 1365 }, { "epoch": 0.17376924055463683, "ewc_loss": 0.020263671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 1.7062273025512695, "learning_rate": 5.786350148367953e-06, "loss": 0.4964, "mean_token_accuracy": 0.8448532819747925, "num_tokens": 51861317.0, "step": 1366 }, { "epoch": 0.17389645083322733, "ewc_loss": 0.020263671875, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 1.7013334035873413, "learning_rate": 5.790589232725732e-06, "loss": 0.4984, "mean_token_accuracy": 0.8414905071258545, "num_tokens": 51902994.0, "step": 1367 }, { "epoch": 0.17402366111181783, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 1.7279473543167114, "learning_rate": 5.794828317083511e-06, "loss": 0.4839, "mean_token_accuracy": 0.8486356735229492, "num_tokens": 51936374.0, "step": 1368 }, { "epoch": 0.17415087139040836, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 3.0517578125e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 1.726078748703003, "learning_rate": 5.79906740144129e-06, "loss": 0.4831, "mean_token_accuracy": 0.8481062650680542, "num_tokens": 51970235.0, "step": 1369 }, { "epoch": 0.17427808166899886, "ewc_loss": 0.020263671875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 1.7101774215698242, "learning_rate": 5.8033064857990685e-06, "loss": 0.4719, "mean_token_accuracy": 0.852635383605957, "num_tokens": 52006220.0, "step": 1370 }, { "epoch": 0.17440529194758936, "ewc_loss": 0.020263671875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 1.695216417312622, "learning_rate": 5.807545570156847e-06, "loss": 0.4541, "mean_token_accuracy": 0.8567306995391846, "num_tokens": 52046634.0, "step": 1371 }, { "epoch": 0.1745325022261799, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 1.7512989044189453, "learning_rate": 5.811784654514625e-06, "loss": 0.5434, "mean_token_accuracy": 0.8303803205490112, "num_tokens": 52079768.0, "step": 1372 }, { "epoch": 0.1746597125047704, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 1.6911855936050415, "learning_rate": 5.8160237388724035e-06, "loss": 0.4679, "mean_token_accuracy": 0.8515126705169678, "num_tokens": 52118160.0, "step": 1373 }, { "epoch": 0.1747869227833609, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 1.7170161008834839, "learning_rate": 5.820262823230182e-06, "loss": 0.5086, "mean_token_accuracy": 0.8415528535842896, "num_tokens": 52160219.0, "step": 1374 }, { "epoch": 0.17491413306195142, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 1.7162786722183228, "learning_rate": 5.824501907587961e-06, "loss": 0.465, "mean_token_accuracy": 0.85260409116745, "num_tokens": 52194746.0, "step": 1375 }, { "epoch": 0.17504134334054192, "ewc_loss": 0.020263671875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 1.7388373613357544, "learning_rate": 5.82874099194574e-06, "loss": 0.5557, "mean_token_accuracy": 0.8238641619682312, "num_tokens": 52232625.0, "step": 1376 }, { "epoch": 0.17516855361913242, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 1.712278962135315, "learning_rate": 5.832980076303519e-06, "loss": 0.5548, "mean_token_accuracy": 0.8271448612213135, "num_tokens": 52273030.0, "step": 1377 }, { "epoch": 0.17529576389772294, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 1.727358341217041, "learning_rate": 5.837219160661297e-06, "loss": 0.4468, "mean_token_accuracy": 0.8570863008499146, "num_tokens": 52306490.0, "step": 1378 }, { "epoch": 0.17542297417631345, "ewc_loss": 0.020263671875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 1.6942745447158813, "learning_rate": 5.841458245019076e-06, "loss": 0.4417, "mean_token_accuracy": 0.860292911529541, "num_tokens": 52346717.0, "step": 1379 }, { "epoch": 0.17555018445490395, "ewc_loss": 0.0205078125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 1.7817161083221436, "learning_rate": 5.845697329376855e-06, "loss": 0.4911, "mean_token_accuracy": 0.8474007248878479, "num_tokens": 52378165.0, "step": 1380 }, { "epoch": 0.17567739473349447, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 1.7137646675109863, "learning_rate": 5.849936413734634e-06, "loss": 0.554, "mean_token_accuracy": 0.8285543918609619, "num_tokens": 52418719.0, "step": 1381 }, { "epoch": 0.17580460501208497, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 1.7180677652359009, "learning_rate": 5.854175498092413e-06, "loss": 0.4644, "mean_token_accuracy": 0.8542900681495667, "num_tokens": 52459094.0, "step": 1382 }, { "epoch": 0.17593181529067548, "ewc_loss": 0.020263671875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 1.7028571367263794, "learning_rate": 5.8584145824501915e-06, "loss": 0.5001, "mean_token_accuracy": 0.8436221480369568, "num_tokens": 52503117.0, "step": 1383 }, { "epoch": 0.176059025569266, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 1.7674843072891235, "learning_rate": 5.86265366680797e-06, "loss": 0.5221, "mean_token_accuracy": 0.835645854473114, "num_tokens": 52542643.0, "step": 1384 }, { "epoch": 0.1761862358478565, "ewc_loss": 0.0201416015625, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.704692840576172e-05, "grad_norm": 1.6960471868515015, "learning_rate": 5.866892751165749e-06, "loss": 0.4993, "mean_token_accuracy": 0.8440000414848328, "num_tokens": 52585885.0, "step": 1385 }, { "epoch": 0.176313446126447, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 1.7605284452438354, "learning_rate": 5.871131835523527e-06, "loss": 0.5473, "mean_token_accuracy": 0.8293224573135376, "num_tokens": 52621629.0, "step": 1386 }, { "epoch": 0.17644065640503753, "ewc_loss": 0.020263671875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 1.7170114517211914, "learning_rate": 5.875370919881306e-06, "loss": 0.5004, "mean_token_accuracy": 0.8393281698226929, "num_tokens": 52667056.0, "step": 1387 }, { "epoch": 0.17656786668362803, "ewc_loss": 0.0205078125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 1.7542657852172852, "learning_rate": 5.879610004239085e-06, "loss": 0.5572, "mean_token_accuracy": 0.8301770091056824, "num_tokens": 52705857.0, "step": 1388 }, { "epoch": 0.17669507696221853, "ewc_loss": 0.020263671875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.71661376953125e-05, "grad_norm": 1.7277096509933472, "learning_rate": 5.883849088596864e-06, "loss": 0.542, "mean_token_accuracy": 0.8307824730873108, "num_tokens": 52748908.0, "step": 1389 }, { "epoch": 0.17682228724080906, "ewc_loss": 0.020751953125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7454885244369507, "learning_rate": 5.888088172954642e-06, "loss": 0.5169, "mean_token_accuracy": 0.8407589197158813, "num_tokens": 52783600.0, "step": 1390 }, { "epoch": 0.17694949751939956, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 1.7219257354736328, "learning_rate": 5.892327257312421e-06, "loss": 0.4407, "mean_token_accuracy": 0.863113522529602, "num_tokens": 52819873.0, "step": 1391 }, { "epoch": 0.1770767077979901, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 1.7245852947235107, "learning_rate": 5.8965663416702e-06, "loss": 0.4426, "mean_token_accuracy": 0.8568962216377258, "num_tokens": 52858883.0, "step": 1392 }, { "epoch": 0.1772039180765806, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 1.7364197969436646, "learning_rate": 5.900805426027978e-06, "loss": 0.4955, "mean_token_accuracy": 0.8457640409469604, "num_tokens": 52895501.0, "step": 1393 }, { "epoch": 0.1773311283551711, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 3.075599670410156e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 1.7141753435134888, "learning_rate": 5.905044510385757e-06, "loss": 0.4425, "mean_token_accuracy": 0.859101414680481, "num_tokens": 52934678.0, "step": 1394 }, { "epoch": 0.17745833863376162, "ewc_loss": 0.020751953125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.741063117980957, "learning_rate": 5.909283594743536e-06, "loss": 0.5477, "mean_token_accuracy": 0.8264004588127136, "num_tokens": 52972540.0, "step": 1395 }, { "epoch": 0.17758554891235212, "ewc_loss": 0.0203857421875, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.728534698486328e-05, "grad_norm": 1.720231056213379, "learning_rate": 5.9135226791013145e-06, "loss": 0.4835, "mean_token_accuracy": 0.8441426753997803, "num_tokens": 53010819.0, "step": 1396 }, { "epoch": 0.17771275919094262, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 1.7408169507980347, "learning_rate": 5.917761763459093e-06, "loss": 0.5478, "mean_token_accuracy": 0.8284918069839478, "num_tokens": 53050142.0, "step": 1397 }, { "epoch": 0.17783996946953315, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 1.7380056381225586, "learning_rate": 5.922000847816872e-06, "loss": 0.4764, "mean_token_accuracy": 0.8518319725990295, "num_tokens": 53087306.0, "step": 1398 }, { "epoch": 0.17796717974812365, "ewc_loss": 0.020751953125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7276850938796997, "learning_rate": 5.926239932174651e-06, "loss": 0.5341, "mean_token_accuracy": 0.8328241109848022, "num_tokens": 53131850.0, "step": 1399 }, { "epoch": 0.17809439002671415, "ewc_loss": 0.0205078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 1.7321498394012451, "learning_rate": 5.93047901653243e-06, "loss": 0.4885, "mean_token_accuracy": 0.8439128398895264, "num_tokens": 53168902.0, "step": 1400 }, { "epoch": 0.17822160030530468, "ewc_loss": 0.0205078125, "ewc_loss_diag": 3.0994415283203125e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 1.7222678661346436, "learning_rate": 5.934718100890208e-06, "loss": 0.4655, "mean_token_accuracy": 0.8518006205558777, "num_tokens": 53208497.0, "step": 1401 }, { "epoch": 0.17834881058389518, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7548972368240356, "learning_rate": 5.938957185247987e-06, "loss": 0.5203, "mean_token_accuracy": 0.8344037532806396, "num_tokens": 53245618.0, "step": 1402 }, { "epoch": 0.17847602086248568, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 1.7453356981277466, "learning_rate": 5.943196269605766e-06, "loss": 0.4815, "mean_token_accuracy": 0.84662264585495, "num_tokens": 53280315.0, "step": 1403 }, { "epoch": 0.1786032311410762, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 1.7561776638031006, "learning_rate": 5.947435353963545e-06, "loss": 0.5196, "mean_token_accuracy": 0.8396028280258179, "num_tokens": 53313291.0, "step": 1404 }, { "epoch": 0.1787304414196667, "ewc_loss": 0.020751953125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 1.740126609802246, "learning_rate": 5.951674438321324e-06, "loss": 0.5058, "mean_token_accuracy": 0.8429505228996277, "num_tokens": 53347764.0, "step": 1405 }, { "epoch": 0.1788576516982572, "ewc_loss": 0.020751953125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 1.7474709749221802, "learning_rate": 5.9559135226791025e-06, "loss": 0.5061, "mean_token_accuracy": 0.8379659652709961, "num_tokens": 53385184.0, "step": 1406 }, { "epoch": 0.17898486197684774, "ewc_loss": 0.020751953125, "ewc_loss_diag": 3.123283386230469e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 1.7380162477493286, "learning_rate": 5.9601526070368805e-06, "loss": 0.5799, "mean_token_accuracy": 0.8179259300231934, "num_tokens": 53425850.0, "step": 1407 }, { "epoch": 0.17911207225543824, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7185757160186768, "learning_rate": 5.964391691394659e-06, "loss": 0.5142, "mean_token_accuracy": 0.8409426212310791, "num_tokens": 53472444.0, "step": 1408 }, { "epoch": 0.17923928253402874, "ewc_loss": 0.020751953125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 1.7638579607009888, "learning_rate": 5.968630775752437e-06, "loss": 0.5517, "mean_token_accuracy": 0.82911217212677, "num_tokens": 53506996.0, "step": 1409 }, { "epoch": 0.17936649281261927, "ewc_loss": 0.0206298828125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 1.7554973363876343, "learning_rate": 5.972869860110216e-06, "loss": 0.526, "mean_token_accuracy": 0.8324185609817505, "num_tokens": 53542005.0, "step": 1410 }, { "epoch": 0.17949370309120977, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7690207958221436, "learning_rate": 5.977108944467995e-06, "loss": 0.5632, "mean_token_accuracy": 0.8309924602508545, "num_tokens": 53582130.0, "step": 1411 }, { "epoch": 0.17962091336980027, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7229496240615845, "learning_rate": 5.981348028825774e-06, "loss": 0.4817, "mean_token_accuracy": 0.8523414731025696, "num_tokens": 53624528.0, "step": 1412 }, { "epoch": 0.1797481236483908, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7822887897491455, "learning_rate": 5.985587113183553e-06, "loss": 0.5291, "mean_token_accuracy": 0.8361489176750183, "num_tokens": 53663701.0, "step": 1413 }, { "epoch": 0.1798753339269813, "ewc_loss": 0.020751953125, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 1.7529460191726685, "learning_rate": 5.989826197541332e-06, "loss": 0.5179, "mean_token_accuracy": 0.8415948152542114, "num_tokens": 53700655.0, "step": 1414 }, { "epoch": 0.18000254420557182, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7280536890029907, "learning_rate": 5.994065281899111e-06, "loss": 0.5386, "mean_token_accuracy": 0.835263729095459, "num_tokens": 53747387.0, "step": 1415 }, { "epoch": 0.18012975448416232, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.731864333152771, "learning_rate": 5.998304366256889e-06, "loss": 0.5063, "mean_token_accuracy": 0.8415218591690063, "num_tokens": 53789416.0, "step": 1416 }, { "epoch": 0.18025696476275282, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7449827194213867, "learning_rate": 6.002543450614668e-06, "loss": 0.4313, "mean_token_accuracy": 0.8674167990684509, "num_tokens": 53824300.0, "step": 1417 }, { "epoch": 0.18038417504134335, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7561140060424805, "learning_rate": 6.0067825349724465e-06, "loss": 0.5062, "mean_token_accuracy": 0.8425949811935425, "num_tokens": 53861366.0, "step": 1418 }, { "epoch": 0.18051138531993385, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.147125244140625e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7594174146652222, "learning_rate": 6.011021619330225e-06, "loss": 0.5296, "mean_token_accuracy": 0.8359942436218262, "num_tokens": 53899655.0, "step": 1419 }, { "epoch": 0.18063859559852435, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7580028772354126, "learning_rate": 6.015260703688004e-06, "loss": 0.5078, "mean_token_accuracy": 0.8372004628181458, "num_tokens": 53935563.0, "step": 1420 }, { "epoch": 0.18076580587711488, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7602545022964478, "learning_rate": 6.019499788045783e-06, "loss": 0.5066, "mean_token_accuracy": 0.8423217535018921, "num_tokens": 53970490.0, "step": 1421 }, { "epoch": 0.18089301615570538, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7703450918197632, "learning_rate": 6.023738872403562e-06, "loss": 0.4722, "mean_token_accuracy": 0.8503224849700928, "num_tokens": 54000460.0, "step": 1422 }, { "epoch": 0.18102022643429588, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7467851638793945, "learning_rate": 6.027977956761341e-06, "loss": 0.4986, "mean_token_accuracy": 0.8483476638793945, "num_tokens": 54038007.0, "step": 1423 }, { "epoch": 0.1811474367128864, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.752251148223877, "learning_rate": 6.032217041119118e-06, "loss": 0.5094, "mean_token_accuracy": 0.8360925912857056, "num_tokens": 54074515.0, "step": 1424 }, { "epoch": 0.1812746469914769, "ewc_loss": 0.020751953125, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.7523765563964844e-05, "grad_norm": 1.7591615915298462, "learning_rate": 6.036456125476897e-06, "loss": 0.5442, "mean_token_accuracy": 0.8302206993103027, "num_tokens": 54109355.0, "step": 1425 }, { "epoch": 0.1814018572700674, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7419830560684204, "learning_rate": 6.040695209834676e-06, "loss": 0.445, "mean_token_accuracy": 0.856960654258728, "num_tokens": 54150966.0, "step": 1426 }, { "epoch": 0.18152906754865794, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7689013481140137, "learning_rate": 6.044934294192455e-06, "loss": 0.5396, "mean_token_accuracy": 0.831405758857727, "num_tokens": 54188398.0, "step": 1427 }, { "epoch": 0.18165627782724844, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7317757606506348, "learning_rate": 6.049173378550234e-06, "loss": 0.4951, "mean_token_accuracy": 0.8440719842910767, "num_tokens": 54233669.0, "step": 1428 }, { "epoch": 0.18178348810583894, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.769765019416809, "learning_rate": 6.0534124629080126e-06, "loss": 0.4775, "mean_token_accuracy": 0.8495999574661255, "num_tokens": 54272377.0, "step": 1429 }, { "epoch": 0.18191069838442947, "ewc_loss": 0.0208740234375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.7642974853515625e-05, "grad_norm": 1.7437644004821777, "learning_rate": 6.0576515472657915e-06, "loss": 0.5471, "mean_token_accuracy": 0.8315657377243042, "num_tokens": 54313831.0, "step": 1430 }, { "epoch": 0.18203790866301997, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.1948089599609375e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.7464886903762817, "learning_rate": 6.0618906316235695e-06, "loss": 0.4525, "mean_token_accuracy": 0.8557614684104919, "num_tokens": 54356724.0, "step": 1431 }, { "epoch": 0.18216511894161047, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7406281232833862, "learning_rate": 6.066129715981348e-06, "loss": 0.4859, "mean_token_accuracy": 0.845644474029541, "num_tokens": 54398407.0, "step": 1432 }, { "epoch": 0.182292329220201, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.218650817871094e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.7660555839538574, "learning_rate": 6.070368800339127e-06, "loss": 0.4443, "mean_token_accuracy": 0.8611345291137695, "num_tokens": 54430023.0, "step": 1433 }, { "epoch": 0.1824195394987915, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.758590579032898, "learning_rate": 6.074607884696906e-06, "loss": 0.4756, "mean_token_accuracy": 0.849163293838501, "num_tokens": 54467154.0, "step": 1434 }, { "epoch": 0.182546749777382, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.780440330505371, "learning_rate": 6.078846969054685e-06, "loss": 0.5057, "mean_token_accuracy": 0.8427325487136841, "num_tokens": 54501139.0, "step": 1435 }, { "epoch": 0.18267396005597253, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7774665355682373, "learning_rate": 6.083086053412464e-06, "loss": 0.486, "mean_token_accuracy": 0.8480328917503357, "num_tokens": 54541084.0, "step": 1436 }, { "epoch": 0.18280117033456303, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7604118585586548, "learning_rate": 6.087325137770243e-06, "loss": 0.4667, "mean_token_accuracy": 0.8527169227600098, "num_tokens": 54576745.0, "step": 1437 }, { "epoch": 0.18292838061315353, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.7574505805969238, "learning_rate": 6.091564222128022e-06, "loss": 0.5455, "mean_token_accuracy": 0.8334915637969971, "num_tokens": 54619039.0, "step": 1438 }, { "epoch": 0.18305559089174406, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.749954104423523, "learning_rate": 6.095803306485799e-06, "loss": 0.4872, "mean_token_accuracy": 0.8450584411621094, "num_tokens": 54660009.0, "step": 1439 }, { "epoch": 0.18318280117033456, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.7630646228790283, "learning_rate": 6.100042390843578e-06, "loss": 0.5212, "mean_token_accuracy": 0.8390791416168213, "num_tokens": 54701253.0, "step": 1440 }, { "epoch": 0.1833100114489251, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7652500867843628, "learning_rate": 6.104281475201357e-06, "loss": 0.4753, "mean_token_accuracy": 0.8501737117767334, "num_tokens": 54738406.0, "step": 1441 }, { "epoch": 0.1834372217275156, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.8029645681381226, "learning_rate": 6.1085205595591355e-06, "loss": 0.5285, "mean_token_accuracy": 0.835218071937561, "num_tokens": 54771808.0, "step": 1442 }, { "epoch": 0.1835644320061061, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7515720129013062, "learning_rate": 6.112759643916914e-06, "loss": 0.4275, "mean_token_accuracy": 0.8650304079055786, "num_tokens": 54807337.0, "step": 1443 }, { "epoch": 0.18369164228469662, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.781786322593689, "learning_rate": 6.116998728274693e-06, "loss": 0.5143, "mean_token_accuracy": 0.8411438465118408, "num_tokens": 54844932.0, "step": 1444 }, { "epoch": 0.18381885256328712, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7509034872055054, "learning_rate": 6.121237812632472e-06, "loss": 0.5163, "mean_token_accuracy": 0.8349399566650391, "num_tokens": 54884492.0, "step": 1445 }, { "epoch": 0.18394606284187762, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.7796788215637207, "learning_rate": 6.12547689699025e-06, "loss": 0.4673, "mean_token_accuracy": 0.856665849685669, "num_tokens": 54924164.0, "step": 1446 }, { "epoch": 0.18407327312046814, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7605761289596558, "learning_rate": 6.129715981348029e-06, "loss": 0.4985, "mean_token_accuracy": 0.8468817472457886, "num_tokens": 54961676.0, "step": 1447 }, { "epoch": 0.18420048339905865, "ewc_loss": 0.02099609375, "ewc_loss_diag": 3.24249267578125e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.7885048389434814, "learning_rate": 6.133955065705808e-06, "loss": 0.5243, "mean_token_accuracy": 0.8334578275680542, "num_tokens": 55000889.0, "step": 1448 }, { "epoch": 0.18432769367764915, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7532799243927002, "learning_rate": 6.138194150063587e-06, "loss": 0.4894, "mean_token_accuracy": 0.847323477268219, "num_tokens": 55039471.0, "step": 1449 }, { "epoch": 0.18445490395623967, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.7898056507110596, "learning_rate": 6.142433234421366e-06, "loss": 0.5227, "mean_token_accuracy": 0.8343798518180847, "num_tokens": 55078408.0, "step": 1450 }, { "epoch": 0.18458211423483017, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7738285064697266, "learning_rate": 6.146672318779145e-06, "loss": 0.5241, "mean_token_accuracy": 0.8365501761436462, "num_tokens": 55116068.0, "step": 1451 }, { "epoch": 0.18470932451342068, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.794779896736145, "learning_rate": 6.1509114031369235e-06, "loss": 0.5104, "mean_token_accuracy": 0.8437179327011108, "num_tokens": 55149933.0, "step": 1452 }, { "epoch": 0.1848365347920112, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.7762184143066406e-05, "grad_norm": 1.7610013484954834, "learning_rate": 6.155150487494702e-06, "loss": 0.4401, "mean_token_accuracy": 0.8615766763687134, "num_tokens": 55184674.0, "step": 1453 }, { "epoch": 0.1849637450706017, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.7630555629730225, "learning_rate": 6.15938957185248e-06, "loss": 0.4739, "mean_token_accuracy": 0.8513391613960266, "num_tokens": 55227918.0, "step": 1454 }, { "epoch": 0.1850909553491922, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.7509052753448486, "learning_rate": 6.1636286562102585e-06, "loss": 0.484, "mean_token_accuracy": 0.8481536507606506, "num_tokens": 55274312.0, "step": 1455 }, { "epoch": 0.18521816562778273, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.8004639148712158, "learning_rate": 6.167867740568037e-06, "loss": 0.5515, "mean_token_accuracy": 0.8276781439781189, "num_tokens": 55310583.0, "step": 1456 }, { "epoch": 0.18534537590637323, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 1.7603580951690674, "learning_rate": 6.172106824925816e-06, "loss": 0.4803, "mean_token_accuracy": 0.8465193510055542, "num_tokens": 55352365.0, "step": 1457 }, { "epoch": 0.18547258618496373, "ewc_loss": 0.0211181640625, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.7881393432617188e-05, "grad_norm": 3.0602269172668457, "learning_rate": 6.176345909283595e-06, "loss": 0.4952, "mean_token_accuracy": 0.8437421321868896, "num_tokens": 55391268.0, "step": 1458 }, { "epoch": 0.18559979646355426, "ewc_loss": 0.03076171875, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 2.7418136596679688e-05, "grad_norm": 2.0689923763275146, "learning_rate": 6.180584993641374e-06, "loss": 0.4837, "mean_token_accuracy": 0.8494849801063538, "num_tokens": 55428842.0, "step": 1459 }, { "epoch": 0.18572700674214476, "ewc_loss": 0.020751953125, "ewc_loss_diag": 3.266334533691406e-05, "ewc_loss_parallel": 1.7404556274414062e-05, "grad_norm": 1.718557357788086, "learning_rate": 6.184824077999153e-06, "loss": 0.4543, "mean_token_accuracy": 0.8557871580123901, "num_tokens": 55464341.0, "step": 1460 }, { "epoch": 0.18585421702073526, "ewc_loss": 0.0234375, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 1.9187772274017334, "learning_rate": 6.189063162356932e-06, "loss": 0.5522, "mean_token_accuracy": 0.8271118998527527, "num_tokens": 55498985.0, "step": 1461 }, { "epoch": 0.1859814272993258, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 1.7897772789001465, "learning_rate": 6.19330224671471e-06, "loss": 0.4793, "mean_token_accuracy": 0.8482354879379272, "num_tokens": 55537762.0, "step": 1462 }, { "epoch": 0.1861086375779163, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 3.2901763916015625e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 1.8627524375915527, "learning_rate": 6.197541331072489e-06, "loss": 0.5248, "mean_token_accuracy": 0.8359522819519043, "num_tokens": 55572571.0, "step": 1463 }, { "epoch": 0.1862358478565068, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.7840496301651, "learning_rate": 6.201780415430268e-06, "loss": 0.4883, "mean_token_accuracy": 0.8486881256103516, "num_tokens": 55611858.0, "step": 1464 }, { "epoch": 0.18636305813509732, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.8424408435821533, "learning_rate": 6.2060194997880465e-06, "loss": 0.5266, "mean_token_accuracy": 0.836156964302063, "num_tokens": 55649168.0, "step": 1465 }, { "epoch": 0.18649026841368782, "ewc_loss": 0.02197265625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 1.7897053956985474, "learning_rate": 6.210258584145825e-06, "loss": 0.4671, "mean_token_accuracy": 0.8503526449203491, "num_tokens": 55686127.0, "step": 1466 }, { "epoch": 0.18661747869227835, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 1.8281415700912476, "learning_rate": 6.214497668503604e-06, "loss": 0.5254, "mean_token_accuracy": 0.8334659337997437, "num_tokens": 55724062.0, "step": 1467 }, { "epoch": 0.18674468897086885, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 1.8300153017044067, "learning_rate": 6.218736752861383e-06, "loss": 0.4819, "mean_token_accuracy": 0.8497512936592102, "num_tokens": 55752137.0, "step": 1468 }, { "epoch": 0.18687189924945935, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 1.806801199913025, "learning_rate": 6.22297583721916e-06, "loss": 0.5195, "mean_token_accuracy": 0.845474362373352, "num_tokens": 55793171.0, "step": 1469 }, { "epoch": 0.18699910952804988, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.8289523124694824, "learning_rate": 6.227214921576939e-06, "loss": 0.5461, "mean_token_accuracy": 0.8356243371963501, "num_tokens": 55831118.0, "step": 1470 }, { "epoch": 0.18712631980664038, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.7849160432815552, "learning_rate": 6.231454005934718e-06, "loss": 0.4438, "mean_token_accuracy": 0.8603877425193787, "num_tokens": 55865568.0, "step": 1471 }, { "epoch": 0.18725353008523088, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 3.314018249511719e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.8438926935195923, "learning_rate": 6.235693090292497e-06, "loss": 0.5597, "mean_token_accuracy": 0.831348180770874, "num_tokens": 55904624.0, "step": 1472 }, { "epoch": 0.1873807403638214, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.7569336891174316, "learning_rate": 6.239932174650276e-06, "loss": 0.4411, "mean_token_accuracy": 0.8615933656692505, "num_tokens": 55944714.0, "step": 1473 }, { "epoch": 0.1875079506424119, "ewc_loss": 0.02197265625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 1.8371286392211914, "learning_rate": 6.244171259008055e-06, "loss": 0.5256, "mean_token_accuracy": 0.8332604765892029, "num_tokens": 55979444.0, "step": 1474 }, { "epoch": 0.1876351609210024, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 1.766302227973938, "learning_rate": 6.248410343365834e-06, "loss": 0.462, "mean_token_accuracy": 0.8554365038871765, "num_tokens": 56017171.0, "step": 1475 }, { "epoch": 0.18776237119959294, "ewc_loss": 0.022216796875, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.807782530784607, "learning_rate": 6.2526494277236125e-06, "loss": 0.4801, "mean_token_accuracy": 0.8517243266105652, "num_tokens": 56060562.0, "step": 1476 }, { "epoch": 0.18788958147818344, "ewc_loss": 0.021484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 1.8665721416473389, "learning_rate": 6.2568885120813906e-06, "loss": 0.4771, "mean_token_accuracy": 0.8502464294433594, "num_tokens": 56098366.0, "step": 1477 }, { "epoch": 0.18801679175677394, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.7806299924850464, "learning_rate": 6.2611275964391694e-06, "loss": 0.5382, "mean_token_accuracy": 0.8298993110656738, "num_tokens": 56143323.0, "step": 1478 }, { "epoch": 0.18814400203536447, "ewc_loss": 0.021484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 1.7841628789901733, "learning_rate": 6.265366680796948e-06, "loss": 0.4346, "mean_token_accuracy": 0.8613772988319397, "num_tokens": 56183846.0, "step": 1479 }, { "epoch": 0.18827121231395497, "ewc_loss": 0.021484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 1.7746808528900146, "learning_rate": 6.269605765154727e-06, "loss": 0.4934, "mean_token_accuracy": 0.8452792763710022, "num_tokens": 56227412.0, "step": 1480 }, { "epoch": 0.18839842259254547, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.8003870248794556, "learning_rate": 6.273844849512506e-06, "loss": 0.4921, "mean_token_accuracy": 0.8450936079025269, "num_tokens": 56267649.0, "step": 1481 }, { "epoch": 0.188525632871136, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.7652820348739624, "learning_rate": 6.278083933870285e-06, "loss": 0.4208, "mean_token_accuracy": 0.8674147129058838, "num_tokens": 56310802.0, "step": 1482 }, { "epoch": 0.1886528431497265, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 1.820716142654419, "learning_rate": 6.282323018228064e-06, "loss": 0.4606, "mean_token_accuracy": 0.8583096265792847, "num_tokens": 56345972.0, "step": 1483 }, { "epoch": 0.188780053428317, "ewc_loss": 0.021484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 1.7779102325439453, "learning_rate": 6.286562102585841e-06, "loss": 0.5436, "mean_token_accuracy": 0.8342170119285583, "num_tokens": 56391778.0, "step": 1484 }, { "epoch": 0.18890726370690752, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 1.8004260063171387, "learning_rate": 6.29080118694362e-06, "loss": 0.4268, "mean_token_accuracy": 0.862289309501648, "num_tokens": 56426306.0, "step": 1485 }, { "epoch": 0.18903447398549802, "ewc_loss": 0.021484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 1.781360387802124, "learning_rate": 6.295040271301399e-06, "loss": 0.4322, "mean_token_accuracy": 0.8627778887748718, "num_tokens": 56461974.0, "step": 1486 }, { "epoch": 0.18916168426408853, "ewc_loss": 0.021484375, "ewc_loss_diag": 3.361701965332031e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 1.7865715026855469, "learning_rate": 6.299279355659178e-06, "loss": 0.4796, "mean_token_accuracy": 0.8511624336242676, "num_tokens": 56507205.0, "step": 1487 }, { "epoch": 0.18928889454267905, "ewc_loss": 0.021484375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 1.8067630529403687, "learning_rate": 6.303518440016957e-06, "loss": 0.4897, "mean_token_accuracy": 0.8436652421951294, "num_tokens": 56542076.0, "step": 1488 }, { "epoch": 0.18941610482126955, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.8073303699493408, "learning_rate": 6.3077575243747355e-06, "loss": 0.4852, "mean_token_accuracy": 0.851647138595581, "num_tokens": 56575980.0, "step": 1489 }, { "epoch": 0.18954331509986008, "ewc_loss": 0.021484375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 1.7989087104797363, "learning_rate": 6.311996608732514e-06, "loss": 0.5175, "mean_token_accuracy": 0.8418610692024231, "num_tokens": 56613410.0, "step": 1490 }, { "epoch": 0.18967052537845058, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.813132643699646, "learning_rate": 6.316235693090293e-06, "loss": 0.5092, "mean_token_accuracy": 0.838465690612793, "num_tokens": 56649105.0, "step": 1491 }, { "epoch": 0.18979773565704108, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 1.8001312017440796, "learning_rate": 6.320474777448071e-06, "loss": 0.5123, "mean_token_accuracy": 0.8374348878860474, "num_tokens": 56686049.0, "step": 1492 }, { "epoch": 0.1899249459356316, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 1.8042302131652832, "learning_rate": 6.32471386180585e-06, "loss": 0.4754, "mean_token_accuracy": 0.8485931754112244, "num_tokens": 56725659.0, "step": 1493 }, { "epoch": 0.1900521562142221, "ewc_loss": 0.021484375, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.811981201171875e-05, "grad_norm": 1.843288540840149, "learning_rate": 6.328952946163629e-06, "loss": 0.5649, "mean_token_accuracy": 0.8278030753135681, "num_tokens": 56760612.0, "step": 1494 }, { "epoch": 0.1901793664928126, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 1.7852129936218262, "learning_rate": 6.333192030521408e-06, "loss": 0.4412, "mean_token_accuracy": 0.8592027425765991, "num_tokens": 56796081.0, "step": 1495 }, { "epoch": 0.19030657677140314, "ewc_loss": 0.022216796875, "ewc_loss_diag": 3.3855438232421875e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.8087211847305298, "learning_rate": 6.337431114879187e-06, "loss": 0.4964, "mean_token_accuracy": 0.8456922173500061, "num_tokens": 56833477.0, "step": 1496 }, { "epoch": 0.19043378704999364, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 1.7981157302856445, "learning_rate": 6.341670199236966e-06, "loss": 0.476, "mean_token_accuracy": 0.8540756702423096, "num_tokens": 56870898.0, "step": 1497 }, { "epoch": 0.19056099732858414, "ewc_loss": 0.022216796875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.7818522453308105, "learning_rate": 6.345909283594745e-06, "loss": 0.4453, "mean_token_accuracy": 0.8591523170471191, "num_tokens": 56916980.0, "step": 1498 }, { "epoch": 0.19068820760717467, "ewc_loss": 0.0216064453125, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.823902130126953e-05, "grad_norm": 1.8126996755599976, "learning_rate": 6.3501483679525235e-06, "loss": 0.495, "mean_token_accuracy": 0.8401301503181458, "num_tokens": 56952564.0, "step": 1499 }, { "epoch": 0.19081541788576517, "ewc_loss": 0.02197265625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 1.8091866970062256, "learning_rate": 6.354387452310301e-06, "loss": 0.4576, "mean_token_accuracy": 0.8525546193122864, "num_tokens": 56987381.0, "step": 1500 }, { "epoch": 0.19094262816435567, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 1.813370943069458, "learning_rate": 6.3586265366680796e-06, "loss": 0.4633, "mean_token_accuracy": 0.8541685342788696, "num_tokens": 57023402.0, "step": 1501 }, { "epoch": 0.1910698384429462, "ewc_loss": 0.0220947265625, "ewc_loss_diag": 3.409385681152344e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 1.7872288227081299, "learning_rate": 6.3628656210258584e-06, "loss": 0.4551, "mean_token_accuracy": 0.8581708669662476, "num_tokens": 57066464.0, "step": 1502 }, { "epoch": 0.1911970487215367, "ewc_loss": 0.022216796875, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.8052414655685425, "learning_rate": 6.367104705383637e-06, "loss": 0.4713, "mean_token_accuracy": 0.850679874420166, "num_tokens": 57104495.0, "step": 1503 }, { "epoch": 0.1913242590001272, "ewc_loss": 0.02197265625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 1.7912561893463135, "learning_rate": 6.371343789741416e-06, "loss": 0.4888, "mean_token_accuracy": 0.8462063074111938, "num_tokens": 57146246.0, "step": 1504 }, { "epoch": 0.19145146927871773, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 1.8059742450714111, "learning_rate": 6.375582874099195e-06, "loss": 0.4817, "mean_token_accuracy": 0.8473347425460815, "num_tokens": 57184861.0, "step": 1505 }, { "epoch": 0.19157867955730823, "ewc_loss": 0.02197265625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 1.8399937152862549, "learning_rate": 6.379821958456974e-06, "loss": 0.5368, "mean_token_accuracy": 0.8309662342071533, "num_tokens": 57219395.0, "step": 1506 }, { "epoch": 0.19170588983589873, "ewc_loss": 0.02197265625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 1.809612512588501, "learning_rate": 6.384061042814752e-06, "loss": 0.5053, "mean_token_accuracy": 0.8430275321006775, "num_tokens": 57255816.0, "step": 1507 }, { "epoch": 0.19183310011448926, "ewc_loss": 0.022216796875, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.8119548559188843, "learning_rate": 6.388300127172531e-06, "loss": 0.498, "mean_token_accuracy": 0.8458724021911621, "num_tokens": 57296347.0, "step": 1508 }, { "epoch": 0.19196031039307976, "ewc_loss": 0.0218505859375, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.8477439880371094e-05, "grad_norm": 1.7857208251953125, "learning_rate": 6.39253921153031e-06, "loss": 0.4979, "mean_token_accuracy": 0.841367244720459, "num_tokens": 57341789.0, "step": 1509 }, { "epoch": 0.19208752067167026, "ewc_loss": 0.02197265625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 1.8156172037124634, "learning_rate": 6.396778295888089e-06, "loss": 0.4666, "mean_token_accuracy": 0.8559560775756836, "num_tokens": 57379399.0, "step": 1510 }, { "epoch": 0.1922147309502608, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.809280276298523, "learning_rate": 6.4010173802458676e-06, "loss": 0.5033, "mean_token_accuracy": 0.8418142795562744, "num_tokens": 57418936.0, "step": 1511 }, { "epoch": 0.1923419412288513, "ewc_loss": 0.022216796875, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.8016769886016846, "learning_rate": 6.4052564646036464e-06, "loss": 0.4886, "mean_token_accuracy": 0.8448852300643921, "num_tokens": 57461772.0, "step": 1512 }, { "epoch": 0.1924691515074418, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.8043272495269775, "learning_rate": 6.409495548961425e-06, "loss": 0.4662, "mean_token_accuracy": 0.8524062037467957, "num_tokens": 57498388.0, "step": 1513 }, { "epoch": 0.19259636178603232, "ewc_loss": 0.022216796875, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.8223860263824463, "learning_rate": 6.413734633319204e-06, "loss": 0.4754, "mean_token_accuracy": 0.8501409888267517, "num_tokens": 57534123.0, "step": 1514 }, { "epoch": 0.19272357206462282, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.4332275390625e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.8108612298965454, "learning_rate": 6.417973717676982e-06, "loss": 0.4511, "mean_token_accuracy": 0.857940673828125, "num_tokens": 57568351.0, "step": 1515 }, { "epoch": 0.19285078234321335, "ewc_loss": 0.022216796875, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 1.7987873554229736, "learning_rate": 6.422212802034761e-06, "loss": 0.5236, "mean_token_accuracy": 0.8353087902069092, "num_tokens": 57617842.0, "step": 1516 }, { "epoch": 0.19297799262180385, "ewc_loss": 0.02197265625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 1.8199641704559326, "learning_rate": 6.426451886392539e-06, "loss": 0.5407, "mean_token_accuracy": 0.8352968692779541, "num_tokens": 57657658.0, "step": 1517 }, { "epoch": 0.19310520290039435, "ewc_loss": 0.02197265625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 1.8088098764419556, "learning_rate": 6.430690970750318e-06, "loss": 0.4651, "mean_token_accuracy": 0.8532601594924927, "num_tokens": 57695964.0, "step": 1518 }, { "epoch": 0.19323241317898487, "ewc_loss": 0.022216796875, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.8715858459472656e-05, "grad_norm": 1.8555816411972046, "learning_rate": 6.434930055108097e-06, "loss": 0.5149, "mean_token_accuracy": 0.8395823240280151, "num_tokens": 57725722.0, "step": 1519 }, { "epoch": 0.19335962345757537, "ewc_loss": 0.021728515625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.8358230590820312e-05, "grad_norm": 1.8175500631332397, "learning_rate": 6.439169139465876e-06, "loss": 0.5262, "mean_token_accuracy": 0.8342026472091675, "num_tokens": 57761944.0, "step": 1520 }, { "epoch": 0.19348683373616588, "ewc_loss": 0.022216796875, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.8328546285629272, "learning_rate": 6.443408223823655e-06, "loss": 0.4856, "mean_token_accuracy": 0.8457428812980652, "num_tokens": 57796853.0, "step": 1521 }, { "epoch": 0.1936140440147564, "ewc_loss": 0.02197265625, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.8596649169921875e-05, "grad_norm": 1.8046306371688843, "learning_rate": 6.447647308181433e-06, "loss": 0.4875, "mean_token_accuracy": 0.8452941179275513, "num_tokens": 57835295.0, "step": 1522 }, { "epoch": 0.1937412542933469, "ewc_loss": 0.0224609375, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 1.8123623132705688, "learning_rate": 6.451886392539212e-06, "loss": 0.4642, "mean_token_accuracy": 0.8531785607337952, "num_tokens": 57876294.0, "step": 1523 }, { "epoch": 0.1938684645719374, "ewc_loss": 0.022216796875, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.8150056600570679, "learning_rate": 6.4561254768969905e-06, "loss": 0.4949, "mean_token_accuracy": 0.8470984697341919, "num_tokens": 57915332.0, "step": 1524 }, { "epoch": 0.19399567485052793, "ewc_loss": 0.0224609375, "ewc_loss_diag": 3.4809112548828125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 1.8168654441833496, "learning_rate": 6.460364561254769e-06, "loss": 0.4511, "mean_token_accuracy": 0.8582727909088135, "num_tokens": 57955624.0, "step": 1525 }, { "epoch": 0.19412288512911843, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.8182224035263062, "learning_rate": 6.464603645612548e-06, "loss": 0.4449, "mean_token_accuracy": 0.8607815504074097, "num_tokens": 57993692.0, "step": 1526 }, { "epoch": 0.19425009540770893, "ewc_loss": 0.0224609375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 1.8104536533355713, "learning_rate": 6.468842729970327e-06, "loss": 0.4865, "mean_token_accuracy": 0.8514037132263184, "num_tokens": 58035813.0, "step": 1527 }, { "epoch": 0.19437730568629946, "ewc_loss": 0.0224609375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 1.8348162174224854, "learning_rate": 6.473081814328106e-06, "loss": 0.5247, "mean_token_accuracy": 0.8424481153488159, "num_tokens": 58077118.0, "step": 1528 }, { "epoch": 0.19450451596488996, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.8101729154586792, "learning_rate": 6.477320898685885e-06, "loss": 0.4993, "mean_token_accuracy": 0.8444714546203613, "num_tokens": 58120853.0, "step": 1529 }, { "epoch": 0.19463172624348046, "ewc_loss": 0.022705078125, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 1.833161473274231, "learning_rate": 6.481559983043663e-06, "loss": 0.5133, "mean_token_accuracy": 0.8369131088256836, "num_tokens": 58160279.0, "step": 1530 }, { "epoch": 0.194758936522071, "ewc_loss": 0.0224609375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 1.810240626335144, "learning_rate": 6.485799067401442e-06, "loss": 0.4606, "mean_token_accuracy": 0.8577144145965576, "num_tokens": 58200902.0, "step": 1531 }, { "epoch": 0.1948861468006615, "ewc_loss": 0.022705078125, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 1.8317466974258423, "learning_rate": 6.490038151759221e-06, "loss": 0.4246, "mean_token_accuracy": 0.8665709495544434, "num_tokens": 58237179.0, "step": 1532 }, { "epoch": 0.195013357079252, "ewc_loss": 0.0224609375, "ewc_loss_diag": 3.504753112792969e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 1.8168904781341553, "learning_rate": 6.494277236117e-06, "loss": 0.4281, "mean_token_accuracy": 0.8637630939483643, "num_tokens": 58271072.0, "step": 1533 }, { "epoch": 0.19514056735784252, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 1.8301249742507935, "learning_rate": 6.4985163204747785e-06, "loss": 0.5745, "mean_token_accuracy": 0.823850154876709, "num_tokens": 58317186.0, "step": 1534 }, { "epoch": 0.19526777763643302, "ewc_loss": 0.0224609375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 1.8244770765304565, "learning_rate": 6.5027554048325566e-06, "loss": 0.4791, "mean_token_accuracy": 0.8473349809646606, "num_tokens": 58353396.0, "step": 1535 }, { "epoch": 0.19539498791502352, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 1.8596532344818115, "learning_rate": 6.5069944891903354e-06, "loss": 0.5537, "mean_token_accuracy": 0.8334155082702637, "num_tokens": 58392466.0, "step": 1536 }, { "epoch": 0.19552219819361405, "ewc_loss": 0.0223388671875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.8835067749023438e-05, "grad_norm": 1.800161361694336, "learning_rate": 6.511233573548114e-06, "loss": 0.4381, "mean_token_accuracy": 0.8619030714035034, "num_tokens": 58432745.0, "step": 1537 }, { "epoch": 0.19564940847220455, "ewc_loss": 0.022705078125, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 1.8707331418991089, "learning_rate": 6.515472657905892e-06, "loss": 0.4906, "mean_token_accuracy": 0.8454175591468811, "num_tokens": 58464288.0, "step": 1538 }, { "epoch": 0.19577661875079505, "ewc_loss": 0.0224609375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 1.832226276397705, "learning_rate": 6.519711742263671e-06, "loss": 0.5434, "mean_token_accuracy": 0.8327654600143433, "num_tokens": 58501232.0, "step": 1539 }, { "epoch": 0.19590382902938558, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 1.8525041341781616, "learning_rate": 6.52395082662145e-06, "loss": 0.5413, "mean_token_accuracy": 0.8295166492462158, "num_tokens": 58539749.0, "step": 1540 }, { "epoch": 0.19603103930797608, "ewc_loss": 0.0224609375, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.895427703857422e-05, "grad_norm": 1.814704418182373, "learning_rate": 6.528189910979229e-06, "loss": 0.4695, "mean_token_accuracy": 0.8539900183677673, "num_tokens": 58576436.0, "step": 1541 }, { "epoch": 0.1961582495865666, "ewc_loss": 0.02294921875, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 1.8586362600326538, "learning_rate": 6.532428995337008e-06, "loss": 0.4785, "mean_token_accuracy": 0.8513462543487549, "num_tokens": 58618677.0, "step": 1542 }, { "epoch": 0.1962854598651571, "ewc_loss": 0.022705078125, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 1.8262544870376587, "learning_rate": 6.536668079694787e-06, "loss": 0.5012, "mean_token_accuracy": 0.840672492980957, "num_tokens": 58654404.0, "step": 1543 }, { "epoch": 0.1964126701437476, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 1.8380117416381836, "learning_rate": 6.540907164052566e-06, "loss": 0.496, "mean_token_accuracy": 0.8438191413879395, "num_tokens": 58697312.0, "step": 1544 }, { "epoch": 0.19653988042233814, "ewc_loss": 0.0225830078125, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.9073486328125e-05, "grad_norm": 1.8262115716934204, "learning_rate": 6.545146248410344e-06, "loss": 0.504, "mean_token_accuracy": 0.8422524929046631, "num_tokens": 58738940.0, "step": 1545 }, { "epoch": 0.19666709070092864, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 3.528594970703125e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 1.845375657081604, "learning_rate": 6.549385332768123e-06, "loss": 0.5107, "mean_token_accuracy": 0.8433235883712769, "num_tokens": 58779626.0, "step": 1546 }, { "epoch": 0.19679430097951914, "ewc_loss": 0.02294921875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 1.83982515335083, "learning_rate": 6.5536244171259015e-06, "loss": 0.5094, "mean_token_accuracy": 0.8395176529884338, "num_tokens": 58818938.0, "step": 1547 }, { "epoch": 0.19692151125810967, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 1.863358497619629, "learning_rate": 6.55786350148368e-06, "loss": 0.5251, "mean_token_accuracy": 0.834799587726593, "num_tokens": 58855973.0, "step": 1548 }, { "epoch": 0.19704872153670017, "ewc_loss": 0.02294921875, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 1.8263720273971558, "learning_rate": 6.562102585841459e-06, "loss": 0.4659, "mean_token_accuracy": 0.8495834469795227, "num_tokens": 58894624.0, "step": 1549 }, { "epoch": 0.19717593181529067, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 3.552436828613281e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 1.8392648696899414, "learning_rate": 6.566341670199238e-06, "loss": 0.4769, "mean_token_accuracy": 0.8508253693580627, "num_tokens": 58934440.0, "step": 1550 }, { "epoch": 0.1973031420938812, "ewc_loss": 0.022705078125, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.919269561767578e-05, "grad_norm": 1.833522081375122, "learning_rate": 6.570580754557017e-06, "loss": 0.4788, "mean_token_accuracy": 0.8505309820175171, "num_tokens": 58972150.0, "step": 1551 }, { "epoch": 0.1974303523724717, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 1.8434741497039795, "learning_rate": 6.574819838914795e-06, "loss": 0.4554, "mean_token_accuracy": 0.8549175262451172, "num_tokens": 59006179.0, "step": 1552 }, { "epoch": 0.1975575626510622, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 1.8301098346710205, "learning_rate": 6.579058923272573e-06, "loss": 0.5088, "mean_token_accuracy": 0.8398947715759277, "num_tokens": 59046583.0, "step": 1553 }, { "epoch": 0.19768477292965272, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 1.8562804460525513, "learning_rate": 6.583298007630352e-06, "loss": 0.4985, "mean_token_accuracy": 0.8445116281509399, "num_tokens": 59078023.0, "step": 1554 }, { "epoch": 0.19781198320824323, "ewc_loss": 0.02294921875, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 1.8279449939727783, "learning_rate": 6.587537091988131e-06, "loss": 0.5044, "mean_token_accuracy": 0.840949296951294, "num_tokens": 59121160.0, "step": 1555 }, { "epoch": 0.19793919348683373, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 1.8391348123550415, "learning_rate": 6.59177617634591e-06, "loss": 0.4744, "mean_token_accuracy": 0.8483231663703918, "num_tokens": 59164564.0, "step": 1556 }, { "epoch": 0.19806640376542425, "ewc_loss": 0.023193359375, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 1.8259015083312988, "learning_rate": 6.596015260703689e-06, "loss": 0.4776, "mean_token_accuracy": 0.8531683683395386, "num_tokens": 59211558.0, "step": 1557 }, { "epoch": 0.19819361404401475, "ewc_loss": 0.02294921875, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 1.8562076091766357, "learning_rate": 6.6002543450614675e-06, "loss": 0.4717, "mean_token_accuracy": 0.8517881631851196, "num_tokens": 59251260.0, "step": 1558 }, { "epoch": 0.19832082432260525, "ewc_loss": 0.02294921875, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 1.8422566652297974, "learning_rate": 6.604493429419246e-06, "loss": 0.5332, "mean_token_accuracy": 0.8351376056671143, "num_tokens": 59293949.0, "step": 1559 }, { "epoch": 0.19844803460119578, "ewc_loss": 0.02294921875, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 1.8553149700164795, "learning_rate": 6.6087325137770244e-06, "loss": 0.5443, "mean_token_accuracy": 0.8388940691947937, "num_tokens": 59341549.0, "step": 1560 }, { "epoch": 0.19857524487978628, "ewc_loss": 0.0228271484375, "ewc_loss_diag": 3.5762786865234375e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 1.8368948698043823, "learning_rate": 6.612971598134803e-06, "loss": 0.4838, "mean_token_accuracy": 0.8472875356674194, "num_tokens": 59379355.0, "step": 1561 }, { "epoch": 0.19870245515837678, "ewc_loss": 0.02294921875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 1.8721985816955566, "learning_rate": 6.617210682492582e-06, "loss": 0.5307, "mean_token_accuracy": 0.8333768248558044, "num_tokens": 59416340.0, "step": 1562 }, { "epoch": 0.1988296654369673, "ewc_loss": 0.02294921875, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 1.838821291923523, "learning_rate": 6.621449766850361e-06, "loss": 0.4577, "mean_token_accuracy": 0.8559402227401733, "num_tokens": 59453100.0, "step": 1563 }, { "epoch": 0.1989568757155578, "ewc_loss": 0.023193359375, "ewc_loss_diag": 3.600120544433594e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 1.8559675216674805, "learning_rate": 6.62568885120814e-06, "loss": 0.5003, "mean_token_accuracy": 0.8436321020126343, "num_tokens": 59491164.0, "step": 1564 }, { "epoch": 0.19908408599414834, "ewc_loss": 0.02294921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 1.860083818435669, "learning_rate": 6.629927935565919e-06, "loss": 0.5063, "mean_token_accuracy": 0.8421509265899658, "num_tokens": 59525752.0, "step": 1565 }, { "epoch": 0.19921129627273884, "ewc_loss": 0.02294921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 1.856402039527893, "learning_rate": 6.634167019923698e-06, "loss": 0.5435, "mean_token_accuracy": 0.8356860280036926, "num_tokens": 59565206.0, "step": 1566 }, { "epoch": 0.19933850655132934, "ewc_loss": 0.023193359375, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 1.8624402284622192, "learning_rate": 6.638406104281477e-06, "loss": 0.4825, "mean_token_accuracy": 0.8477989435195923, "num_tokens": 59597578.0, "step": 1567 }, { "epoch": 0.19946571682991987, "ewc_loss": 0.02294921875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.9311904907226562e-05, "grad_norm": 1.825919270515442, "learning_rate": 6.642645188639254e-06, "loss": 0.457, "mean_token_accuracy": 0.8552976250648499, "num_tokens": 59640073.0, "step": 1568 }, { "epoch": 0.19959292710851037, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 1.8504451513290405, "learning_rate": 6.646884272997033e-06, "loss": 0.5104, "mean_token_accuracy": 0.83841472864151, "num_tokens": 59680259.0, "step": 1569 }, { "epoch": 0.19972013738710087, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 1.853867769241333, "learning_rate": 6.651123357354812e-06, "loss": 0.4962, "mean_token_accuracy": 0.844011127948761, "num_tokens": 59716372.0, "step": 1570 }, { "epoch": 0.1998473476656914, "ewc_loss": 0.0234375, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 1.8715542554855347, "learning_rate": 6.6553624417125905e-06, "loss": 0.5151, "mean_token_accuracy": 0.8370561599731445, "num_tokens": 59751294.0, "step": 1571 }, { "epoch": 0.1999745579442819, "ewc_loss": 0.023193359375, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 1.837640404701233, "learning_rate": 6.659601526070369e-06, "loss": 0.5483, "mean_token_accuracy": 0.8353050947189331, "num_tokens": 59794395.0, "step": 1572 }, { "epoch": 0.2001017682228724, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 1.907698392868042, "learning_rate": 6.663840610428148e-06, "loss": 0.5742, "mean_token_accuracy": 0.8275759220123291, "num_tokens": 59825503.0, "step": 1573 }, { "epoch": 0.20022897850146293, "ewc_loss": 0.0230712890625, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.9431114196777344e-05, "grad_norm": 1.8452074527740479, "learning_rate": 6.668079694785927e-06, "loss": 0.459, "mean_token_accuracy": 0.8581146001815796, "num_tokens": 59864634.0, "step": 1574 }, { "epoch": 0.20035618878005343, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 1.8825870752334595, "learning_rate": 6.672318779143705e-06, "loss": 0.4785, "mean_token_accuracy": 0.8485272526741028, "num_tokens": 59898693.0, "step": 1575 }, { "epoch": 0.20048339905864393, "ewc_loss": 0.0234375, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 1.865103006362915, "learning_rate": 6.676557863501484e-06, "loss": 0.501, "mean_token_accuracy": 0.8447126746177673, "num_tokens": 59930234.0, "step": 1576 }, { "epoch": 0.20061060933723446, "ewc_loss": 0.0233154296875, "ewc_loss_diag": 3.647804260253906e-05, "ewc_loss_parallel": 1.9669532775878906e-05, "grad_norm": 1.8618745803833008, "learning_rate": 6.680796947859263e-06, "loss": 0.4946, "mean_token_accuracy": 0.8454045057296753, "num_tokens": 59970932.0, "step": 1577 }, { "epoch": 0.20073781961582496, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 1.858310341835022, "learning_rate": 6.685036032217042e-06, "loss": 0.4882, "mean_token_accuracy": 0.8464950323104858, "num_tokens": 60007282.0, "step": 1578 }, { "epoch": 0.20086502989441546, "ewc_loss": 0.023193359375, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.9550323486328125e-05, "grad_norm": 1.8591517210006714, "learning_rate": 6.689275116574821e-06, "loss": 0.4718, "mean_token_accuracy": 0.855445384979248, "num_tokens": 60043336.0, "step": 1579 }, { "epoch": 0.200992240173006, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 1.8506616353988647, "learning_rate": 6.6935142009326e-06, "loss": 0.4996, "mean_token_accuracy": 0.8474187254905701, "num_tokens": 60083284.0, "step": 1580 }, { "epoch": 0.2011194504515965, "ewc_loss": 0.0234375, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 1.8556244373321533, "learning_rate": 6.6977532852903785e-06, "loss": 0.5008, "mean_token_accuracy": 0.8419921398162842, "num_tokens": 60125115.0, "step": 1581 }, { "epoch": 0.201246660730187, "ewc_loss": 0.023681640625, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 1.8569098711013794, "learning_rate": 6.701992369648157e-06, "loss": 0.4178, "mean_token_accuracy": 0.8678036332130432, "num_tokens": 60161497.0, "step": 1582 }, { "epoch": 0.20137387100877752, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 1.8608721494674683, "learning_rate": 6.7062314540059345e-06, "loss": 0.4456, "mean_token_accuracy": 0.8601248860359192, "num_tokens": 60202282.0, "step": 1583 }, { "epoch": 0.20150108128736802, "ewc_loss": 0.023681640625, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 1.8635802268981934, "learning_rate": 6.7104705383637134e-06, "loss": 0.4791, "mean_token_accuracy": 0.8494077324867249, "num_tokens": 60241211.0, "step": 1584 }, { "epoch": 0.20162829156595852, "ewc_loss": 0.0238037109375, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 1.8951977491378784, "learning_rate": 6.714709622721492e-06, "loss": 0.5197, "mean_token_accuracy": 0.8388866186141968, "num_tokens": 60273677.0, "step": 1585 }, { "epoch": 0.20175550184454905, "ewc_loss": 0.023681640625, "ewc_loss_diag": 3.6716461181640625e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 1.8452022075653076, "learning_rate": 6.718948707079271e-06, "loss": 0.4769, "mean_token_accuracy": 0.8511536717414856, "num_tokens": 60313904.0, "step": 1586 }, { "epoch": 0.20188271212313955, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 3.695487976074219e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 1.873579978942871, "learning_rate": 6.72318779143705e-06, "loss": 0.4768, "mean_token_accuracy": 0.8508650660514832, "num_tokens": 60351534.0, "step": 1587 }, { "epoch": 0.20200992240173005, "ewc_loss": 0.0234375, "ewc_loss_diag": 3.695487976074219e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 1.8508777618408203, "learning_rate": 6.727426875794829e-06, "loss": 0.5203, "mean_token_accuracy": 0.8401368260383606, "num_tokens": 60393362.0, "step": 1588 }, { "epoch": 0.20213713268032057, "ewc_loss": 0.02392578125, "ewc_loss_diag": 3.695487976074219e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 1.864973783493042, "learning_rate": 6.731665960152608e-06, "loss": 0.4928, "mean_token_accuracy": 0.8440083265304565, "num_tokens": 60437735.0, "step": 1589 }, { "epoch": 0.20226434295891108, "ewc_loss": 0.0235595703125, "ewc_loss_diag": 3.695487976074219e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 1.8815691471099854, "learning_rate": 6.735905044510387e-06, "loss": 0.551, "mean_token_accuracy": 0.8235282897949219, "num_tokens": 60476032.0, "step": 1590 }, { "epoch": 0.2023915532375016, "ewc_loss": 0.023681640625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 1.8759151697158813, "learning_rate": 6.740144128868165e-06, "loss": 0.5028, "mean_token_accuracy": 0.8434937000274658, "num_tokens": 60513330.0, "step": 1591 }, { "epoch": 0.2025187635160921, "ewc_loss": 0.023681640625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 1.8642328977584839, "learning_rate": 6.744383213225944e-06, "loss": 0.5302, "mean_token_accuracy": 0.8348170518875122, "num_tokens": 60552829.0, "step": 1592 }, { "epoch": 0.2026459737946826, "ewc_loss": 0.0234375, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 1.8748273849487305, "learning_rate": 6.7486222975837225e-06, "loss": 0.4631, "mean_token_accuracy": 0.8511234521865845, "num_tokens": 60585333.0, "step": 1593 }, { "epoch": 0.20277318407327313, "ewc_loss": 0.023681640625, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 2.002716064453125e-05, "grad_norm": 1.8767223358154297, "learning_rate": 6.7528613819415014e-06, "loss": 0.4946, "mean_token_accuracy": 0.8421211242675781, "num_tokens": 60621091.0, "step": 1594 }, { "epoch": 0.20290039435186363, "ewc_loss": 0.0234375, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 1.9788742065429688e-05, "grad_norm": 1.864056944847107, "learning_rate": 6.75710046629928e-06, "loss": 0.4965, "mean_token_accuracy": 0.847262978553772, "num_tokens": 60661602.0, "step": 1595 }, { "epoch": 0.20302760463045413, "ewc_loss": 0.02392578125, "ewc_loss_diag": 3.719329833984375e-05, "ewc_loss_parallel": 2.0265579223632812e-05, "grad_norm": 1.9145781993865967, "learning_rate": 6.761339550657059e-06, "loss": 0.5029, "mean_token_accuracy": 0.839435875415802, "num_tokens": 60689085.0, "step": 1596 }, { "epoch": 0.20315481490904466, "ewc_loss": 0.023681640625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 1.990795135498047e-05, "grad_norm": 1.8541918992996216, "learning_rate": 6.765578635014838e-06, "loss": 0.4974, "mean_token_accuracy": 0.8478367328643799, "num_tokens": 60727259.0, "step": 1597 }, { "epoch": 0.20328202518763516, "ewc_loss": 0.024169921875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 1.8980692625045776, "learning_rate": 6.769817719372615e-06, "loss": 0.4724, "mean_token_accuracy": 0.8515351414680481, "num_tokens": 60761880.0, "step": 1598 }, { "epoch": 0.20340923546622566, "ewc_loss": 0.02392578125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 1.8686811923980713, "learning_rate": 6.774056803730394e-06, "loss": 0.5139, "mean_token_accuracy": 0.8402057886123657, "num_tokens": 60795416.0, "step": 1599 }, { "epoch": 0.2035364457448162, "ewc_loss": 0.024658203125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 1.881366491317749, "learning_rate": 6.778295888088173e-06, "loss": 0.4925, "mean_token_accuracy": 0.8468992710113525, "num_tokens": 60833373.0, "step": 1600 }, { "epoch": 0.2036636560234067, "ewc_loss": 0.024169921875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 1.8723030090332031, "learning_rate": 6.782534972445952e-06, "loss": 0.5072, "mean_token_accuracy": 0.8389226198196411, "num_tokens": 60870024.0, "step": 1601 }, { "epoch": 0.2037908663019972, "ewc_loss": 0.0244140625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 1.89757239818573, "learning_rate": 6.786774056803731e-06, "loss": 0.5006, "mean_token_accuracy": 0.8455121517181396, "num_tokens": 60909790.0, "step": 1602 }, { "epoch": 0.20391807658058772, "ewc_loss": 0.0244140625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 1.8699229955673218, "learning_rate": 6.79101314116151e-06, "loss": 0.5084, "mean_token_accuracy": 0.8412259221076965, "num_tokens": 60948793.0, "step": 1603 }, { "epoch": 0.20404528685917822, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 1.8837050199508667, "learning_rate": 6.795252225519289e-06, "loss": 0.5348, "mean_token_accuracy": 0.8331310749053955, "num_tokens": 60991370.0, "step": 1604 }, { "epoch": 0.20417249713776872, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 1.872687816619873, "learning_rate": 6.7994913098770675e-06, "loss": 0.4995, "mean_token_accuracy": 0.842491865158081, "num_tokens": 61028137.0, "step": 1605 }, { "epoch": 0.20429970741635925, "ewc_loss": 0.0244140625, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 1.9000802040100098, "learning_rate": 6.8037303942348455e-06, "loss": 0.4559, "mean_token_accuracy": 0.8571075201034546, "num_tokens": 61060843.0, "step": 1606 }, { "epoch": 0.20442691769494975, "ewc_loss": 0.02392578125, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 2.014636993408203e-05, "grad_norm": 1.8871488571166992, "learning_rate": 6.807969478592624e-06, "loss": 0.5007, "mean_token_accuracy": 0.8450469374656677, "num_tokens": 61096272.0, "step": 1607 }, { "epoch": 0.20455412797354025, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 3.743171691894531e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 1.8676538467407227, "learning_rate": 6.812208562950403e-06, "loss": 0.5221, "mean_token_accuracy": 0.8440649509429932, "num_tokens": 61139119.0, "step": 1608 }, { "epoch": 0.20468133825213078, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 1.8752655982971191, "learning_rate": 6.816447647308182e-06, "loss": 0.5538, "mean_token_accuracy": 0.8283004760742188, "num_tokens": 61183327.0, "step": 1609 }, { "epoch": 0.20480854853072128, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 1.884448766708374, "learning_rate": 6.820686731665961e-06, "loss": 0.4886, "mean_token_accuracy": 0.8478177785873413, "num_tokens": 61218787.0, "step": 1610 }, { "epoch": 0.20493575880931178, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 1.8867357969284058, "learning_rate": 6.82492581602374e-06, "loss": 0.4946, "mean_token_accuracy": 0.8460983037948608, "num_tokens": 61256517.0, "step": 1611 }, { "epoch": 0.2050629690879023, "ewc_loss": 0.0244140625, "ewc_loss_diag": 3.7670135498046875e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 1.8661915063858032, "learning_rate": 6.829164900381519e-06, "loss": 0.4831, "mean_token_accuracy": 0.8497382402420044, "num_tokens": 61301065.0, "step": 1612 }, { "epoch": 0.2051901793664928, "ewc_loss": 0.024658203125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 1.8921769857406616, "learning_rate": 6.833403984739296e-06, "loss": 0.4875, "mean_token_accuracy": 0.8465508222579956, "num_tokens": 61336368.0, "step": 1613 }, { "epoch": 0.2053173896450833, "ewc_loss": 0.024169921875, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 2.0384788513183594e-05, "grad_norm": 1.8997279405593872, "learning_rate": 6.837643069097075e-06, "loss": 0.5663, "mean_token_accuracy": 0.8244602680206299, "num_tokens": 61373742.0, "step": 1614 }, { "epoch": 0.20544459992367384, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 1.8891557455062866, "learning_rate": 6.841882153454854e-06, "loss": 0.4755, "mean_token_accuracy": 0.8501623868942261, "num_tokens": 61408454.0, "step": 1615 }, { "epoch": 0.20557181020226434, "ewc_loss": 0.0244140625, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 1.8819013833999634, "learning_rate": 6.846121237812633e-06, "loss": 0.4455, "mean_token_accuracy": 0.8604936599731445, "num_tokens": 61445420.0, "step": 1616 }, { "epoch": 0.20569902048085487, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 1.871605396270752, "learning_rate": 6.8503603221704115e-06, "loss": 0.4762, "mean_token_accuracy": 0.8509780764579773, "num_tokens": 61487607.0, "step": 1617 }, { "epoch": 0.20582623075944537, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 1.9149852991104126, "learning_rate": 6.85459940652819e-06, "loss": 0.4825, "mean_token_accuracy": 0.8478959798812866, "num_tokens": 61521053.0, "step": 1618 }, { "epoch": 0.20595344103803587, "ewc_loss": 0.024658203125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 1.873587965965271, "learning_rate": 6.858838490885969e-06, "loss": 0.5422, "mean_token_accuracy": 0.8295085430145264, "num_tokens": 61562211.0, "step": 1619 }, { "epoch": 0.2060806513166264, "ewc_loss": 0.024658203125, "ewc_loss_diag": 3.790855407714844e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 1.9145597219467163, "learning_rate": 6.863077575243748e-06, "loss": 0.5537, "mean_token_accuracy": 0.8329310417175293, "num_tokens": 61600089.0, "step": 1620 }, { "epoch": 0.2062078615952169, "ewc_loss": 0.0242919921875, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 2.0503997802734375e-05, "grad_norm": 1.880868911743164, "learning_rate": 6.867316659601526e-06, "loss": 0.4754, "mean_token_accuracy": 0.8493077754974365, "num_tokens": 61636226.0, "step": 1621 }, { "epoch": 0.2063350718738074, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 1.8941785097122192, "learning_rate": 6.871555743959305e-06, "loss": 0.5035, "mean_token_accuracy": 0.8441159129142761, "num_tokens": 61677224.0, "step": 1622 }, { "epoch": 0.20646228215239792, "ewc_loss": 0.0245361328125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 1.8799314498901367, "learning_rate": 6.875794828317084e-06, "loss": 0.4789, "mean_token_accuracy": 0.8508758544921875, "num_tokens": 61715174.0, "step": 1623 }, { "epoch": 0.20658949243098843, "ewc_loss": 0.02490234375, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 1.8878523111343384, "learning_rate": 6.880033912674863e-06, "loss": 0.4672, "mean_token_accuracy": 0.8554109334945679, "num_tokens": 61760819.0, "step": 1624 }, { "epoch": 0.20671670270957893, "ewc_loss": 0.0244140625, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 2.0623207092285156e-05, "grad_norm": 1.8944449424743652, "learning_rate": 6.884272997032642e-06, "loss": 0.5756, "mean_token_accuracy": 0.8238868117332458, "num_tokens": 61800502.0, "step": 1625 }, { "epoch": 0.20684391298816945, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 1.9021283388137817, "learning_rate": 6.888512081390421e-06, "loss": 0.4612, "mean_token_accuracy": 0.857028603553772, "num_tokens": 61834922.0, "step": 1626 }, { "epoch": 0.20697112326675995, "ewc_loss": 0.024658203125, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 2.086162567138672e-05, "grad_norm": 1.9278684854507446, "learning_rate": 6.8927511657481995e-06, "loss": 0.5199, "mean_token_accuracy": 0.8382523059844971, "num_tokens": 61874675.0, "step": 1627 }, { "epoch": 0.20709833354535045, "ewc_loss": 0.02490234375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 1.8793269395828247, "learning_rate": 6.896990250105978e-06, "loss": 0.4654, "mean_token_accuracy": 0.8573927283287048, "num_tokens": 61914310.0, "step": 1628 }, { "epoch": 0.20722554382394098, "ewc_loss": 0.0247802734375, "ewc_loss_diag": 3.814697265625e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 1.906269907951355, "learning_rate": 6.901229334463756e-06, "loss": 0.4696, "mean_token_accuracy": 0.8498460054397583, "num_tokens": 61949142.0, "step": 1629 }, { "epoch": 0.20735275410253148, "ewc_loss": 0.024658203125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 1.8766955137252808, "learning_rate": 6.9054684188215345e-06, "loss": 0.488, "mean_token_accuracy": 0.853139340877533, "num_tokens": 61989939.0, "step": 1630 }, { "epoch": 0.20747996438112198, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 1.9141209125518799, "learning_rate": 6.909707503179313e-06, "loss": 0.4916, "mean_token_accuracy": 0.8447675704956055, "num_tokens": 62029263.0, "step": 1631 }, { "epoch": 0.2076071746597125, "ewc_loss": 0.024658203125, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.0742416381835938e-05, "grad_norm": 1.8743469715118408, "learning_rate": 6.913946587537092e-06, "loss": 0.4784, "mean_token_accuracy": 0.8514785766601562, "num_tokens": 62067963.0, "step": 1632 }, { "epoch": 0.207734384938303, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 1.927449107170105, "learning_rate": 6.918185671894871e-06, "loss": 0.5621, "mean_token_accuracy": 0.8262515664100647, "num_tokens": 62105313.0, "step": 1633 }, { "epoch": 0.2078615952168935, "ewc_loss": 0.02490234375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 1.9106570482254028, "learning_rate": 6.92242475625265e-06, "loss": 0.5303, "mean_token_accuracy": 0.8469378352165222, "num_tokens": 62140836.0, "step": 1634 }, { "epoch": 0.20798880549548404, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 1.8961377143859863, "learning_rate": 6.926663840610429e-06, "loss": 0.4466, "mean_token_accuracy": 0.8612090349197388, "num_tokens": 62178951.0, "step": 1635 }, { "epoch": 0.20811601577407454, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 1.8978410959243774, "learning_rate": 6.930902924968207e-06, "loss": 0.4832, "mean_token_accuracy": 0.8468406796455383, "num_tokens": 62216941.0, "step": 1636 }, { "epoch": 0.20824322605266504, "ewc_loss": 0.02490234375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 1.911210536956787, "learning_rate": 6.935142009325986e-06, "loss": 0.5203, "mean_token_accuracy": 0.8394648432731628, "num_tokens": 62257676.0, "step": 1637 }, { "epoch": 0.20837043633125557, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 1.9031678438186646, "learning_rate": 6.939381093683765e-06, "loss": 0.5317, "mean_token_accuracy": 0.8357111215591431, "num_tokens": 62296878.0, "step": 1638 }, { "epoch": 0.20849764660984607, "ewc_loss": 0.02490234375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 1.8922723531723022, "learning_rate": 6.943620178041544e-06, "loss": 0.4536, "mean_token_accuracy": 0.8582363724708557, "num_tokens": 62335292.0, "step": 1639 }, { "epoch": 0.2086248568884366, "ewc_loss": 0.02490234375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.110004425048828e-05, "grad_norm": 1.9193729162216187, "learning_rate": 6.9478592623993225e-06, "loss": 0.4938, "mean_token_accuracy": 0.8530323505401611, "num_tokens": 62371182.0, "step": 1640 }, { "epoch": 0.2087520671670271, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.838539123535156e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 1.900424838066101, "learning_rate": 6.952098346757101e-06, "loss": 0.4832, "mean_token_accuracy": 0.8481450080871582, "num_tokens": 62407803.0, "step": 1641 }, { "epoch": 0.2088792774456176, "ewc_loss": 0.02490234375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.09808349609375e-05, "grad_norm": 1.9033657312393188, "learning_rate": 6.95633743111488e-06, "loss": 0.4717, "mean_token_accuracy": 0.8541579246520996, "num_tokens": 62446320.0, "step": 1642 }, { "epoch": 0.20900648772420813, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 1.9089558124542236, "learning_rate": 6.960576515472659e-06, "loss": 0.5162, "mean_token_accuracy": 0.8411210179328918, "num_tokens": 62486919.0, "step": 1643 }, { "epoch": 0.20913369800279863, "ewc_loss": 0.025390625, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 1.9173616170883179, "learning_rate": 6.964815599830436e-06, "loss": 0.4948, "mean_token_accuracy": 0.8490734100341797, "num_tokens": 62521511.0, "step": 1644 }, { "epoch": 0.20926090828138913, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 1.896924614906311, "learning_rate": 6.969054684188215e-06, "loss": 0.5514, "mean_token_accuracy": 0.8252145051956177, "num_tokens": 62564702.0, "step": 1645 }, { "epoch": 0.20938811855997966, "ewc_loss": 0.025390625, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 1.9046900272369385, "learning_rate": 6.973293768545994e-06, "loss": 0.4628, "mean_token_accuracy": 0.8566292524337769, "num_tokens": 62605850.0, "step": 1646 }, { "epoch": 0.20951532883857016, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 1.886620044708252, "learning_rate": 6.977532852903773e-06, "loss": 0.4917, "mean_token_accuracy": 0.8448683619499207, "num_tokens": 62650497.0, "step": 1647 }, { "epoch": 0.20964253911716066, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 1.9389708042144775, "learning_rate": 6.981771937261552e-06, "loss": 0.5448, "mean_token_accuracy": 0.8338335752487183, "num_tokens": 62685538.0, "step": 1648 }, { "epoch": 0.2097697493957512, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 1.9009817838668823, "learning_rate": 6.986011021619331e-06, "loss": 0.5013, "mean_token_accuracy": 0.8455811738967896, "num_tokens": 62725076.0, "step": 1649 }, { "epoch": 0.2098969596743417, "ewc_loss": 0.025634765625, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 1.9427284002304077, "learning_rate": 6.99025010597711e-06, "loss": 0.5187, "mean_token_accuracy": 0.8401114344596863, "num_tokens": 62762335.0, "step": 1650 }, { "epoch": 0.2100241699529322, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 1.8769404888153076, "learning_rate": 6.994489190334888e-06, "loss": 0.4769, "mean_token_accuracy": 0.8504078388214111, "num_tokens": 62804507.0, "step": 1651 }, { "epoch": 0.21015138023152272, "ewc_loss": 0.02587890625, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 1.9199243783950806, "learning_rate": 6.9987282746926666e-06, "loss": 0.4474, "mean_token_accuracy": 0.8610703349113464, "num_tokens": 62844499.0, "step": 1652 }, { "epoch": 0.21027859051011322, "ewc_loss": 0.025390625, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 1.9038728475570679, "learning_rate": 7.0029673590504455e-06, "loss": 0.4889, "mean_token_accuracy": 0.8524706363677979, "num_tokens": 62880061.0, "step": 1653 }, { "epoch": 0.21040580078870372, "ewc_loss": 0.02587890625, "ewc_loss_diag": 3.886222839355469e-05, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 1.9193634986877441, "learning_rate": 7.007206443408224e-06, "loss": 0.4375, "mean_token_accuracy": 0.8618732690811157, "num_tokens": 62916726.0, "step": 1654 }, { "epoch": 0.21053301106729425, "ewc_loss": 0.0252685546875, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 1.8977482318878174, "learning_rate": 7.011445527766003e-06, "loss": 0.4832, "mean_token_accuracy": 0.846699595451355, "num_tokens": 62953341.0, "step": 1655 }, { "epoch": 0.21066022134588475, "ewc_loss": 0.0255126953125, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 1.9131401777267456, "learning_rate": 7.015684612123782e-06, "loss": 0.4768, "mean_token_accuracy": 0.8504464626312256, "num_tokens": 62995047.0, "step": 1656 }, { "epoch": 0.21078743162447525, "ewc_loss": 0.025634765625, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 1.9105490446090698, "learning_rate": 7.019923696481561e-06, "loss": 0.5062, "mean_token_accuracy": 0.843714714050293, "num_tokens": 63039520.0, "step": 1657 }, { "epoch": 0.21091464190306577, "ewc_loss": 0.025634765625, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 1.954450249671936, "learning_rate": 7.02416278083934e-06, "loss": 0.5234, "mean_token_accuracy": 0.833182692527771, "num_tokens": 63071243.0, "step": 1658 }, { "epoch": 0.21104185218165628, "ewc_loss": 0.025390625, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 1.8944123983383179, "learning_rate": 7.028401865197118e-06, "loss": 0.5194, "mean_token_accuracy": 0.8367856740951538, "num_tokens": 63114496.0, "step": 1659 }, { "epoch": 0.21116906246024678, "ewc_loss": 0.0257568359375, "ewc_loss_diag": 3.933906555175781e-05, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 1.9063427448272705, "learning_rate": 7.032640949554897e-06, "loss": 0.4296, "mean_token_accuracy": 0.8613653182983398, "num_tokens": 63158111.0, "step": 1660 }, { "epoch": 0.2112962727388373, "ewc_loss": 0.025390625, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 2.1338462829589844e-05, "grad_norm": 1.9160444736480713, "learning_rate": 7.036880033912676e-06, "loss": 0.5282, "mean_token_accuracy": 0.835873544216156, "num_tokens": 63198305.0, "step": 1661 }, { "epoch": 0.2114234830174278, "ewc_loss": 0.025634765625, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 1.917009949684143, "learning_rate": 7.041119118270454e-06, "loss": 0.4766, "mean_token_accuracy": 0.8501961827278137, "num_tokens": 63235053.0, "step": 1662 }, { "epoch": 0.2115506932960183, "ewc_loss": 0.025634765625, "ewc_loss_diag": 3.9577484130859375e-05, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 1.918806791305542, "learning_rate": 7.045358202628233e-06, "loss": 0.525, "mean_token_accuracy": 0.8401305675506592, "num_tokens": 63275495.0, "step": 1663 }, { "epoch": 0.21167790357460883, "ewc_loss": 0.025634765625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 1.92464017868042, "learning_rate": 7.0495972869860115e-06, "loss": 0.5205, "mean_token_accuracy": 0.8387153148651123, "num_tokens": 63314338.0, "step": 1664 }, { "epoch": 0.21180511385319933, "ewc_loss": 0.025390625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 1.9270864725112915, "learning_rate": 7.05383637134379e-06, "loss": 0.473, "mean_token_accuracy": 0.8516475558280945, "num_tokens": 63348582.0, "step": 1665 }, { "epoch": 0.21193232413178986, "ewc_loss": 0.025390625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 1.930910348892212, "learning_rate": 7.058075455701568e-06, "loss": 0.5172, "mean_token_accuracy": 0.833588719367981, "num_tokens": 63388496.0, "step": 1666 }, { "epoch": 0.21205953441038036, "ewc_loss": 0.025390625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 1.9495434761047363, "learning_rate": 7.062314540059347e-06, "loss": 0.4676, "mean_token_accuracy": 0.8570119142532349, "num_tokens": 63426216.0, "step": 1667 }, { "epoch": 0.21218674468897086, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 1.9151082038879395, "learning_rate": 7.066553624417126e-06, "loss": 0.4689, "mean_token_accuracy": 0.8521719574928284, "num_tokens": 63460627.0, "step": 1668 }, { "epoch": 0.2123139549675614, "ewc_loss": 0.02587890625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 1.9456431865692139, "learning_rate": 7.070792708774905e-06, "loss": 0.5027, "mean_token_accuracy": 0.8435120582580566, "num_tokens": 63501911.0, "step": 1669 }, { "epoch": 0.2124411652461519, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 1.8896490335464478, "learning_rate": 7.075031793132684e-06, "loss": 0.5242, "mean_token_accuracy": 0.8412965536117554, "num_tokens": 63548764.0, "step": 1670 }, { "epoch": 0.2125683755247424, "ewc_loss": 0.026123046875, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 1.9492380619049072, "learning_rate": 7.079270877490463e-06, "loss": 0.4744, "mean_token_accuracy": 0.8517524003982544, "num_tokens": 63583347.0, "step": 1671 }, { "epoch": 0.21269558580333292, "ewc_loss": 0.025146484375, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.1219253540039062e-05, "grad_norm": 1.9079777002334595, "learning_rate": 7.083509961848242e-06, "loss": 0.5541, "mean_token_accuracy": 0.8306704163551331, "num_tokens": 63624115.0, "step": 1672 }, { "epoch": 0.21282279608192342, "ewc_loss": 0.0263671875, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 1.944866418838501, "learning_rate": 7.087749046206021e-06, "loss": 0.4746, "mean_token_accuracy": 0.8564709424972534, "num_tokens": 63660850.0, "step": 1673 }, { "epoch": 0.21295000636051392, "ewc_loss": 0.025390625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.1457672119140625e-05, "grad_norm": 1.909010887145996, "learning_rate": 7.091988130563799e-06, "loss": 0.5193, "mean_token_accuracy": 0.8385133743286133, "num_tokens": 63701530.0, "step": 1674 }, { "epoch": 0.21307721663910445, "ewc_loss": 0.0263671875, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 1.9452908039093018, "learning_rate": 7.0962272149215775e-06, "loss": 0.5173, "mean_token_accuracy": 0.8379671573638916, "num_tokens": 63740400.0, "step": 1675 }, { "epoch": 0.21320442691769495, "ewc_loss": 0.025634765625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.1576881408691406e-05, "grad_norm": 2.020831823348999, "learning_rate": 7.100466299279356e-06, "loss": 0.4987, "mean_token_accuracy": 0.8463350534439087, "num_tokens": 63775019.0, "step": 1676 }, { "epoch": 0.21333163719628545, "ewc_loss": 0.02587890625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 1.9174615144729614, "learning_rate": 7.104705383637135e-06, "loss": 0.5205, "mean_token_accuracy": 0.8424829244613647, "num_tokens": 63813154.0, "step": 1677 }, { "epoch": 0.21345884747487598, "ewc_loss": 0.02587890625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 1.9373043775558472, "learning_rate": 7.108944467994914e-06, "loss": 0.4839, "mean_token_accuracy": 0.8513494729995728, "num_tokens": 63851982.0, "step": 1678 }, { "epoch": 0.21358605775346648, "ewc_loss": 0.025634765625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 1.9276533126831055, "learning_rate": 7.113183552352693e-06, "loss": 0.5527, "mean_token_accuracy": 0.8274615406990051, "num_tokens": 63888983.0, "step": 1679 }, { "epoch": 0.21371326803205698, "ewc_loss": 0.0263671875, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 1.9381983280181885, "learning_rate": 7.117422636710471e-06, "loss": 0.4417, "mean_token_accuracy": 0.8625953197479248, "num_tokens": 63927131.0, "step": 1680 }, { "epoch": 0.2138404783106475, "ewc_loss": 0.02587890625, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 1.9588857889175415, "learning_rate": 7.12166172106825e-06, "loss": 0.5753, "mean_token_accuracy": 0.8205705881118774, "num_tokens": 63962362.0, "step": 1681 }, { "epoch": 0.213967688589238, "ewc_loss": 0.0263671875, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 1.925659418106079, "learning_rate": 7.125900805426028e-06, "loss": 0.4516, "mean_token_accuracy": 0.8611322641372681, "num_tokens": 63997189.0, "step": 1682 }, { "epoch": 0.2140948988678285, "ewc_loss": 0.026123046875, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 1.958691120147705, "learning_rate": 7.130139889783807e-06, "loss": 0.4894, "mean_token_accuracy": 0.8475872874259949, "num_tokens": 64032763.0, "step": 1683 }, { "epoch": 0.21422210914641904, "ewc_loss": 0.026123046875, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 1.918921947479248, "learning_rate": 7.134378974141586e-06, "loss": 0.4843, "mean_token_accuracy": 0.8488472700119019, "num_tokens": 64070176.0, "step": 1684 }, { "epoch": 0.21434931942500954, "ewc_loss": 0.026611328125, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 1.9668070077896118, "learning_rate": 7.138618058499365e-06, "loss": 0.4852, "mean_token_accuracy": 0.8511697053909302, "num_tokens": 64105632.0, "step": 1685 }, { "epoch": 0.21447652970360004, "ewc_loss": 0.026123046875, "ewc_loss_diag": 3.981590270996094e-05, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 1.9010937213897705, "learning_rate": 7.1428571428571436e-06, "loss": 0.4567, "mean_token_accuracy": 0.8597413897514343, "num_tokens": 64148729.0, "step": 1686 }, { "epoch": 0.21460373998219057, "ewc_loss": 0.026611328125, "ewc_loss_diag": 4.00543212890625e-05, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 1.9601627588272095, "learning_rate": 7.1470962272149224e-06, "loss": 0.4825, "mean_token_accuracy": 0.8510452508926392, "num_tokens": 64189466.0, "step": 1687 }, { "epoch": 0.21473095026078107, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 1.9168411493301392, "learning_rate": 7.151335311572701e-06, "loss": 0.4598, "mean_token_accuracy": 0.8567023277282715, "num_tokens": 64226325.0, "step": 1688 }, { "epoch": 0.21485816053937157, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 1.972744107246399, "learning_rate": 7.155574395930479e-06, "loss": 0.5027, "mean_token_accuracy": 0.8413240909576416, "num_tokens": 64262178.0, "step": 1689 }, { "epoch": 0.2149853708179621, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 1.9208592176437378, "learning_rate": 7.159813480288258e-06, "loss": 0.4798, "mean_token_accuracy": 0.847603440284729, "num_tokens": 64299933.0, "step": 1690 }, { "epoch": 0.2151125810965526, "ewc_loss": 0.02685546875, "ewc_loss_diag": 4.029273986816406e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 1.9684582948684692, "learning_rate": 7.164052564646037e-06, "loss": 0.4454, "mean_token_accuracy": 0.8579851388931274, "num_tokens": 64335836.0, "step": 1691 }, { "epoch": 0.21523979137514312, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 4.0531158447265625e-05, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 1.9081618785858154, "learning_rate": 7.168291649003816e-06, "loss": 0.4732, "mean_token_accuracy": 0.8562220931053162, "num_tokens": 64377190.0, "step": 1692 }, { "epoch": 0.21536700165373363, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 4.0531158447265625e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 1.9754818677902222, "learning_rate": 7.172530733361595e-06, "loss": 0.519, "mean_token_accuracy": 0.8400535583496094, "num_tokens": 64414693.0, "step": 1693 }, { "epoch": 0.21549421193232413, "ewc_loss": 0.0260009765625, "ewc_loss_diag": 4.0531158447265625e-05, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 1.9027734994888306, "learning_rate": 7.176769817719374e-06, "loss": 0.4343, "mean_token_accuracy": 0.8703521490097046, "num_tokens": 64454256.0, "step": 1694 }, { "epoch": 0.21562142221091465, "ewc_loss": 0.02685546875, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 1.9775018692016602, "learning_rate": 7.181008902077153e-06, "loss": 0.4896, "mean_token_accuracy": 0.851211667060852, "num_tokens": 64491513.0, "step": 1695 }, { "epoch": 0.21574863248950515, "ewc_loss": 0.026123046875, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 2.193450927734375e-05, "grad_norm": 1.9248359203338623, "learning_rate": 7.1852479864349316e-06, "loss": 0.5089, "mean_token_accuracy": 0.8461453914642334, "num_tokens": 64531349.0, "step": 1696 }, { "epoch": 0.21587584276809565, "ewc_loss": 0.026611328125, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 1.9750843048095703, "learning_rate": 7.189487070792709e-06, "loss": 0.4885, "mean_token_accuracy": 0.8508462905883789, "num_tokens": 64568007.0, "step": 1697 }, { "epoch": 0.21600305304668618, "ewc_loss": 0.02587890625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 1.915908694267273, "learning_rate": 7.193726155150488e-06, "loss": 0.4908, "mean_token_accuracy": 0.8491228222846985, "num_tokens": 64610287.0, "step": 1698 }, { "epoch": 0.21613026332527668, "ewc_loss": 0.02685546875, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 1.9814515113830566, "learning_rate": 7.1979652395082665e-06, "loss": 0.4928, "mean_token_accuracy": 0.8485713005065918, "num_tokens": 64648757.0, "step": 1699 }, { "epoch": 0.21625747360386718, "ewc_loss": 0.02587890625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 2.1696090698242188e-05, "grad_norm": 1.9119703769683838, "learning_rate": 7.202204323866045e-06, "loss": 0.4908, "mean_token_accuracy": 0.85074782371521, "num_tokens": 64690137.0, "step": 1700 }, { "epoch": 0.2163846838824577, "ewc_loss": 0.02685546875, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 1.9762330055236816, "learning_rate": 7.206443408223824e-06, "loss": 0.4889, "mean_token_accuracy": 0.845969021320343, "num_tokens": 64730218.0, "step": 1701 }, { "epoch": 0.2165118941610482, "ewc_loss": 0.02587890625, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 2.181529998779297e-05, "grad_norm": 1.9280389547348022, "learning_rate": 7.210682492581603e-06, "loss": 0.4669, "mean_token_accuracy": 0.8535759449005127, "num_tokens": 64762213.0, "step": 1702 }, { "epoch": 0.2166391044396387, "ewc_loss": 0.02685546875, "ewc_loss_diag": 4.076957702636719e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 1.99734365940094, "learning_rate": 7.214921576939382e-06, "loss": 0.5096, "mean_token_accuracy": 0.8391210436820984, "num_tokens": 64802471.0, "step": 1703 }, { "epoch": 0.21676631471822924, "ewc_loss": 0.0262451171875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.205371856689453e-05, "grad_norm": 1.923405647277832, "learning_rate": 7.21916066129716e-06, "loss": 0.4565, "mean_token_accuracy": 0.8567031025886536, "num_tokens": 64838139.0, "step": 1704 }, { "epoch": 0.21689352499681974, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 1.9684922695159912, "learning_rate": 7.223399745654939e-06, "loss": 0.4975, "mean_token_accuracy": 0.845605194568634, "num_tokens": 64877721.0, "step": 1705 }, { "epoch": 0.21702073527541024, "ewc_loss": 0.0263671875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 1.931854009628296, "learning_rate": 7.227638830012718e-06, "loss": 0.5133, "mean_token_accuracy": 0.8395076990127563, "num_tokens": 64916315.0, "step": 1706 }, { "epoch": 0.21714794555400077, "ewc_loss": 0.027099609375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 1.9876450300216675, "learning_rate": 7.231877914370497e-06, "loss": 0.4685, "mean_token_accuracy": 0.8571164608001709, "num_tokens": 64951303.0, "step": 1707 }, { "epoch": 0.21727515583259127, "ewc_loss": 0.0263671875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2172927856445312e-05, "grad_norm": 1.94390070438385, "learning_rate": 7.236116998728276e-06, "loss": 0.5241, "mean_token_accuracy": 0.8382269144058228, "num_tokens": 64984677.0, "step": 1708 }, { "epoch": 0.21740236611118177, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 1.963319182395935, "learning_rate": 7.2403560830860545e-06, "loss": 0.4158, "mean_token_accuracy": 0.8703505992889404, "num_tokens": 65021960.0, "step": 1709 }, { "epoch": 0.2175295763897723, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 1.9359381198883057, "learning_rate": 7.244595167443833e-06, "loss": 0.5177, "mean_token_accuracy": 0.8393124341964722, "num_tokens": 65065576.0, "step": 1710 }, { "epoch": 0.2176567866683628, "ewc_loss": 0.02685546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 1.9591026306152344, "learning_rate": 7.248834251801612e-06, "loss": 0.4857, "mean_token_accuracy": 0.8535864353179932, "num_tokens": 65107188.0, "step": 1711 }, { "epoch": 0.2177839969469533, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 1.9397398233413696, "learning_rate": 7.2530733361593895e-06, "loss": 0.4558, "mean_token_accuracy": 0.8575535416603088, "num_tokens": 65145740.0, "step": 1712 }, { "epoch": 0.21791120722554383, "ewc_loss": 0.027099609375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 1.9836984872817993, "learning_rate": 7.257312420517168e-06, "loss": 0.4443, "mean_token_accuracy": 0.8599331974983215, "num_tokens": 65176572.0, "step": 1713 }, { "epoch": 0.21803841750413433, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 1.9390126466751099, "learning_rate": 7.261551504874947e-06, "loss": 0.4718, "mean_token_accuracy": 0.8501207828521729, "num_tokens": 65214839.0, "step": 1714 }, { "epoch": 0.21816562778272486, "ewc_loss": 0.02685546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 1.9771778583526611, "learning_rate": 7.265790589232726e-06, "loss": 0.4674, "mean_token_accuracy": 0.8556133508682251, "num_tokens": 65251438.0, "step": 1715 }, { "epoch": 0.21829283806131536, "ewc_loss": 0.026611328125, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 1.9315651655197144, "learning_rate": 7.270029673590505e-06, "loss": 0.5038, "mean_token_accuracy": 0.8447415828704834, "num_tokens": 65294256.0, "step": 1716 }, { "epoch": 0.21842004833990586, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 1.9944112300872803, "learning_rate": 7.274268757948284e-06, "loss": 0.5511, "mean_token_accuracy": 0.8309995532035828, "num_tokens": 65332917.0, "step": 1717 }, { "epoch": 0.2185472586184964, "ewc_loss": 0.026611328125, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2411346435546875e-05, "grad_norm": 1.9446464776992798, "learning_rate": 7.278507842306063e-06, "loss": 0.4627, "mean_token_accuracy": 0.8541655540466309, "num_tokens": 65366678.0, "step": 1718 }, { "epoch": 0.2186744688970869, "ewc_loss": 0.027099609375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 1.9953279495239258, "learning_rate": 7.282746926663842e-06, "loss": 0.5524, "mean_token_accuracy": 0.829998791217804, "num_tokens": 65404797.0, "step": 1719 }, { "epoch": 0.2188016791756774, "ewc_loss": 0.0264892578125, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2292137145996094e-05, "grad_norm": 1.9285783767700195, "learning_rate": 7.28698601102162e-06, "loss": 0.4633, "mean_token_accuracy": 0.854759931564331, "num_tokens": 65441855.0, "step": 1720 }, { "epoch": 0.21892888945426792, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 1.9759536981582642, "learning_rate": 7.291225095379399e-06, "loss": 0.509, "mean_token_accuracy": 0.845632791519165, "num_tokens": 65486895.0, "step": 1721 }, { "epoch": 0.21905609973285842, "ewc_loss": 0.02685546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 1.9393123388290405, "learning_rate": 7.2954641797371775e-06, "loss": 0.4904, "mean_token_accuracy": 0.8503320813179016, "num_tokens": 65524862.0, "step": 1722 }, { "epoch": 0.21918331001144892, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 1.9823180437088013, "learning_rate": 7.299703264094956e-06, "loss": 0.4863, "mean_token_accuracy": 0.8474562168121338, "num_tokens": 65561538.0, "step": 1723 }, { "epoch": 0.21931052029003945, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 1.9442023038864136, "learning_rate": 7.303942348452735e-06, "loss": 0.5064, "mean_token_accuracy": 0.8425847291946411, "num_tokens": 65602376.0, "step": 1724 }, { "epoch": 0.21943773056862995, "ewc_loss": 0.02734375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 1.9736778736114502, "learning_rate": 7.308181432810514e-06, "loss": 0.4534, "mean_token_accuracy": 0.8553340435028076, "num_tokens": 65640817.0, "step": 1725 }, { "epoch": 0.21956494084722045, "ewc_loss": 0.02685546875, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 1.959222435951233, "learning_rate": 7.312420517168293e-06, "loss": 0.4818, "mean_token_accuracy": 0.846147358417511, "num_tokens": 65679238.0, "step": 1726 }, { "epoch": 0.21969215112581097, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 1.967673420906067, "learning_rate": 7.31665960152607e-06, "loss": 0.5051, "mean_token_accuracy": 0.8388041257858276, "num_tokens": 65720141.0, "step": 1727 }, { "epoch": 0.21981936140440148, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 4.124641418457031e-05, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 1.9608598947525024, "learning_rate": 7.320898685883849e-06, "loss": 0.4773, "mean_token_accuracy": 0.8498402833938599, "num_tokens": 65758511.0, "step": 1728 }, { "epoch": 0.21994657168299198, "ewc_loss": 0.02685546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 1.9842363595962524, "learning_rate": 7.325137770241628e-06, "loss": 0.5992, "mean_token_accuracy": 0.8175067901611328, "num_tokens": 65798485.0, "step": 1729 }, { "epoch": 0.2200737819615825, "ewc_loss": 0.02685546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 1.9379674196243286, "learning_rate": 7.329376854599407e-06, "loss": 0.5061, "mean_token_accuracy": 0.8436046838760376, "num_tokens": 65840414.0, "step": 1730 }, { "epoch": 0.220200992240173, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 1.9864789247512817, "learning_rate": 7.333615938957186e-06, "loss": 0.4842, "mean_token_accuracy": 0.8510947823524475, "num_tokens": 65878425.0, "step": 1731 }, { "epoch": 0.2203282025187635, "ewc_loss": 0.0267333984375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.2530555725097656e-05, "grad_norm": 1.9479715824127197, "learning_rate": 7.337855023314965e-06, "loss": 0.4884, "mean_token_accuracy": 0.846603512763977, "num_tokens": 65916958.0, "step": 1732 }, { "epoch": 0.22045541279735403, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 1.983605146408081, "learning_rate": 7.3420941076727435e-06, "loss": 0.5404, "mean_token_accuracy": 0.8315072059631348, "num_tokens": 65958511.0, "step": 1733 }, { "epoch": 0.22058262307594453, "ewc_loss": 0.02685546875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.2649765014648438e-05, "grad_norm": 1.9597725868225098, "learning_rate": 7.346333192030522e-06, "loss": 0.4885, "mean_token_accuracy": 0.8504594564437866, "num_tokens": 65994735.0, "step": 1734 }, { "epoch": 0.22070983335453503, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 1.9899225234985352, "learning_rate": 7.3505722763883004e-06, "loss": 0.5443, "mean_token_accuracy": 0.8300291299819946, "num_tokens": 66036017.0, "step": 1735 }, { "epoch": 0.22083704363312556, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 1.999814748764038, "learning_rate": 7.354811360746079e-06, "loss": 0.4925, "mean_token_accuracy": 0.8456565141677856, "num_tokens": 66073203.0, "step": 1736 }, { "epoch": 0.22096425391171606, "ewc_loss": 0.027099609375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 1.9785456657409668, "learning_rate": 7.359050445103858e-06, "loss": 0.5101, "mean_token_accuracy": 0.8423912525177002, "num_tokens": 66109623.0, "step": 1737 }, { "epoch": 0.22109146419030656, "ewc_loss": 0.0269775390625, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.276897430419922e-05, "grad_norm": 1.9765174388885498, "learning_rate": 7.363289529461637e-06, "loss": 0.4851, "mean_token_accuracy": 0.8501443266868591, "num_tokens": 66148426.0, "step": 1738 }, { "epoch": 0.2212186744688971, "ewc_loss": 0.027099609375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 1.9607698917388916, "learning_rate": 7.367528613819416e-06, "loss": 0.4445, "mean_token_accuracy": 0.8617923259735107, "num_tokens": 66189164.0, "step": 1739 }, { "epoch": 0.2213458847474876, "ewc_loss": 0.027099609375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 1.9746006727218628, "learning_rate": 7.371767698177195e-06, "loss": 0.4621, "mean_token_accuracy": 0.8542687892913818, "num_tokens": 66223523.0, "step": 1740 }, { "epoch": 0.22147309502607812, "ewc_loss": 0.027099609375, "ewc_loss_diag": 4.1484832763671875e-05, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 1.967686653137207, "learning_rate": 7.376006782534974e-06, "loss": 0.4506, "mean_token_accuracy": 0.8609272241592407, "num_tokens": 66256819.0, "step": 1741 }, { "epoch": 0.22160030530466862, "ewc_loss": 0.02734375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 1.987064242362976, "learning_rate": 7.380245866892751e-06, "loss": 0.4803, "mean_token_accuracy": 0.8504362106323242, "num_tokens": 66291619.0, "step": 1742 }, { "epoch": 0.22172751558325912, "ewc_loss": 0.02734375, "ewc_loss_diag": 4.172325134277344e-05, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 1.9525238275527954, "learning_rate": 7.38448495125053e-06, "loss": 0.4836, "mean_token_accuracy": 0.8499370217323303, "num_tokens": 66334034.0, "step": 1743 }, { "epoch": 0.22185472586184965, "ewc_loss": 0.027587890625, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 1.9874528646469116, "learning_rate": 7.388724035608309e-06, "loss": 0.472, "mean_token_accuracy": 0.8567482829093933, "num_tokens": 66368786.0, "step": 1744 }, { "epoch": 0.22198193614044015, "ewc_loss": 0.027099609375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 1.984465479850769, "learning_rate": 7.392963119966088e-06, "loss": 0.4643, "mean_token_accuracy": 0.854239821434021, "num_tokens": 66400034.0, "step": 1745 }, { "epoch": 0.22210914641903065, "ewc_loss": 0.02783203125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 1.9862926006317139, "learning_rate": 7.3972022043238665e-06, "loss": 0.5, "mean_token_accuracy": 0.8475196361541748, "num_tokens": 66436954.0, "step": 1746 }, { "epoch": 0.22223635669762118, "ewc_loss": 0.027587890625, "ewc_loss_diag": 4.1961669921875e-05, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 1.9656260013580322, "learning_rate": 7.401441288681645e-06, "loss": 0.4658, "mean_token_accuracy": 0.8559657335281372, "num_tokens": 66476360.0, "step": 1747 }, { "epoch": 0.22236356697621168, "ewc_loss": 0.027587890625, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 1.973924160003662, "learning_rate": 7.405680373039424e-06, "loss": 0.4448, "mean_token_accuracy": 0.864157497882843, "num_tokens": 66516205.0, "step": 1748 }, { "epoch": 0.22249077725480218, "ewc_loss": 0.02734375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 1.9980390071868896, "learning_rate": 7.409919457397203e-06, "loss": 0.5078, "mean_token_accuracy": 0.840546727180481, "num_tokens": 66548732.0, "step": 1749 }, { "epoch": 0.2226179875333927, "ewc_loss": 0.02734375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 1.9901858568191528, "learning_rate": 7.414158541754981e-06, "loss": 0.5466, "mean_token_accuracy": 0.8272818326950073, "num_tokens": 66587888.0, "step": 1750 }, { "epoch": 0.2227451978119832, "ewc_loss": 0.027587890625, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 2.0061981678009033, "learning_rate": 7.41839762611276e-06, "loss": 0.4819, "mean_token_accuracy": 0.8586956858634949, "num_tokens": 66617846.0, "step": 1751 }, { "epoch": 0.2228724080905737, "ewc_loss": 0.02734375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 1.9670655727386475, "learning_rate": 7.422636710470539e-06, "loss": 0.4954, "mean_token_accuracy": 0.8459208607673645, "num_tokens": 66659703.0, "step": 1752 }, { "epoch": 0.22299961836916424, "ewc_loss": 0.02783203125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 2.006052255630493, "learning_rate": 7.426875794828318e-06, "loss": 0.5093, "mean_token_accuracy": 0.8458389043807983, "num_tokens": 66697604.0, "step": 1753 }, { "epoch": 0.22312682864775474, "ewc_loss": 0.02734375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 1.9777008295059204, "learning_rate": 7.431114879186097e-06, "loss": 0.5372, "mean_token_accuracy": 0.8351269960403442, "num_tokens": 66736839.0, "step": 1754 }, { "epoch": 0.22325403892634524, "ewc_loss": 0.02783203125, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 1.9967501163482666, "learning_rate": 7.435353963543876e-06, "loss": 0.4995, "mean_token_accuracy": 0.8472557663917542, "num_tokens": 66777477.0, "step": 1755 }, { "epoch": 0.22338124920493577, "ewc_loss": 0.027587890625, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 1.9586259126663208, "learning_rate": 7.4395930479016545e-06, "loss": 0.4463, "mean_token_accuracy": 0.8593878149986267, "num_tokens": 66816294.0, "step": 1756 }, { "epoch": 0.22350845948352627, "ewc_loss": 0.028076171875, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 2.023512125015259, "learning_rate": 7.443832132259432e-06, "loss": 0.5119, "mean_token_accuracy": 0.8396633863449097, "num_tokens": 66852740.0, "step": 1757 }, { "epoch": 0.22363566976211677, "ewc_loss": 0.027099609375, "ewc_loss_diag": 4.220008850097656e-05, "ewc_loss_parallel": 2.300739288330078e-05, "grad_norm": 1.9662092924118042, "learning_rate": 7.4480712166172105e-06, "loss": 0.4883, "mean_token_accuracy": 0.8461986184120178, "num_tokens": 66890511.0, "step": 1758 }, { "epoch": 0.2237628800407073, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 2.0931973457336426, "learning_rate": 7.4523103009749894e-06, "loss": 0.4508, "mean_token_accuracy": 0.8579733371734619, "num_tokens": 66924724.0, "step": 1759 }, { "epoch": 0.2238900903192978, "ewc_loss": 0.0272216796875, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 2.288818359375e-05, "grad_norm": 1.9484306573867798, "learning_rate": 7.456549385332768e-06, "loss": 0.5305, "mean_token_accuracy": 0.8372554183006287, "num_tokens": 66964102.0, "step": 1760 }, { "epoch": 0.2240173005978883, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.0248963832855225, "learning_rate": 7.460788469690547e-06, "loss": 0.414, "mean_token_accuracy": 0.8699199557304382, "num_tokens": 67003234.0, "step": 1761 }, { "epoch": 0.22414451087647883, "ewc_loss": 0.02734375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 2.3126602172851562e-05, "grad_norm": 1.9515764713287354, "learning_rate": 7.465027554048326e-06, "loss": 0.5689, "mean_token_accuracy": 0.8234261870384216, "num_tokens": 67045952.0, "step": 1762 }, { "epoch": 0.22427172115506933, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 2.0307464599609375, "learning_rate": 7.469266638406105e-06, "loss": 0.5537, "mean_token_accuracy": 0.8310378789901733, "num_tokens": 67090169.0, "step": 1763 }, { "epoch": 0.22439893143365983, "ewc_loss": 0.0274658203125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 2.3245811462402344e-05, "grad_norm": 1.9613234996795654, "learning_rate": 7.473505722763884e-06, "loss": 0.4726, "mean_token_accuracy": 0.852080762386322, "num_tokens": 67125075.0, "step": 1764 }, { "epoch": 0.22452614171225035, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 2.041841983795166, "learning_rate": 7.477744807121662e-06, "loss": 0.519, "mean_token_accuracy": 0.8402878046035767, "num_tokens": 67162892.0, "step": 1765 }, { "epoch": 0.22465335199084085, "ewc_loss": 0.027587890625, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 2.3365020751953125e-05, "grad_norm": 1.9871934652328491, "learning_rate": 7.481983891479441e-06, "loss": 0.5, "mean_token_accuracy": 0.8463594913482666, "num_tokens": 67198324.0, "step": 1766 }, { "epoch": 0.22478056226943138, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.020780563354492, "learning_rate": 7.48622297583722e-06, "loss": 0.483, "mean_token_accuracy": 0.8472558259963989, "num_tokens": 67239548.0, "step": 1767 }, { "epoch": 0.22490777254802188, "ewc_loss": 0.02783203125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 1.9835683107376099, "learning_rate": 7.4904620601949985e-06, "loss": 0.4475, "mean_token_accuracy": 0.8576175570487976, "num_tokens": 67273373.0, "step": 1768 }, { "epoch": 0.22503498282661238, "ewc_loss": 0.0283203125, "ewc_loss_diag": 4.2438507080078125e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 2.0321524143218994, "learning_rate": 7.4947011445527774e-06, "loss": 0.5269, "mean_token_accuracy": 0.8374007940292358, "num_tokens": 67307074.0, "step": 1769 }, { "epoch": 0.2251621931052029, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 1.9764984846115112, "learning_rate": 7.498940228910556e-06, "loss": 0.5041, "mean_token_accuracy": 0.8415560126304626, "num_tokens": 67347905.0, "step": 1770 }, { "epoch": 0.2252894033837934, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.029646396636963, "learning_rate": 7.503179313268335e-06, "loss": 0.4758, "mean_token_accuracy": 0.8489432334899902, "num_tokens": 67382594.0, "step": 1771 }, { "epoch": 0.2254166136623839, "ewc_loss": 0.0277099609375, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 2.3484230041503906e-05, "grad_norm": 1.9711660146713257, "learning_rate": 7.507418397626114e-06, "loss": 0.5205, "mean_token_accuracy": 0.8378697633743286, "num_tokens": 67422867.0, "step": 1772 }, { "epoch": 0.22554382394097444, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 2.0179250240325928, "learning_rate": 7.511657481983891e-06, "loss": 0.5076, "mean_token_accuracy": 0.8402306437492371, "num_tokens": 67463524.0, "step": 1773 }, { "epoch": 0.22567103421956494, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 4.291534423828125e-05, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 1.9793670177459717, "learning_rate": 7.51589656634167e-06, "loss": 0.4643, "mean_token_accuracy": 0.849419116973877, "num_tokens": 67499224.0, "step": 1774 }, { "epoch": 0.22579824449815544, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.0263655185699463, "learning_rate": 7.520135650699449e-06, "loss": 0.4938, "mean_token_accuracy": 0.844022810459137, "num_tokens": 67536745.0, "step": 1775 }, { "epoch": 0.22592545477674597, "ewc_loss": 0.02783203125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 2.3603439331054688e-05, "grad_norm": 1.9971144199371338, "learning_rate": 7.524374735057228e-06, "loss": 0.5176, "mean_token_accuracy": 0.8437838554382324, "num_tokens": 67572725.0, "step": 1776 }, { "epoch": 0.22605266505533647, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 2.024815559387207, "learning_rate": 7.528613819415007e-06, "loss": 0.4741, "mean_token_accuracy": 0.8575283288955688, "num_tokens": 67607588.0, "step": 1777 }, { "epoch": 0.22617987533392697, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 1.995893120765686, "learning_rate": 7.532852903772786e-06, "loss": 0.5076, "mean_token_accuracy": 0.8423904180526733, "num_tokens": 67643064.0, "step": 1778 }, { "epoch": 0.2263070856125175, "ewc_loss": 0.0283203125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 2.0335750579833984, "learning_rate": 7.537091988130565e-06, "loss": 0.5099, "mean_token_accuracy": 0.8393579721450806, "num_tokens": 67677985.0, "step": 1779 }, { "epoch": 0.226434295891108, "ewc_loss": 0.0279541015625, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 1.9916865825653076, "learning_rate": 7.541331072488343e-06, "loss": 0.4973, "mean_token_accuracy": 0.8458591103553772, "num_tokens": 67713814.0, "step": 1780 }, { "epoch": 0.2265615061696985, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.0208911895751953, "learning_rate": 7.5455701568461215e-06, "loss": 0.4977, "mean_token_accuracy": 0.8456447124481201, "num_tokens": 67752806.0, "step": 1781 }, { "epoch": 0.22668871644828903, "ewc_loss": 0.028076171875, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 1.9858739376068115, "learning_rate": 7.5498092412039e-06, "loss": 0.4814, "mean_token_accuracy": 0.8515069484710693, "num_tokens": 67792804.0, "step": 1782 }, { "epoch": 0.22681592672687953, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 4.315376281738281e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.011711359024048, "learning_rate": 7.554048325561679e-06, "loss": 0.4444, "mean_token_accuracy": 0.8613329529762268, "num_tokens": 67831361.0, "step": 1783 }, { "epoch": 0.22694313700547003, "ewc_loss": 0.028076171875, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 2.017047882080078, "learning_rate": 7.558287409919458e-06, "loss": 0.5344, "mean_token_accuracy": 0.8402211666107178, "num_tokens": 67868981.0, "step": 1784 }, { "epoch": 0.22707034728406056, "ewc_loss": 0.0283203125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 2.0224356651306152, "learning_rate": 7.562526494277237e-06, "loss": 0.5133, "mean_token_accuracy": 0.8380342125892639, "num_tokens": 67905501.0, "step": 1785 }, { "epoch": 0.22719755756265106, "ewc_loss": 0.028076171875, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 1.9942727088928223, "learning_rate": 7.566765578635016e-06, "loss": 0.4818, "mean_token_accuracy": 0.8528806567192078, "num_tokens": 67946001.0, "step": 1786 }, { "epoch": 0.22732476784124156, "ewc_loss": 0.0283203125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 2.0129892826080322, "learning_rate": 7.571004662992795e-06, "loss": 0.5045, "mean_token_accuracy": 0.8467682600021362, "num_tokens": 67983868.0, "step": 1787 }, { "epoch": 0.2274519781198321, "ewc_loss": 0.028076171875, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 2.0020108222961426, "learning_rate": 7.575243747350573e-06, "loss": 0.5602, "mean_token_accuracy": 0.8277537226676941, "num_tokens": 68023664.0, "step": 1788 }, { "epoch": 0.2275791883984226, "ewc_loss": 0.0283203125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 2.006448984146118, "learning_rate": 7.579482831708351e-06, "loss": 0.4979, "mean_token_accuracy": 0.8479939103126526, "num_tokens": 68063950.0, "step": 1789 }, { "epoch": 0.2277063986770131, "ewc_loss": 0.0283203125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 2.016961097717285, "learning_rate": 7.58372191606613e-06, "loss": 0.496, "mean_token_accuracy": 0.8467650413513184, "num_tokens": 68097812.0, "step": 1790 }, { "epoch": 0.22783360895560362, "ewc_loss": 0.0283203125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 2.0105857849121094, "learning_rate": 7.587961000423909e-06, "loss": 0.4966, "mean_token_accuracy": 0.8460653424263, "num_tokens": 68135219.0, "step": 1791 }, { "epoch": 0.22796081923419412, "ewc_loss": 0.028564453125, "ewc_loss_diag": 4.3392181396484375e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.0358614921569824, "learning_rate": 7.5922000847816875e-06, "loss": 0.5495, "mean_token_accuracy": 0.8347948789596558, "num_tokens": 68169757.0, "step": 1792 }, { "epoch": 0.22808802951278465, "ewc_loss": 0.028076171875, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 2.0118520259857178, "learning_rate": 7.5964391691394664e-06, "loss": 0.4938, "mean_token_accuracy": 0.8465367555618286, "num_tokens": 68204807.0, "step": 1793 }, { "epoch": 0.22821523979137515, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 2.01629900932312, "learning_rate": 7.600678253497245e-06, "loss": 0.4896, "mean_token_accuracy": 0.8516702651977539, "num_tokens": 68245037.0, "step": 1794 }, { "epoch": 0.22834245006996565, "ewc_loss": 0.028076171875, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.372264862060547e-05, "grad_norm": 1.9994531869888306, "learning_rate": 7.604917337855023e-06, "loss": 0.4604, "mean_token_accuracy": 0.856503963470459, "num_tokens": 68281938.0, "step": 1795 }, { "epoch": 0.22846966034855618, "ewc_loss": 0.028564453125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.015740394592285, "learning_rate": 7.609156422212802e-06, "loss": 0.4444, "mean_token_accuracy": 0.8618041276931763, "num_tokens": 68319363.0, "step": 1796 }, { "epoch": 0.22859687062714668, "ewc_loss": 0.028564453125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 1.9993945360183716, "learning_rate": 7.613395506570581e-06, "loss": 0.4684, "mean_token_accuracy": 0.8548779487609863, "num_tokens": 68357428.0, "step": 1797 }, { "epoch": 0.22872408090573718, "ewc_loss": 0.028564453125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.0397307872772217, "learning_rate": 7.61763459092836e-06, "loss": 0.4735, "mean_token_accuracy": 0.853420615196228, "num_tokens": 68390558.0, "step": 1798 }, { "epoch": 0.2288512911843277, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 1.9992256164550781, "learning_rate": 7.621873675286139e-06, "loss": 0.5087, "mean_token_accuracy": 0.8433102369308472, "num_tokens": 68432250.0, "step": 1799 }, { "epoch": 0.2289785014629182, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 2.0402214527130127, "learning_rate": 7.626112759643918e-06, "loss": 0.5098, "mean_token_accuracy": 0.842806339263916, "num_tokens": 68469427.0, "step": 1800 }, { "epoch": 0.2291057117415087, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 2.0105483531951904, "learning_rate": 7.630351844001696e-06, "loss": 0.4869, "mean_token_accuracy": 0.8499265313148499, "num_tokens": 68503442.0, "step": 1801 }, { "epoch": 0.22923292202009923, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 2.027507781982422, "learning_rate": 7.634590928359475e-06, "loss": 0.472, "mean_token_accuracy": 0.8557039499282837, "num_tokens": 68540836.0, "step": 1802 }, { "epoch": 0.22936013229868973, "ewc_loss": 0.0281982421875, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.384185791015625e-05, "grad_norm": 1.9969439506530762, "learning_rate": 7.638830012717254e-06, "loss": 0.5065, "mean_token_accuracy": 0.841219961643219, "num_tokens": 68581118.0, "step": 1803 }, { "epoch": 0.22948734257728023, "ewc_loss": 0.029052734375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 2.02852725982666, "learning_rate": 7.643069097075032e-06, "loss": 0.4763, "mean_token_accuracy": 0.8508021831512451, "num_tokens": 68620988.0, "step": 1804 }, { "epoch": 0.22961455285587076, "ewc_loss": 0.0283203125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 2.0631139278411865, "learning_rate": 7.647308181432811e-06, "loss": 0.4487, "mean_token_accuracy": 0.8582140207290649, "num_tokens": 68662141.0, "step": 1805 }, { "epoch": 0.22974176313446126, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 2.02630615234375, "learning_rate": 7.65154726579059e-06, "loss": 0.5347, "mean_token_accuracy": 0.8353098630905151, "num_tokens": 68700922.0, "step": 1806 }, { "epoch": 0.22986897341305176, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 2.030538320541382, "learning_rate": 7.655786350148369e-06, "loss": 0.5115, "mean_token_accuracy": 0.842662513256073, "num_tokens": 68741137.0, "step": 1807 }, { "epoch": 0.2299961836916423, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 2.0101194381713867, "learning_rate": 7.660025434506148e-06, "loss": 0.5122, "mean_token_accuracy": 0.8427402973175049, "num_tokens": 68785679.0, "step": 1808 }, { "epoch": 0.2301233939702328, "ewc_loss": 0.02880859375, "ewc_loss_diag": 4.363059997558594e-05, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 2.0292296409606934, "learning_rate": 7.664264518863927e-06, "loss": 0.4938, "mean_token_accuracy": 0.8484190702438354, "num_tokens": 68823482.0, "step": 1809 }, { "epoch": 0.2302506042488233, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 2.0193228721618652, "learning_rate": 7.668503603221706e-06, "loss": 0.5511, "mean_token_accuracy": 0.8274661302566528, "num_tokens": 68863854.0, "step": 1810 }, { "epoch": 0.23037781452741382, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 2.0191192626953125, "learning_rate": 7.672742687579483e-06, "loss": 0.4458, "mean_token_accuracy": 0.859027087688446, "num_tokens": 68901341.0, "step": 1811 }, { "epoch": 0.23050502480600432, "ewc_loss": 0.02880859375, "ewc_loss_diag": 4.38690185546875e-05, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 2.041144847869873, "learning_rate": 7.676981771937262e-06, "loss": 0.4256, "mean_token_accuracy": 0.8634551763534546, "num_tokens": 68931774.0, "step": 1812 }, { "epoch": 0.23063223508459482, "ewc_loss": 0.028564453125, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.004765033721924, "learning_rate": 7.68122085629504e-06, "loss": 0.4646, "mean_token_accuracy": 0.8542909622192383, "num_tokens": 68972114.0, "step": 1813 }, { "epoch": 0.23075944536318535, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 2.043443202972412, "learning_rate": 7.68545994065282e-06, "loss": 0.4874, "mean_token_accuracy": 0.8467977643013, "num_tokens": 69007272.0, "step": 1814 }, { "epoch": 0.23088665564177585, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 1.9995535612106323, "learning_rate": 7.689699025010598e-06, "loss": 0.4821, "mean_token_accuracy": 0.8498742580413818, "num_tokens": 69050011.0, "step": 1815 }, { "epoch": 0.23101386592036638, "ewc_loss": 0.029052734375, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 2.0573513507843018, "learning_rate": 7.693938109368377e-06, "loss": 0.5035, "mean_token_accuracy": 0.8416939973831177, "num_tokens": 69087914.0, "step": 1816 }, { "epoch": 0.23114107619895688, "ewc_loss": 0.028564453125, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.010059356689453, "learning_rate": 7.698177193726156e-06, "loss": 0.4935, "mean_token_accuracy": 0.8450789451599121, "num_tokens": 69125301.0, "step": 1817 }, { "epoch": 0.23126828647754738, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 2.0615506172180176, "learning_rate": 7.702416278083933e-06, "loss": 0.5074, "mean_token_accuracy": 0.843010663986206, "num_tokens": 69159268.0, "step": 1818 }, { "epoch": 0.2313954967561379, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 1.9927361011505127, "learning_rate": 7.706655362441712e-06, "loss": 0.4865, "mean_token_accuracy": 0.8497405052185059, "num_tokens": 69202207.0, "step": 1819 }, { "epoch": 0.2315227070347284, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 2.069328784942627, "learning_rate": 7.710894446799491e-06, "loss": 0.4638, "mean_token_accuracy": 0.8534941673278809, "num_tokens": 69234757.0, "step": 1820 }, { "epoch": 0.2316499173133189, "ewc_loss": 0.0283203125, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 1.9900152683258057, "learning_rate": 7.71513353115727e-06, "loss": 0.467, "mean_token_accuracy": 0.8589240312576294, "num_tokens": 69274995.0, "step": 1821 }, { "epoch": 0.23177712759190944, "ewc_loss": 0.029541015625, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 2.071441888809204, "learning_rate": 7.719372615515049e-06, "loss": 0.5331, "mean_token_accuracy": 0.8355202078819275, "num_tokens": 69318706.0, "step": 1822 }, { "epoch": 0.23190433787049994, "ewc_loss": 0.0283203125, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 2.396106719970703e-05, "grad_norm": 1.9871717691421509, "learning_rate": 7.723611699872828e-06, "loss": 0.4803, "mean_token_accuracy": 0.8491742014884949, "num_tokens": 69361895.0, "step": 1823 }, { "epoch": 0.23203154814909044, "ewc_loss": 0.0301513671875, "ewc_loss_diag": 4.410743713378906e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.088702917098999, "learning_rate": 7.727850784230607e-06, "loss": 0.4993, "mean_token_accuracy": 0.8424242734909058, "num_tokens": 69400877.0, "step": 1824 }, { "epoch": 0.23215875842768097, "ewc_loss": 0.0284423828125, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 2.4080276489257812e-05, "grad_norm": 2.001746416091919, "learning_rate": 7.732089868588386e-06, "loss": 0.5215, "mean_token_accuracy": 0.8421801328659058, "num_tokens": 69437875.0, "step": 1825 }, { "epoch": 0.23228596870627147, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 2.06520938873291, "learning_rate": 7.736328952946165e-06, "loss": 0.4359, "mean_token_accuracy": 0.8653350472450256, "num_tokens": 69481770.0, "step": 1826 }, { "epoch": 0.23241317898486197, "ewc_loss": 0.028564453125, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.0134565830230713, "learning_rate": 7.740568037303943e-06, "loss": 0.4574, "mean_token_accuracy": 0.8544983863830566, "num_tokens": 69514643.0, "step": 1827 }, { "epoch": 0.2325403892634525, "ewc_loss": 0.030029296875, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 2.0585060119628906, "learning_rate": 7.744807121661722e-06, "loss": 0.426, "mean_token_accuracy": 0.870572566986084, "num_tokens": 69551906.0, "step": 1828 }, { "epoch": 0.232667599542043, "ewc_loss": 0.0286865234375, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 2.0161166191101074, "learning_rate": 7.749046206019501e-06, "loss": 0.497, "mean_token_accuracy": 0.8494295477867126, "num_tokens": 69593621.0, "step": 1829 }, { "epoch": 0.2327948098206335, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 2.064631700515747, "learning_rate": 7.75328529037728e-06, "loss": 0.4345, "mean_token_accuracy": 0.864751935005188, "num_tokens": 69629586.0, "step": 1830 }, { "epoch": 0.23292202009922403, "ewc_loss": 0.02880859375, "ewc_loss_diag": 4.4345855712890625e-05, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 2.0310661792755127, "learning_rate": 7.757524374735059e-06, "loss": 0.5367, "mean_token_accuracy": 0.8334258198738098, "num_tokens": 69667409.0, "step": 1831 }, { "epoch": 0.23304923037781453, "ewc_loss": 0.029296875, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.0530529022216797, "learning_rate": 7.761763459092838e-06, "loss": 0.4448, "mean_token_accuracy": 0.860264003276825, "num_tokens": 69704645.0, "step": 1832 }, { "epoch": 0.23317644065640503, "ewc_loss": 0.02880859375, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 2.0218749046325684, "learning_rate": 7.766002543450615e-06, "loss": 0.474, "mean_token_accuracy": 0.8517978191375732, "num_tokens": 69742216.0, "step": 1833 }, { "epoch": 0.23330365093499555, "ewc_loss": 0.029541015625, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 2.0564582347869873, "learning_rate": 7.770241627808394e-06, "loss": 0.4376, "mean_token_accuracy": 0.8622555732727051, "num_tokens": 69779119.0, "step": 1834 }, { "epoch": 0.23343086121358606, "ewc_loss": 0.02880859375, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 2.0258865356445312, "learning_rate": 7.774480712166173e-06, "loss": 0.5135, "mean_token_accuracy": 0.8407641649246216, "num_tokens": 69819980.0, "step": 1835 }, { "epoch": 0.23355807149217656, "ewc_loss": 0.029541015625, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 2.0426995754241943, "learning_rate": 7.778719796523952e-06, "loss": 0.4943, "mean_token_accuracy": 0.8446447849273682, "num_tokens": 69859840.0, "step": 1836 }, { "epoch": 0.23368528177076708, "ewc_loss": 0.028564453125, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 2.4199485778808594e-05, "grad_norm": 2.0445919036865234, "learning_rate": 7.78295888088173e-06, "loss": 0.4823, "mean_token_accuracy": 0.8461202383041382, "num_tokens": 69893446.0, "step": 1837 }, { "epoch": 0.23381249204935758, "ewc_loss": 0.029296875, "ewc_loss_diag": 4.458427429199219e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.0321290493011475, "learning_rate": 7.78719796523951e-06, "loss": 0.4459, "mean_token_accuracy": 0.8580881357192993, "num_tokens": 69930258.0, "step": 1838 }, { "epoch": 0.23393970232794808, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 2.0767295360565186, "learning_rate": 7.791437049597288e-06, "loss": 0.5519, "mean_token_accuracy": 0.830604612827301, "num_tokens": 69960691.0, "step": 1839 }, { "epoch": 0.2340669126065386, "ewc_loss": 0.029296875, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 2.025686740875244, "learning_rate": 7.795676133955067e-06, "loss": 0.5016, "mean_token_accuracy": 0.843887209892273, "num_tokens": 69998503.0, "step": 1840 }, { "epoch": 0.2341941228851291, "ewc_loss": 0.029541015625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 2.0782864093780518, "learning_rate": 7.799915218312844e-06, "loss": 0.499, "mean_token_accuracy": 0.8443391919136047, "num_tokens": 70032381.0, "step": 1841 }, { "epoch": 0.23432133316371964, "ewc_loss": 0.02880859375, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 2.4318695068359375e-05, "grad_norm": 2.0106616020202637, "learning_rate": 7.804154302670623e-06, "loss": 0.4591, "mean_token_accuracy": 0.8575118780136108, "num_tokens": 70070616.0, "step": 1842 }, { "epoch": 0.23444854344231014, "ewc_loss": 0.030029296875, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 2.0650641918182373, "learning_rate": 7.808393387028402e-06, "loss": 0.4831, "mean_token_accuracy": 0.8481691479682922, "num_tokens": 70111709.0, "step": 1843 }, { "epoch": 0.23457575372090064, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 4.482269287109375e-05, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 2.0284321308135986, "learning_rate": 7.812632471386181e-06, "loss": 0.5284, "mean_token_accuracy": 0.8334488868713379, "num_tokens": 70148516.0, "step": 1844 }, { "epoch": 0.23470296399949117, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 2.072861909866333, "learning_rate": 7.81687155574396e-06, "loss": 0.4934, "mean_token_accuracy": 0.8470090627670288, "num_tokens": 70185688.0, "step": 1845 }, { "epoch": 0.23483017427808167, "ewc_loss": 0.029052734375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 2.01603627204895, "learning_rate": 7.821110640101739e-06, "loss": 0.5095, "mean_token_accuracy": 0.8397254943847656, "num_tokens": 70228901.0, "step": 1846 }, { "epoch": 0.23495738455667217, "ewc_loss": 0.0302734375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.074413537979126, "learning_rate": 7.825349724459518e-06, "loss": 0.4916, "mean_token_accuracy": 0.8489623665809631, "num_tokens": 70268341.0, "step": 1847 }, { "epoch": 0.2350845948352627, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 2.0268685817718506, "learning_rate": 7.829588808817295e-06, "loss": 0.4865, "mean_token_accuracy": 0.8474460244178772, "num_tokens": 70309285.0, "step": 1848 }, { "epoch": 0.2352118051138532, "ewc_loss": 0.0302734375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.103245973587036, "learning_rate": 7.833827893175074e-06, "loss": 0.5204, "mean_token_accuracy": 0.8357397317886353, "num_tokens": 70339694.0, "step": 1849 }, { "epoch": 0.2353390153924437, "ewc_loss": 0.0289306640625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.4437904357910156e-05, "grad_norm": 2.024498462677002, "learning_rate": 7.838066977532853e-06, "loss": 0.4806, "mean_token_accuracy": 0.8506453037261963, "num_tokens": 70377684.0, "step": 1850 }, { "epoch": 0.23546622567103423, "ewc_loss": 0.030517578125, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 2.082754611968994, "learning_rate": 7.842306061890632e-06, "loss": 0.5193, "mean_token_accuracy": 0.8410196304321289, "num_tokens": 70417690.0, "step": 1851 }, { "epoch": 0.23559343594962473, "ewc_loss": 0.029052734375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.4557113647460938e-05, "grad_norm": 2.0323646068573, "learning_rate": 7.84654514624841e-06, "loss": 0.5069, "mean_token_accuracy": 0.8412713408470154, "num_tokens": 70454485.0, "step": 1852 }, { "epoch": 0.23572064622821523, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 2.1097571849823, "learning_rate": 7.85078423060619e-06, "loss": 0.5322, "mean_token_accuracy": 0.8396432399749756, "num_tokens": 70490337.0, "step": 1853 }, { "epoch": 0.23584785650680576, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 1.9990452527999878, "learning_rate": 7.855023314963968e-06, "loss": 0.5089, "mean_token_accuracy": 0.845186710357666, "num_tokens": 70535194.0, "step": 1854 }, { "epoch": 0.23597506678539626, "ewc_loss": 0.031005859375, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 2.102301597595215, "learning_rate": 7.859262399321747e-06, "loss": 0.5137, "mean_token_accuracy": 0.8399428129196167, "num_tokens": 70577046.0, "step": 1855 }, { "epoch": 0.23610227706398676, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 2.0095815658569336, "learning_rate": 7.863501483679526e-06, "loss": 0.4722, "mean_token_accuracy": 0.8532389402389526, "num_tokens": 70616483.0, "step": 1856 }, { "epoch": 0.2362294873425773, "ewc_loss": 0.031494140625, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.6941299438476562e-05, "grad_norm": 2.1141295433044434, "learning_rate": 7.867740568037305e-06, "loss": 0.4725, "mean_token_accuracy": 0.849673330783844, "num_tokens": 70654073.0, "step": 1857 }, { "epoch": 0.2363566976211678, "ewc_loss": 0.029296875, "ewc_loss_diag": 4.506111145019531e-05, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 2.0240085124969482, "learning_rate": 7.871979652395084e-06, "loss": 0.4965, "mean_token_accuracy": 0.8431960344314575, "num_tokens": 70687530.0, "step": 1858 }, { "epoch": 0.2364839078997583, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.0945026874542236, "learning_rate": 7.876218736752863e-06, "loss": 0.4547, "mean_token_accuracy": 0.8631075620651245, "num_tokens": 70729156.0, "step": 1859 }, { "epoch": 0.23661111817834882, "ewc_loss": 0.0291748046875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 2.467632293701172e-05, "grad_norm": 2.0196077823638916, "learning_rate": 7.880457821110642e-06, "loss": 0.5234, "mean_token_accuracy": 0.8390983939170837, "num_tokens": 70769781.0, "step": 1860 }, { "epoch": 0.23673832845693932, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.088536500930786, "learning_rate": 7.88469690546842e-06, "loss": 0.458, "mean_token_accuracy": 0.860756516456604, "num_tokens": 70811822.0, "step": 1861 }, { "epoch": 0.23686553873552982, "ewc_loss": 0.029541015625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 2.041778564453125, "learning_rate": 7.8889359898262e-06, "loss": 0.5037, "mean_token_accuracy": 0.8409573435783386, "num_tokens": 70850470.0, "step": 1862 }, { "epoch": 0.23699274901412035, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 2.08819317817688, "learning_rate": 7.893175074183978e-06, "loss": 0.4964, "mean_token_accuracy": 0.8453043699264526, "num_tokens": 70889270.0, "step": 1863 }, { "epoch": 0.23711995929271085, "ewc_loss": 0.029296875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 2.47955322265625e-05, "grad_norm": 2.0289957523345947, "learning_rate": 7.897414158541755e-06, "loss": 0.4871, "mean_token_accuracy": 0.8481631278991699, "num_tokens": 70928787.0, "step": 1864 }, { "epoch": 0.23724716957130135, "ewc_loss": 0.0308837890625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 2.1101324558258057, "learning_rate": 7.901653242899534e-06, "loss": 0.5241, "mean_token_accuracy": 0.8416668176651001, "num_tokens": 70965726.0, "step": 1865 }, { "epoch": 0.23737437984989188, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 2.5153160095214844e-05, "grad_norm": 2.035226821899414, "learning_rate": 7.905892327257313e-06, "loss": 0.4865, "mean_token_accuracy": 0.8460821509361267, "num_tokens": 71002603.0, "step": 1866 }, { "epoch": 0.23750159012848238, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.0996174812316895, "learning_rate": 7.910131411615092e-06, "loss": 0.4806, "mean_token_accuracy": 0.8534862399101257, "num_tokens": 71039930.0, "step": 1867 }, { "epoch": 0.2376288004070729, "ewc_loss": 0.029541015625, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 2.0289323329925537, "learning_rate": 7.914370495972871e-06, "loss": 0.486, "mean_token_accuracy": 0.8483579754829407, "num_tokens": 71077789.0, "step": 1868 }, { "epoch": 0.2377560106856634, "ewc_loss": 0.03125, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 2.0990042686462402, "learning_rate": 7.91860958033065e-06, "loss": 0.5298, "mean_token_accuracy": 0.8381139039993286, "num_tokens": 71121325.0, "step": 1869 }, { "epoch": 0.2378832209642539, "ewc_loss": 0.0294189453125, "ewc_loss_diag": 4.553794860839844e-05, "ewc_loss_parallel": 2.491474151611328e-05, "grad_norm": 2.043682813644409, "learning_rate": 7.922848664688429e-06, "loss": 0.5277, "mean_token_accuracy": 0.8412994742393494, "num_tokens": 71159764.0, "step": 1870 }, { "epoch": 0.23801043124284443, "ewc_loss": 0.031494140625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 2.0867578983306885, "learning_rate": 7.927087749046206e-06, "loss": 0.4349, "mean_token_accuracy": 0.8680909872055054, "num_tokens": 71196559.0, "step": 1871 }, { "epoch": 0.23813764152143493, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 2.0428810119628906, "learning_rate": 7.931326833403985e-06, "loss": 0.4695, "mean_token_accuracy": 0.8508525490760803, "num_tokens": 71229925.0, "step": 1872 }, { "epoch": 0.23826485180002543, "ewc_loss": 0.031005859375, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 2.1054821014404297, "learning_rate": 7.935565917761764e-06, "loss": 0.4816, "mean_token_accuracy": 0.8504911661148071, "num_tokens": 71273764.0, "step": 1873 }, { "epoch": 0.23839206207861596, "ewc_loss": 0.0299072265625, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 2.5272369384765625e-05, "grad_norm": 2.0531516075134277, "learning_rate": 7.939805002119543e-06, "loss": 0.5026, "mean_token_accuracy": 0.8439772129058838, "num_tokens": 71309090.0, "step": 1874 }, { "epoch": 0.23851927235720646, "ewc_loss": 0.03125, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.08858323097229, "learning_rate": 7.944044086477321e-06, "loss": 0.5012, "mean_token_accuracy": 0.8447844982147217, "num_tokens": 71348867.0, "step": 1875 }, { "epoch": 0.23864648263579696, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 2.040350914001465, "learning_rate": 7.9482831708351e-06, "loss": 0.4479, "mean_token_accuracy": 0.859078586101532, "num_tokens": 71385707.0, "step": 1876 }, { "epoch": 0.2387736929143875, "ewc_loss": 0.03076171875, "ewc_loss_diag": 4.601478576660156e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.0664334297180176, "learning_rate": 7.95252225519288e-06, "loss": 0.4183, "mean_token_accuracy": 0.8682895302772522, "num_tokens": 71428516.0, "step": 1877 }, { "epoch": 0.238900903192978, "ewc_loss": 0.030029296875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 2.044621229171753, "learning_rate": 7.956761339550658e-06, "loss": 0.4434, "mean_token_accuracy": 0.8618917465209961, "num_tokens": 71466913.0, "step": 1878 }, { "epoch": 0.2390281134715685, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 2.088080883026123, "learning_rate": 7.961000423908435e-06, "loss": 0.4975, "mean_token_accuracy": 0.8505932092666626, "num_tokens": 71506362.0, "step": 1879 }, { "epoch": 0.23915532375015902, "ewc_loss": 0.030029296875, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 2.051316499710083, "learning_rate": 7.965239508266214e-06, "loss": 0.4695, "mean_token_accuracy": 0.8547536134719849, "num_tokens": 71544303.0, "step": 1880 }, { "epoch": 0.23928253402874952, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 2.077359914779663, "learning_rate": 7.969478592623993e-06, "loss": 0.4515, "mean_token_accuracy": 0.8594312071800232, "num_tokens": 71582295.0, "step": 1881 }, { "epoch": 0.23940974430734002, "ewc_loss": 0.0296630859375, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 2.5033950805664062e-05, "grad_norm": 2.047562599182129, "learning_rate": 7.973717676981772e-06, "loss": 0.4739, "mean_token_accuracy": 0.8537161946296692, "num_tokens": 71621880.0, "step": 1882 }, { "epoch": 0.23953695458593055, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 4.649162292480469e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.076559543609619, "learning_rate": 7.97795676133955e-06, "loss": 0.4375, "mean_token_accuracy": 0.8619986176490784, "num_tokens": 71657997.0, "step": 1883 }, { "epoch": 0.23966416486452105, "ewc_loss": 0.0302734375, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 2.072068214416504, "learning_rate": 7.98219584569733e-06, "loss": 0.5802, "mean_token_accuracy": 0.819523811340332, "num_tokens": 71697622.0, "step": 1884 }, { "epoch": 0.23979137514311155, "ewc_loss": 0.030517578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 2.0740413665771484, "learning_rate": 7.986434930055109e-06, "loss": 0.4429, "mean_token_accuracy": 0.8590186834335327, "num_tokens": 71733354.0, "step": 1885 }, { "epoch": 0.23991858542170208, "ewc_loss": 0.030029296875, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 2.5391578674316406e-05, "grad_norm": 2.0667617321014404, "learning_rate": 7.990674014412887e-06, "loss": 0.4971, "mean_token_accuracy": 0.8461245894432068, "num_tokens": 71777172.0, "step": 1886 }, { "epoch": 0.24004579570029258, "ewc_loss": 0.030517578125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 2.072908401489258, "learning_rate": 7.994913098770666e-06, "loss": 0.4872, "mean_token_accuracy": 0.8464639186859131, "num_tokens": 71816165.0, "step": 1887 }, { "epoch": 0.24017300597888308, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 4.673004150390625e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.0718512535095215, "learning_rate": 7.999152183128445e-06, "loss": 0.5163, "mean_token_accuracy": 0.8382241129875183, "num_tokens": 71852313.0, "step": 1888 }, { "epoch": 0.2403002162574736, "ewc_loss": 0.030517578125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 2.0643362998962402, "learning_rate": 8.003391267486224e-06, "loss": 0.4664, "mean_token_accuracy": 0.8544268608093262, "num_tokens": 71892462.0, "step": 1889 }, { "epoch": 0.2404274265360641, "ewc_loss": 0.030517578125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 2.0961616039276123, "learning_rate": 8.007630351844003e-06, "loss": 0.4759, "mean_token_accuracy": 0.8545559644699097, "num_tokens": 71924079.0, "step": 1890 }, { "epoch": 0.24055463681465464, "ewc_loss": 0.0302734375, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 2.5510787963867188e-05, "grad_norm": 2.071027994155884, "learning_rate": 8.011869436201782e-06, "loss": 0.4639, "mean_token_accuracy": 0.8559225797653198, "num_tokens": 71959546.0, "step": 1891 }, { "epoch": 0.24068184709324514, "ewc_loss": 0.03076171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.078892946243286, "learning_rate": 8.01610852055956e-06, "loss": 0.498, "mean_token_accuracy": 0.8452212810516357, "num_tokens": 72002565.0, "step": 1892 }, { "epoch": 0.24080905737183564, "ewc_loss": 0.030517578125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.0814216136932373, "learning_rate": 8.02034760491734e-06, "loss": 0.4569, "mean_token_accuracy": 0.8587125539779663, "num_tokens": 72037178.0, "step": 1893 }, { "epoch": 0.24093626765042617, "ewc_loss": 0.030517578125, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.0755369663238525, "learning_rate": 8.024586689275117e-06, "loss": 0.4822, "mean_token_accuracy": 0.84996497631073, "num_tokens": 72076562.0, "step": 1894 }, { "epoch": 0.24106347792901667, "ewc_loss": 0.03076171875, "ewc_loss_diag": 4.696846008300781e-05, "ewc_loss_parallel": 2.6106834411621094e-05, "grad_norm": 2.089745283126831, "learning_rate": 8.028825773632896e-06, "loss": 0.4912, "mean_token_accuracy": 0.851464569568634, "num_tokens": 72113379.0, "step": 1895 }, { "epoch": 0.24119068820760717, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 2.075378894805908, "learning_rate": 8.033064857990675e-06, "loss": 0.4194, "mean_token_accuracy": 0.8686447143554688, "num_tokens": 72149225.0, "step": 1896 }, { "epoch": 0.2413178984861977, "ewc_loss": 0.03076171875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 2.0802886486053467, "learning_rate": 8.037303942348453e-06, "loss": 0.465, "mean_token_accuracy": 0.8541080951690674, "num_tokens": 72185774.0, "step": 1897 }, { "epoch": 0.2414451087647882, "ewc_loss": 0.030517578125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.064927339553833, "learning_rate": 8.041543026706232e-06, "loss": 0.4596, "mean_token_accuracy": 0.857730507850647, "num_tokens": 72224270.0, "step": 1898 }, { "epoch": 0.2415723190433787, "ewc_loss": 0.03076171875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.5987625122070312e-05, "grad_norm": 2.091979742050171, "learning_rate": 8.045782111064011e-06, "loss": 0.4691, "mean_token_accuracy": 0.8529156446456909, "num_tokens": 72260629.0, "step": 1899 }, { "epoch": 0.24169952932196923, "ewc_loss": 0.030517578125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.072756290435791, "learning_rate": 8.05002119542179e-06, "loss": 0.5152, "mean_token_accuracy": 0.8396848440170288, "num_tokens": 72300040.0, "step": 1900 }, { "epoch": 0.24182673960055973, "ewc_loss": 0.031005859375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 2.1026434898376465, "learning_rate": 8.054260279779569e-06, "loss": 0.4829, "mean_token_accuracy": 0.8479076027870178, "num_tokens": 72335396.0, "step": 1901 }, { "epoch": 0.24195394987915023, "ewc_loss": 0.030517578125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.0670111179351807, "learning_rate": 8.058499364137346e-06, "loss": 0.5211, "mean_token_accuracy": 0.8424999117851257, "num_tokens": 72375215.0, "step": 1902 }, { "epoch": 0.24208116015774075, "ewc_loss": 0.031005859375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.6226043701171875e-05, "grad_norm": 2.0970518589019775, "learning_rate": 8.062738448495125e-06, "loss": 0.4362, "mean_token_accuracy": 0.8645603656768799, "num_tokens": 72416407.0, "step": 1903 }, { "epoch": 0.24220837043633126, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 2.0627074241638184, "learning_rate": 8.066977532852904e-06, "loss": 0.4361, "mean_token_accuracy": 0.8677095770835876, "num_tokens": 72455031.0, "step": 1904 }, { "epoch": 0.24233558071492176, "ewc_loss": 0.03125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.1232199668884277, "learning_rate": 8.071216617210683e-06, "loss": 0.5016, "mean_token_accuracy": 0.8443907499313354, "num_tokens": 72487186.0, "step": 1905 }, { "epoch": 0.24246279099351228, "ewc_loss": 0.030517578125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.574920654296875e-05, "grad_norm": 2.05749773979187, "learning_rate": 8.075455701568462e-06, "loss": 0.4686, "mean_token_accuracy": 0.853216290473938, "num_tokens": 72525659.0, "step": 1906 }, { "epoch": 0.24259000127210278, "ewc_loss": 0.031982421875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.7179718017578125e-05, "grad_norm": 2.1056604385375977, "learning_rate": 8.07969478592624e-06, "loss": 0.4641, "mean_token_accuracy": 0.858776867389679, "num_tokens": 72567912.0, "step": 1907 }, { "epoch": 0.24271721155069328, "ewc_loss": 0.0303955078125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.562999725341797e-05, "grad_norm": 2.068465232849121, "learning_rate": 8.08393387028402e-06, "loss": 0.4202, "mean_token_accuracy": 0.8687143325805664, "num_tokens": 72599525.0, "step": 1908 }, { "epoch": 0.2428444218292838, "ewc_loss": 0.031982421875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.7298927307128906e-05, "grad_norm": 2.1244516372680664, "learning_rate": 8.088172954641797e-06, "loss": 0.4762, "mean_token_accuracy": 0.8513078689575195, "num_tokens": 72635391.0, "step": 1909 }, { "epoch": 0.2429716321078743, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 2.10176944732666, "learning_rate": 8.092412038999576e-06, "loss": 0.5163, "mean_token_accuracy": 0.8389639258384705, "num_tokens": 72666058.0, "step": 1910 }, { "epoch": 0.24309884238646481, "ewc_loss": 0.031494140625, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 2.1190590858459473, "learning_rate": 8.096651123357354e-06, "loss": 0.5241, "mean_token_accuracy": 0.8336457014083862, "num_tokens": 72699869.0, "step": 1911 }, { "epoch": 0.24322605266505534, "ewc_loss": 0.0306396484375, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.586841583251953e-05, "grad_norm": 2.090388774871826, "learning_rate": 8.100890207715133e-06, "loss": 0.5195, "mean_token_accuracy": 0.842136800289154, "num_tokens": 72735660.0, "step": 1912 }, { "epoch": 0.24335326294364584, "ewc_loss": 0.031494140625, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 2.0988821983337402, "learning_rate": 8.105129292072912e-06, "loss": 0.4892, "mean_token_accuracy": 0.8487051725387573, "num_tokens": 72774334.0, "step": 1913 }, { "epoch": 0.24348047322223634, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 2.0945143699645996, "learning_rate": 8.109368376430691e-06, "loss": 0.5149, "mean_token_accuracy": 0.8417400121688843, "num_tokens": 72808387.0, "step": 1914 }, { "epoch": 0.24360768350082687, "ewc_loss": 0.031494140625, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 2.0926449298858643, "learning_rate": 8.11360746078847e-06, "loss": 0.4546, "mean_token_accuracy": 0.8598339557647705, "num_tokens": 72846144.0, "step": 1915 }, { "epoch": 0.24373489377941737, "ewc_loss": 0.031494140625, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 2.079177141189575, "learning_rate": 8.117846545146249e-06, "loss": 0.4315, "mean_token_accuracy": 0.8642640709877014, "num_tokens": 72887381.0, "step": 1916 }, { "epoch": 0.2438621040580079, "ewc_loss": 0.031982421875, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.7179718017578125e-05, "grad_norm": 2.1036739349365234, "learning_rate": 8.122085629504028e-06, "loss": 0.516, "mean_token_accuracy": 0.8373554348945618, "num_tokens": 72927126.0, "step": 1917 }, { "epoch": 0.2439893143365984, "ewc_loss": 0.031494140625, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.6702880859375e-05, "grad_norm": 2.087395668029785, "learning_rate": 8.126324713861807e-06, "loss": 0.4919, "mean_token_accuracy": 0.8490407466888428, "num_tokens": 72968770.0, "step": 1918 }, { "epoch": 0.2441165246151889, "ewc_loss": 0.03173828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.7060508728027344e-05, "grad_norm": 2.099949359893799, "learning_rate": 8.130563798219586e-06, "loss": 0.4698, "mean_token_accuracy": 0.8578018546104431, "num_tokens": 73008831.0, "step": 1919 }, { "epoch": 0.24424373489377943, "ewc_loss": 0.03125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.0892865657806396, "learning_rate": 8.134802882577364e-06, "loss": 0.5042, "mean_token_accuracy": 0.8476414680480957, "num_tokens": 73048469.0, "step": 1920 }, { "epoch": 0.24437094517236993, "ewc_loss": 0.03173828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.6941299438476562e-05, "grad_norm": 2.085559368133545, "learning_rate": 8.139041966935143e-06, "loss": 0.4849, "mean_token_accuracy": 0.8517560362815857, "num_tokens": 73096067.0, "step": 1921 }, { "epoch": 0.24449815545096043, "ewc_loss": 0.031494140625, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 2.1224288940429688, "learning_rate": 8.143281051292922e-06, "loss": 0.546, "mean_token_accuracy": 0.8374395966529846, "num_tokens": 73132680.0, "step": 1922 }, { "epoch": 0.24462536572955096, "ewc_loss": 0.031494140625, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 2.0943710803985596, "learning_rate": 8.147520135650701e-06, "loss": 0.5004, "mean_token_accuracy": 0.8448023796081543, "num_tokens": 73168436.0, "step": 1923 }, { "epoch": 0.24475257600814146, "ewc_loss": 0.03173828125, "ewc_loss_diag": 4.7206878662109375e-05, "ewc_loss_parallel": 2.6941299438476562e-05, "grad_norm": 2.1000638008117676, "learning_rate": 8.151759220008478e-06, "loss": 0.4774, "mean_token_accuracy": 0.8503654599189758, "num_tokens": 73210961.0, "step": 1924 }, { "epoch": 0.24487978628673196, "ewc_loss": 0.03125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.1026697158813477, "learning_rate": 8.155998304366257e-06, "loss": 0.4825, "mean_token_accuracy": 0.8558545112609863, "num_tokens": 73244580.0, "step": 1925 }, { "epoch": 0.2450069965653225, "ewc_loss": 0.031494140625, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 2.130213975906372, "learning_rate": 8.160237388724036e-06, "loss": 0.5219, "mean_token_accuracy": 0.8387817144393921, "num_tokens": 73278176.0, "step": 1926 }, { "epoch": 0.245134206843913, "ewc_loss": 0.03125, "ewc_loss_diag": 4.744529724121094e-05, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 2.087130546569824, "learning_rate": 8.164476473081815e-06, "loss": 0.4961, "mean_token_accuracy": 0.847242534160614, "num_tokens": 73315678.0, "step": 1927 }, { "epoch": 0.2452614171225035, "ewc_loss": 0.03173828125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 2.7060508728027344e-05, "grad_norm": 2.1123507022857666, "learning_rate": 8.168715557439594e-06, "loss": 0.4419, "mean_token_accuracy": 0.8619803190231323, "num_tokens": 73356336.0, "step": 1928 }, { "epoch": 0.24538862740109402, "ewc_loss": 0.03125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 2.658367156982422e-05, "grad_norm": 2.0897374153137207, "learning_rate": 8.172954641797373e-06, "loss": 0.4639, "mean_token_accuracy": 0.8548648357391357, "num_tokens": 73396605.0, "step": 1929 }, { "epoch": 0.24551583767968452, "ewc_loss": 0.031982421875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 2.7179718017578125e-05, "grad_norm": 2.127272605895996, "learning_rate": 8.177193726155152e-06, "loss": 0.5733, "mean_token_accuracy": 0.8250792026519775, "num_tokens": 73444278.0, "step": 1930 }, { "epoch": 0.24564304795827502, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 2.0863325595855713, "learning_rate": 8.18143281051293e-06, "loss": 0.5314, "mean_token_accuracy": 0.8371790647506714, "num_tokens": 73483755.0, "step": 1931 }, { "epoch": 0.24577025823686555, "ewc_loss": 0.0322265625, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 2.7418136596679688e-05, "grad_norm": 2.129385471343994, "learning_rate": 8.185671894870708e-06, "loss": 0.5455, "mean_token_accuracy": 0.8320791721343994, "num_tokens": 73522819.0, "step": 1932 }, { "epoch": 0.24589746851545605, "ewc_loss": 0.0311279296875, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 2.6345252990722656e-05, "grad_norm": 2.08667254447937, "learning_rate": 8.189910979228487e-06, "loss": 0.4458, "mean_token_accuracy": 0.8579939603805542, "num_tokens": 73554973.0, "step": 1933 }, { "epoch": 0.24602467879404655, "ewc_loss": 0.0322265625, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 2.7298927307128906e-05, "grad_norm": 2.13309645652771, "learning_rate": 8.194150063586265e-06, "loss": 0.5329, "mean_token_accuracy": 0.8360967636108398, "num_tokens": 73592541.0, "step": 1934 }, { "epoch": 0.24615188907263708, "ewc_loss": 0.03125, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 2.6464462280273438e-05, "grad_norm": 2.089085340499878, "learning_rate": 8.198389147944044e-06, "loss": 0.5185, "mean_token_accuracy": 0.8380995392799377, "num_tokens": 73629328.0, "step": 1935 }, { "epoch": 0.24627909935122758, "ewc_loss": 0.0322265625, "ewc_loss_diag": 4.76837158203125e-05, "ewc_loss_parallel": 2.7418136596679688e-05, "grad_norm": 2.1282293796539307, "learning_rate": 8.202628232301823e-06, "loss": 0.4946, "mean_token_accuracy": 0.8497388362884521, "num_tokens": 73664944.0, "step": 1936 }, { "epoch": 0.24640630962981808, "ewc_loss": 0.03173828125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 2.6941299438476562e-05, "grad_norm": 2.0763437747955322, "learning_rate": 8.206867316659602e-06, "loss": 0.5289, "mean_token_accuracy": 0.8357219099998474, "num_tokens": 73712403.0, "step": 1937 }, { "epoch": 0.2465335199084086, "ewc_loss": 0.03271484375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 2.777576446533203e-05, "grad_norm": 2.1196088790893555, "learning_rate": 8.211106401017381e-06, "loss": 0.407, "mean_token_accuracy": 0.8711059093475342, "num_tokens": 73752716.0, "step": 1938 }, { "epoch": 0.2466607301869991, "ewc_loss": 0.03173828125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 2.682209014892578e-05, "grad_norm": 2.861953020095825, "learning_rate": 8.21534548537516e-06, "loss": 0.5346, "mean_token_accuracy": 0.8333420753479004, "num_tokens": 73796900.0, "step": 1939 }, { "epoch": 0.2467879404655896, "ewc_loss": 0.03515625, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 2.139319896697998, "learning_rate": 8.219584569732939e-06, "loss": 0.5075, "mean_token_accuracy": 0.8408917188644409, "num_tokens": 73836426.0, "step": 1940 }, { "epoch": 0.24691515074418013, "ewc_loss": 0.0341796875, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 2.944469451904297e-05, "grad_norm": 2.276449680328369, "learning_rate": 8.223823654090718e-06, "loss": 0.4445, "mean_token_accuracy": 0.8610784411430359, "num_tokens": 73872147.0, "step": 1941 }, { "epoch": 0.24704236102277063, "ewc_loss": 0.032470703125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 2.753734588623047e-05, "grad_norm": 2.0525457859039307, "learning_rate": 8.228062738448497e-06, "loss": 0.4801, "mean_token_accuracy": 0.8482316732406616, "num_tokens": 73912878.0, "step": 1942 }, { "epoch": 0.24716957130136116, "ewc_loss": 0.036865234375, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.3338584899902344, "learning_rate": 8.232301822806274e-06, "loss": 0.5111, "mean_token_accuracy": 0.8444835543632507, "num_tokens": 73952152.0, "step": 1943 }, { "epoch": 0.24729678157995166, "ewc_loss": 0.033203125, "ewc_loss_diag": 4.792213439941406e-05, "ewc_loss_parallel": 2.849102020263672e-05, "grad_norm": 2.0883684158325195, "learning_rate": 8.236540907164053e-06, "loss": 0.4945, "mean_token_accuracy": 0.8493932485580444, "num_tokens": 73988760.0, "step": 1944 }, { "epoch": 0.24742399185854216, "ewc_loss": 0.036865234375, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.3188490867614746, "learning_rate": 8.240779991521831e-06, "loss": 0.4747, "mean_token_accuracy": 0.8535350561141968, "num_tokens": 74028107.0, "step": 1945 }, { "epoch": 0.2475512021371327, "ewc_loss": 0.033447265625, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 2.86102294921875e-05, "grad_norm": 2.089097499847412, "learning_rate": 8.24501907587961e-06, "loss": 0.5124, "mean_token_accuracy": 0.8409093022346497, "num_tokens": 74071236.0, "step": 1946 }, { "epoch": 0.2476784124157232, "ewc_loss": 0.036865234375, "ewc_loss_diag": 4.8160552978515625e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.331315040588379, "learning_rate": 8.24925816023739e-06, "loss": 0.5158, "mean_token_accuracy": 0.8482450842857361, "num_tokens": 74105734.0, "step": 1947 }, { "epoch": 0.2478056226943137, "ewc_loss": 0.033935546875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 2.9087066650390625e-05, "grad_norm": 2.1115238666534424, "learning_rate": 8.253497244595168e-06, "loss": 0.5123, "mean_token_accuracy": 0.8422327637672424, "num_tokens": 74142569.0, "step": 1948 }, { "epoch": 0.24793283297290422, "ewc_loss": 0.03662109375, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.301215648651123, "learning_rate": 8.257736328952947e-06, "loss": 0.5353, "mean_token_accuracy": 0.8353307247161865, "num_tokens": 74185395.0, "step": 1949 }, { "epoch": 0.24806004325149472, "ewc_loss": 0.033935546875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 2.9087066650390625e-05, "grad_norm": 2.1022701263427734, "learning_rate": 8.261975413310726e-06, "loss": 0.4908, "mean_token_accuracy": 0.8484742641448975, "num_tokens": 74223416.0, "step": 1950 }, { "epoch": 0.24818725353008522, "ewc_loss": 0.036376953125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.2781293392181396, "learning_rate": 8.266214497668505e-06, "loss": 0.4538, "mean_token_accuracy": 0.8631294965744019, "num_tokens": 74265356.0, "step": 1951 }, { "epoch": 0.24831446380867575, "ewc_loss": 0.033935546875, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 2.9087066650390625e-05, "grad_norm": 2.117852210998535, "learning_rate": 8.270453582026284e-06, "loss": 0.5255, "mean_token_accuracy": 0.8380625247955322, "num_tokens": 74303807.0, "step": 1952 }, { "epoch": 0.24844167408726625, "ewc_loss": 0.0361328125, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 2.2726752758026123, "learning_rate": 8.274692666384063e-06, "loss": 0.4778, "mean_token_accuracy": 0.8543065786361694, "num_tokens": 74337299.0, "step": 1953 }, { "epoch": 0.24856888436585675, "ewc_loss": 0.03369140625, "ewc_loss_diag": 4.839897155761719e-05, "ewc_loss_parallel": 2.8848648071289062e-05, "grad_norm": 2.1139543056488037, "learning_rate": 8.278931750741841e-06, "loss": 0.4945, "mean_token_accuracy": 0.8459365367889404, "num_tokens": 74370817.0, "step": 1954 }, { "epoch": 0.24869609464444728, "ewc_loss": 0.03564453125, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 2.2655515670776367, "learning_rate": 8.283170835099619e-06, "loss": 0.4846, "mean_token_accuracy": 0.8574491739273071, "num_tokens": 74408009.0, "step": 1955 }, { "epoch": 0.24882330492303778, "ewc_loss": 0.033447265625, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 2.872943878173828e-05, "grad_norm": 2.0995686054229736, "learning_rate": 8.287409919457398e-06, "loss": 0.4609, "mean_token_accuracy": 0.8576889038085938, "num_tokens": 74446669.0, "step": 1956 }, { "epoch": 0.24895051520162828, "ewc_loss": 0.03564453125, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 2.2432596683502197, "learning_rate": 8.291649003815176e-06, "loss": 0.5052, "mean_token_accuracy": 0.8454148769378662, "num_tokens": 74484264.0, "step": 1957 }, { "epoch": 0.2490777254802188, "ewc_loss": 0.033203125, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 2.8371810913085938e-05, "grad_norm": 2.095193862915039, "learning_rate": 8.295888088172955e-06, "loss": 0.5216, "mean_token_accuracy": 0.8439808487892151, "num_tokens": 74523956.0, "step": 1958 }, { "epoch": 0.2492049357588093, "ewc_loss": 0.03515625, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 3.039836883544922e-05, "grad_norm": 2.2286417484283447, "learning_rate": 8.300127172530734e-06, "loss": 0.4242, "mean_token_accuracy": 0.8685464859008789, "num_tokens": 74560748.0, "step": 1959 }, { "epoch": 0.2493321460373998, "ewc_loss": 0.032958984375, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 2.8133392333984375e-05, "grad_norm": 2.1152126789093018, "learning_rate": 8.304366256888513e-06, "loss": 0.4716, "mean_token_accuracy": 0.8544859886169434, "num_tokens": 74596465.0, "step": 1960 }, { "epoch": 0.24945935631599034, "ewc_loss": 0.035400390625, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.2300310134887695, "learning_rate": 8.308605341246292e-06, "loss": 0.4809, "mean_token_accuracy": 0.8517245650291443, "num_tokens": 74632788.0, "step": 1961 }, { "epoch": 0.24958656659458084, "ewc_loss": 0.03271484375, "ewc_loss_diag": 4.863739013671875e-05, "ewc_loss_parallel": 2.7894973754882812e-05, "grad_norm": 2.104215621948242, "learning_rate": 8.312844425604069e-06, "loss": 0.478, "mean_token_accuracy": 0.8576333522796631, "num_tokens": 74667723.0, "step": 1962 }, { "epoch": 0.24971377687317134, "ewc_loss": 0.03466796875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 2.9802322387695312e-05, "grad_norm": 2.1998562812805176, "learning_rate": 8.317083509961848e-06, "loss": 0.5113, "mean_token_accuracy": 0.8466854095458984, "num_tokens": 74709205.0, "step": 1963 }, { "epoch": 0.24984098715176187, "ewc_loss": 0.03271484375, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 2.765655517578125e-05, "grad_norm": 2.089672803878784, "learning_rate": 8.321322594319627e-06, "loss": 0.4524, "mean_token_accuracy": 0.8585488796234131, "num_tokens": 74749437.0, "step": 1964 }, { "epoch": 0.24996819743035237, "ewc_loss": 0.034912109375, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 3.0040740966796875e-05, "grad_norm": 2.20731258392334, "learning_rate": 8.325561678677406e-06, "loss": 0.5218, "mean_token_accuracy": 0.8385169506072998, "num_tokens": 74790602.0, "step": 1965 }, { "epoch": 0.2500954077089429, "ewc_loss": 0.03271484375, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 2.8014183044433594e-05, "grad_norm": 2.099052906036377, "learning_rate": 8.329800763035185e-06, "loss": 0.5262, "mean_token_accuracy": 0.839640200138092, "num_tokens": 74830845.0, "step": 1966 }, { "epoch": 0.25022261798753337, "ewc_loss": 0.03466796875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 2.968311309814453e-05, "grad_norm": 2.210339069366455, "learning_rate": 8.334039847392964e-06, "loss": 0.4772, "mean_token_accuracy": 0.8505111932754517, "num_tokens": 74864208.0, "step": 1967 }, { "epoch": 0.2503498282661239, "ewc_loss": 0.0322265625, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 2.7418136596679688e-05, "grad_norm": 2.1023736000061035, "learning_rate": 8.338278931750742e-06, "loss": 0.4479, "mean_token_accuracy": 0.8622412085533142, "num_tokens": 74897217.0, "step": 1968 }, { "epoch": 0.2504770385447144, "ewc_loss": 0.03466796875, "ewc_loss_diag": 4.887580871582031e-05, "ewc_loss_parallel": 2.968311309814453e-05, "grad_norm": 2.206927537918091, "learning_rate": 8.342518016108521e-06, "loss": 0.4716, "mean_token_accuracy": 0.8521310091018677, "num_tokens": 74932105.0, "step": 1969 }, { "epoch": 0.2506042488233049, "ewc_loss": 0.03271484375, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 2.765655517578125e-05, "grad_norm": 2.117525100708008, "learning_rate": 8.3467571004663e-06, "loss": 0.4747, "mean_token_accuracy": 0.8546377420425415, "num_tokens": 74965229.0, "step": 1970 }, { "epoch": 0.2507314591018954, "ewc_loss": 0.0341796875, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 2.9206275939941406e-05, "grad_norm": 2.199080228805542, "learning_rate": 8.350996184824079e-06, "loss": 0.5213, "mean_token_accuracy": 0.8394551277160645, "num_tokens": 75002290.0, "step": 1971 }, { "epoch": 0.25085866938048595, "ewc_loss": 0.03271484375, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 2.765655517578125e-05, "grad_norm": 2.1097309589385986, "learning_rate": 8.355235269181858e-06, "loss": 0.4987, "mean_token_accuracy": 0.8444002270698547, "num_tokens": 75038113.0, "step": 1972 }, { "epoch": 0.2509858796590764, "ewc_loss": 0.0341796875, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 2.9325485229492188e-05, "grad_norm": 2.1802871227264404, "learning_rate": 8.359474353539637e-06, "loss": 0.463, "mean_token_accuracy": 0.8569153547286987, "num_tokens": 75078145.0, "step": 1973 }, { "epoch": 0.25111308993766696, "ewc_loss": 0.03271484375, "ewc_loss_diag": 4.935264587402344e-05, "ewc_loss_parallel": 2.765655517578125e-05, "grad_norm": 2.102508068084717, "learning_rate": 8.363713437897416e-06, "loss": 0.55, "mean_token_accuracy": 0.8332717418670654, "num_tokens": 75126473.0, "step": 1974 }, { "epoch": 0.2512403002162575, "ewc_loss": 0.033935546875, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 2.9087066650390625e-05, "grad_norm": 2.182858467102051, "learning_rate": 8.367952522255195e-06, "loss": 0.4859, "mean_token_accuracy": 0.8502926230430603, "num_tokens": 75166785.0, "step": 1975 }, { "epoch": 0.25136751049484796, "ewc_loss": 0.032470703125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 2.7418136596679688e-05, "grad_norm": 2.107106924057007, "learning_rate": 8.372191606612974e-06, "loss": 0.4884, "mean_token_accuracy": 0.8476194739341736, "num_tokens": 75204599.0, "step": 1976 }, { "epoch": 0.2514947207734385, "ewc_loss": 0.033935546875, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 2.9087066650390625e-05, "grad_norm": 2.191913366317749, "learning_rate": 8.37643069097075e-06, "loss": 0.4454, "mean_token_accuracy": 0.8602221012115479, "num_tokens": 75235295.0, "step": 1977 }, { "epoch": 0.251621931052029, "ewc_loss": 0.03271484375, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 2.765655517578125e-05, "grad_norm": 2.092048168182373, "learning_rate": 8.38066977532853e-06, "loss": 0.448, "mean_token_accuracy": 0.8626804947853088, "num_tokens": 75275549.0, "step": 1978 }, { "epoch": 0.25174914133061954, "ewc_loss": 0.0341796875, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 2.9325485229492188e-05, "grad_norm": 2.1903460025787354, "learning_rate": 8.384908859686308e-06, "loss": 0.4497, "mean_token_accuracy": 0.8618941307067871, "num_tokens": 75316920.0, "step": 1979 }, { "epoch": 0.25187635160921, "ewc_loss": 0.032470703125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 2.753734588623047e-05, "grad_norm": 2.0888564586639404, "learning_rate": 8.389147944044087e-06, "loss": 0.4851, "mean_token_accuracy": 0.850962221622467, "num_tokens": 75360091.0, "step": 1980 }, { "epoch": 0.25200356188780054, "ewc_loss": 0.034423828125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 2.956390380859375e-05, "grad_norm": 2.2103042602539062, "learning_rate": 8.393387028401866e-06, "loss": 0.5143, "mean_token_accuracy": 0.8406240940093994, "num_tokens": 75397123.0, "step": 1981 }, { "epoch": 0.25213077216639107, "ewc_loss": 0.032470703125, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 2.753734588623047e-05, "grad_norm": 2.1210970878601074, "learning_rate": 8.397626112759645e-06, "loss": 0.536, "mean_token_accuracy": 0.8343985080718994, "num_tokens": 75429983.0, "step": 1982 }, { "epoch": 0.25225798244498154, "ewc_loss": 0.03466796875, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 2.968311309814453e-05, "grad_norm": 2.2131145000457764, "learning_rate": 8.401865197117424e-06, "loss": 0.4678, "mean_token_accuracy": 0.8566421866416931, "num_tokens": 75460322.0, "step": 1983 }, { "epoch": 0.25238519272357207, "ewc_loss": 0.0322265625, "ewc_loss_diag": 4.982948303222656e-05, "ewc_loss_parallel": 2.7298927307128906e-05, "grad_norm": 2.0933165550231934, "learning_rate": 8.406104281475203e-06, "loss": 0.4542, "mean_token_accuracy": 0.8574044108390808, "num_tokens": 75499733.0, "step": 1984 }, { "epoch": 0.2525124030021626, "ewc_loss": 0.03466796875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 2.9802322387695312e-05, "grad_norm": 2.1928749084472656, "learning_rate": 8.41034336583298e-06, "loss": 0.4633, "mean_token_accuracy": 0.856342077255249, "num_tokens": 75541182.0, "step": 1985 }, { "epoch": 0.2526396132807531, "ewc_loss": 0.03271484375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 2.753734588623047e-05, "grad_norm": 2.085277557373047, "learning_rate": 8.414582450190759e-06, "loss": 0.4766, "mean_token_accuracy": 0.8512586355209351, "num_tokens": 75585052.0, "step": 1986 }, { "epoch": 0.2527668235593436, "ewc_loss": 0.034912109375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 2.2278048992156982, "learning_rate": 8.418821534548538e-06, "loss": 0.5672, "mean_token_accuracy": 0.8287312984466553, "num_tokens": 75625208.0, "step": 1987 }, { "epoch": 0.25289403383793413, "ewc_loss": 0.03271484375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 2.753734588623047e-05, "grad_norm": 2.0930137634277344, "learning_rate": 8.423060618906317e-06, "loss": 0.486, "mean_token_accuracy": 0.85086989402771, "num_tokens": 75665205.0, "step": 1988 }, { "epoch": 0.2530212441165246, "ewc_loss": 0.03466796875, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 2.9802322387695312e-05, "grad_norm": 2.216118335723877, "learning_rate": 8.427299703264096e-06, "loss": 0.4839, "mean_token_accuracy": 0.8517088890075684, "num_tokens": 75708278.0, "step": 1989 }, { "epoch": 0.25314845439511513, "ewc_loss": 0.0322265625, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 2.7298927307128906e-05, "grad_norm": 2.0884032249450684, "learning_rate": 8.431538787621875e-06, "loss": 0.5023, "mean_token_accuracy": 0.8450596332550049, "num_tokens": 75750565.0, "step": 1990 }, { "epoch": 0.25327566467370566, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 2.229416847229004, "learning_rate": 8.435777871979653e-06, "loss": 0.47, "mean_token_accuracy": 0.8549354076385498, "num_tokens": 75790519.0, "step": 1991 }, { "epoch": 0.25340287495229613, "ewc_loss": 0.03271484375, "ewc_loss_diag": 5.0067901611328125e-05, "ewc_loss_parallel": 2.753734588623047e-05, "grad_norm": 2.1018788814544678, "learning_rate": 8.440016956337432e-06, "loss": 0.4959, "mean_token_accuracy": 0.8490117192268372, "num_tokens": 75828256.0, "step": 1992 }, { "epoch": 0.25353008523088666, "ewc_loss": 0.035400390625, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 2.229203462600708, "learning_rate": 8.44425604069521e-06, "loss": 0.5102, "mean_token_accuracy": 0.8474535346031189, "num_tokens": 75869238.0, "step": 1993 }, { "epoch": 0.2536572955094772, "ewc_loss": 0.03271484375, "ewc_loss_diag": 5.054473876953125e-05, "ewc_loss_parallel": 2.753734588623047e-05, "grad_norm": 2.100104331970215, "learning_rate": 8.448495125052988e-06, "loss": 0.4965, "mean_token_accuracy": 0.8476938009262085, "num_tokens": 75906776.0, "step": 1994 }, { "epoch": 0.25378450578806766, "ewc_loss": 0.034912109375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 2.9802322387695312e-05, "grad_norm": 2.232588529586792, "learning_rate": 8.452734209410767e-06, "loss": 0.4309, "mean_token_accuracy": 0.868349552154541, "num_tokens": 75940271.0, "step": 1995 }, { "epoch": 0.2539117160666582, "ewc_loss": 0.03271484375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 2.753734588623047e-05, "grad_norm": 2.102957010269165, "learning_rate": 8.456973293768546e-06, "loss": 0.4623, "mean_token_accuracy": 0.8554906845092773, "num_tokens": 75974536.0, "step": 1996 }, { "epoch": 0.2540389263452487, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 3.0040740966796875e-05, "grad_norm": 2.2388627529144287, "learning_rate": 8.461212378126325e-06, "loss": 0.5177, "mean_token_accuracy": 0.8415248394012451, "num_tokens": 76013809.0, "step": 1997 }, { "epoch": 0.2541661366238392, "ewc_loss": 0.032958984375, "ewc_loss_diag": 5.078315734863281e-05, "ewc_loss_parallel": 2.777576446533203e-05, "grad_norm": 2.120795726776123, "learning_rate": 8.465451462484104e-06, "loss": 0.5422, "mean_token_accuracy": 0.8350527286529541, "num_tokens": 76049559.0, "step": 1998 }, { "epoch": 0.2542933469024297, "ewc_loss": 0.035400390625, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 2.234973192214966, "learning_rate": 8.469690546841883e-06, "loss": 0.4865, "mean_token_accuracy": 0.8495902419090271, "num_tokens": 76087091.0, "step": 1999 }, { "epoch": 0.25442055718102025, "ewc_loss": 0.033203125, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 2.7894973754882812e-05, "grad_norm": 2.1185972690582275, "learning_rate": 8.473929631199662e-06, "loss": 0.4617, "mean_token_accuracy": 0.8569697737693787, "num_tokens": 76121466.0, "step": 2000 }, { "epoch": 0.2545477674596107, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 2.2244725227355957, "learning_rate": 8.47816871555744e-06, "loss": 0.4838, "mean_token_accuracy": 0.8548383712768555, "num_tokens": 76158291.0, "step": 2001 }, { "epoch": 0.25467497773820125, "ewc_loss": 0.033203125, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 2.8014183044433594e-05, "grad_norm": 2.113396644592285, "learning_rate": 8.48240779991522e-06, "loss": 0.4953, "mean_token_accuracy": 0.8463501930236816, "num_tokens": 76197350.0, "step": 2002 }, { "epoch": 0.2548021880167918, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 2.221972703933716, "learning_rate": 8.486646884272998e-06, "loss": 0.4897, "mean_token_accuracy": 0.8474797606468201, "num_tokens": 76235367.0, "step": 2003 }, { "epoch": 0.25492939829538225, "ewc_loss": 0.033203125, "ewc_loss_diag": 5.125999450683594e-05, "ewc_loss_parallel": 2.7894973754882812e-05, "grad_norm": 2.1300113201141357, "learning_rate": 8.490885968630777e-06, "loss": 0.5227, "mean_token_accuracy": 0.8373392820358276, "num_tokens": 76270462.0, "step": 2004 }, { "epoch": 0.2550566085739728, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 2.216221809387207, "learning_rate": 8.495125052988556e-06, "loss": 0.5129, "mean_token_accuracy": 0.8466493487358093, "num_tokens": 76315032.0, "step": 2005 }, { "epoch": 0.2551838188525633, "ewc_loss": 0.033203125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 2.8133392333984375e-05, "grad_norm": 2.1195504665374756, "learning_rate": 8.499364137346335e-06, "loss": 0.4922, "mean_token_accuracy": 0.8501177430152893, "num_tokens": 76356337.0, "step": 2006 }, { "epoch": 0.2553110291311538, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 3.0040740966796875e-05, "grad_norm": 2.2286815643310547, "learning_rate": 8.503603221704114e-06, "loss": 0.516, "mean_token_accuracy": 0.8413819074630737, "num_tokens": 76391329.0, "step": 2007 }, { "epoch": 0.2554382394097443, "ewc_loss": 0.033203125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 2.8133392333984375e-05, "grad_norm": 2.111427068710327, "learning_rate": 8.507842306061891e-06, "loss": 0.4768, "mean_token_accuracy": 0.8517071604728699, "num_tokens": 76429015.0, "step": 2008 }, { "epoch": 0.25556544968833483, "ewc_loss": 0.035400390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 2.228177309036255, "learning_rate": 8.51208139041967e-06, "loss": 0.5137, "mean_token_accuracy": 0.8513039946556091, "num_tokens": 76470343.0, "step": 2009 }, { "epoch": 0.2556926599669253, "ewc_loss": 0.033203125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 2.8014183044433594e-05, "grad_norm": 2.128220796585083, "learning_rate": 8.516320474777449e-06, "loss": 0.4957, "mean_token_accuracy": 0.8487573862075806, "num_tokens": 76505140.0, "step": 2010 }, { "epoch": 0.25581987024551583, "ewc_loss": 0.035400390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 2.2230722904205322, "learning_rate": 8.520559559135228e-06, "loss": 0.4558, "mean_token_accuracy": 0.8639568090438843, "num_tokens": 76541701.0, "step": 2011 }, { "epoch": 0.25594708052410636, "ewc_loss": 0.033203125, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 2.7894973754882812e-05, "grad_norm": 2.1268036365509033, "learning_rate": 8.524798643493007e-06, "loss": 0.5143, "mean_token_accuracy": 0.84796142578125, "num_tokens": 76580208.0, "step": 2012 }, { "epoch": 0.25607429080269684, "ewc_loss": 0.035400390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 2.214008092880249, "learning_rate": 8.529037727850785e-06, "loss": 0.463, "mean_token_accuracy": 0.8581802845001221, "num_tokens": 76617750.0, "step": 2013 }, { "epoch": 0.25620150108128736, "ewc_loss": 0.033447265625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 2.8252601623535156e-05, "grad_norm": 2.126708507537842, "learning_rate": 8.533276812208564e-06, "loss": 0.4313, "mean_token_accuracy": 0.8655340671539307, "num_tokens": 76652678.0, "step": 2014 }, { "epoch": 0.2563287113598779, "ewc_loss": 0.035400390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 2.208164691925049, "learning_rate": 8.537515896566342e-06, "loss": 0.4685, "mean_token_accuracy": 0.8543262481689453, "num_tokens": 76695375.0, "step": 2015 }, { "epoch": 0.25645592163846836, "ewc_loss": 0.033447265625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 2.8252601623535156e-05, "grad_norm": 2.1448583602905273, "learning_rate": 8.54175498092412e-06, "loss": 0.4791, "mean_token_accuracy": 0.8537566065788269, "num_tokens": 76728695.0, "step": 2016 }, { "epoch": 0.2565831319170589, "ewc_loss": 0.035400390625, "ewc_loss_diag": 5.14984130859375e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 2.2243990898132324, "learning_rate": 8.5459940652819e-06, "loss": 0.4562, "mean_token_accuracy": 0.8606176376342773, "num_tokens": 76761081.0, "step": 2017 }, { "epoch": 0.2567103421956494, "ewc_loss": 0.033447265625, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 2.8252601623535156e-05, "grad_norm": 2.1363489627838135, "learning_rate": 8.550233149639678e-06, "loss": 0.5493, "mean_token_accuracy": 0.8298390507698059, "num_tokens": 76800335.0, "step": 2018 }, { "epoch": 0.2568375524742399, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 2.2081964015960693, "learning_rate": 8.554472233997457e-06, "loss": 0.5152, "mean_token_accuracy": 0.8417763710021973, "num_tokens": 76839653.0, "step": 2019 }, { "epoch": 0.2569647627528304, "ewc_loss": 0.033447265625, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 2.8371810913085938e-05, "grad_norm": 2.1551597118377686, "learning_rate": 8.558711318355236e-06, "loss": 0.5444, "mean_token_accuracy": 0.8297748565673828, "num_tokens": 76876557.0, "step": 2020 }, { "epoch": 0.25709197303142095, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 2.207988739013672, "learning_rate": 8.562950402713015e-06, "loss": 0.4384, "mean_token_accuracy": 0.8684675693511963, "num_tokens": 76910000.0, "step": 2021 }, { "epoch": 0.2572191833100114, "ewc_loss": 0.033935546875, "ewc_loss_diag": 5.1975250244140625e-05, "ewc_loss_parallel": 2.872943878173828e-05, "grad_norm": 2.1457626819610596, "learning_rate": 8.567189487070794e-06, "loss": 0.5275, "mean_token_accuracy": 0.8410676121711731, "num_tokens": 76946861.0, "step": 2022 }, { "epoch": 0.25734639358860195, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.221366882324219e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 2.203517436981201, "learning_rate": 8.571428571428571e-06, "loss": 0.5443, "mean_token_accuracy": 0.8375667929649353, "num_tokens": 76990831.0, "step": 2023 }, { "epoch": 0.2574736038671925, "ewc_loss": 0.03369140625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 2.86102294921875e-05, "grad_norm": 2.124798059463501, "learning_rate": 8.57566765578635e-06, "loss": 0.4461, "mean_token_accuracy": 0.8662116527557373, "num_tokens": 77033145.0, "step": 2024 }, { "epoch": 0.25760081414578295, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.245208740234375e-05, "ewc_loss_parallel": 3.0040740966796875e-05, "grad_norm": 2.2015328407287598, "learning_rate": 8.579906740144129e-06, "loss": 0.4822, "mean_token_accuracy": 0.8534374237060547, "num_tokens": 77076621.0, "step": 2025 }, { "epoch": 0.2577280244243735, "ewc_loss": 0.033935546875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 2.86102294921875e-05, "grad_norm": 2.152742624282837, "learning_rate": 8.584145824501908e-06, "loss": 0.5167, "mean_token_accuracy": 0.8404185771942139, "num_tokens": 77115951.0, "step": 2026 }, { "epoch": 0.257855234702964, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 2.2038097381591797, "learning_rate": 8.588384908859686e-06, "loss": 0.5145, "mean_token_accuracy": 0.8473915457725525, "num_tokens": 77154674.0, "step": 2027 }, { "epoch": 0.25798244498155454, "ewc_loss": 0.03369140625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 2.849102020263672e-05, "grad_norm": 2.140207290649414, "learning_rate": 8.592623993217465e-06, "loss": 0.4934, "mean_token_accuracy": 0.8503355979919434, "num_tokens": 77192970.0, "step": 2028 }, { "epoch": 0.258109655260145, "ewc_loss": 0.035400390625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 3.0040740966796875e-05, "grad_norm": 2.212214708328247, "learning_rate": 8.596863077575244e-06, "loss": 0.5271, "mean_token_accuracy": 0.8423432111740112, "num_tokens": 77231258.0, "step": 2029 }, { "epoch": 0.25823686553873554, "ewc_loss": 0.033935546875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 2.86102294921875e-05, "grad_norm": 2.1429502964019775, "learning_rate": 8.601102161933023e-06, "loss": 0.5011, "mean_token_accuracy": 0.8475391268730164, "num_tokens": 77269829.0, "step": 2030 }, { "epoch": 0.25836407581732607, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 2.2113842964172363, "learning_rate": 8.605341246290802e-06, "loss": 0.4804, "mean_token_accuracy": 0.8517545461654663, "num_tokens": 77309611.0, "step": 2031 }, { "epoch": 0.25849128609591654, "ewc_loss": 0.033447265625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 2.8133392333984375e-05, "grad_norm": 2.1549017429351807, "learning_rate": 8.609580330648581e-06, "loss": 0.4738, "mean_token_accuracy": 0.8560795187950134, "num_tokens": 77341937.0, "step": 2032 }, { "epoch": 0.25861849637450707, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 2.9802322387695312e-05, "grad_norm": 2.224789619445801, "learning_rate": 8.61381941500636e-06, "loss": 0.5432, "mean_token_accuracy": 0.8390237092971802, "num_tokens": 77380484.0, "step": 2033 }, { "epoch": 0.2587457066530976, "ewc_loss": 0.03369140625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 2.849102020263672e-05, "grad_norm": 2.136439085006714, "learning_rate": 8.618058499364139e-06, "loss": 0.5357, "mean_token_accuracy": 0.8368790745735168, "num_tokens": 77421220.0, "step": 2034 }, { "epoch": 0.25887291693168807, "ewc_loss": 0.035400390625, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 2.2197928428649902, "learning_rate": 8.622297583721918e-06, "loss": 0.4847, "mean_token_accuracy": 0.8550234436988831, "num_tokens": 77465136.0, "step": 2035 }, { "epoch": 0.2590001272102786, "ewc_loss": 0.0341796875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 2.8967857360839844e-05, "grad_norm": 2.1434574127197266, "learning_rate": 8.626536668079696e-06, "loss": 0.4936, "mean_token_accuracy": 0.8480312824249268, "num_tokens": 77500120.0, "step": 2036 }, { "epoch": 0.2591273374888691, "ewc_loss": 0.03564453125, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 2.2361814975738525, "learning_rate": 8.630775752437475e-06, "loss": 0.4813, "mean_token_accuracy": 0.8511059284210205, "num_tokens": 77532860.0, "step": 2037 }, { "epoch": 0.2592545477674596, "ewc_loss": 0.033935546875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 2.872943878173828e-05, "grad_norm": 2.1453330516815186, "learning_rate": 8.635014836795252e-06, "loss": 0.5174, "mean_token_accuracy": 0.8415349721908569, "num_tokens": 77569154.0, "step": 2038 }, { "epoch": 0.2593817580460501, "ewc_loss": 0.03564453125, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 3.039836883544922e-05, "grad_norm": 2.2344131469726562, "learning_rate": 8.639253921153031e-06, "loss": 0.4507, "mean_token_accuracy": 0.8601167798042297, "num_tokens": 77602918.0, "step": 2039 }, { "epoch": 0.25950896832464065, "ewc_loss": 0.033935546875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 2.872943878173828e-05, "grad_norm": 2.1437137126922607, "learning_rate": 8.64349300551081e-06, "loss": 0.4666, "mean_token_accuracy": 0.8551895618438721, "num_tokens": 77637233.0, "step": 2040 }, { "epoch": 0.2596361786032311, "ewc_loss": 0.035888671875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.2342820167541504, "learning_rate": 8.647732089868589e-06, "loss": 0.5502, "mean_token_accuracy": 0.8319684863090515, "num_tokens": 77678344.0, "step": 2041 }, { "epoch": 0.25976338888182166, "ewc_loss": 0.0341796875, "ewc_loss_diag": 5.269050598144531e-05, "ewc_loss_parallel": 2.8967857360839844e-05, "grad_norm": 2.1391448974609375, "learning_rate": 8.651971174226368e-06, "loss": 0.4584, "mean_token_accuracy": 0.8575796484947205, "num_tokens": 77712706.0, "step": 2042 }, { "epoch": 0.2598905991604122, "ewc_loss": 0.035888671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.254249095916748, "learning_rate": 8.656210258584147e-06, "loss": 0.5242, "mean_token_accuracy": 0.8380182981491089, "num_tokens": 77749841.0, "step": 2043 }, { "epoch": 0.26001780943900266, "ewc_loss": 0.0341796875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 2.8967857360839844e-05, "grad_norm": 2.1453804969787598, "learning_rate": 8.660449342941926e-06, "loss": 0.5215, "mean_token_accuracy": 0.839826226234436, "num_tokens": 77788464.0, "step": 2044 }, { "epoch": 0.2601450197175932, "ewc_loss": 0.035888671875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.2500412464141846, "learning_rate": 8.664688427299705e-06, "loss": 0.4983, "mean_token_accuracy": 0.8455182313919067, "num_tokens": 77826493.0, "step": 2045 }, { "epoch": 0.2602722299961837, "ewc_loss": 0.0341796875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 2.8848648071289062e-05, "grad_norm": 2.144826889038086, "learning_rate": 8.668927511657482e-06, "loss": 0.4551, "mean_token_accuracy": 0.8587350249290466, "num_tokens": 77857814.0, "step": 2046 }, { "epoch": 0.2603994402747742, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 2.2653849124908447, "learning_rate": 8.67316659601526e-06, "loss": 0.5046, "mean_token_accuracy": 0.8472673892974854, "num_tokens": 77891263.0, "step": 2047 }, { "epoch": 0.2605266505533647, "ewc_loss": 0.034423828125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 2.9087066650390625e-05, "grad_norm": 2.1264748573303223, "learning_rate": 8.67740568037304e-06, "loss": 0.4709, "mean_token_accuracy": 0.8590312600135803, "num_tokens": 77931683.0, "step": 2048 }, { "epoch": 0.26065386083195524, "ewc_loss": 0.036376953125, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 2.2839667797088623, "learning_rate": 8.681644764730819e-06, "loss": 0.4982, "mean_token_accuracy": 0.8470777273178101, "num_tokens": 77965204.0, "step": 2049 }, { "epoch": 0.2607810711105457, "ewc_loss": 0.0341796875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 2.8848648071289062e-05, "grad_norm": 2.1193525791168213, "learning_rate": 8.685883849088597e-06, "loss": 0.4367, "mean_token_accuracy": 0.867087721824646, "num_tokens": 78003427.0, "step": 2050 }, { "epoch": 0.26090828138913624, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.2907185554504395, "learning_rate": 8.690122933446376e-06, "loss": 0.4977, "mean_token_accuracy": 0.8506008386611938, "num_tokens": 78045524.0, "step": 2051 }, { "epoch": 0.26103549166772677, "ewc_loss": 0.0341796875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 2.8848648071289062e-05, "grad_norm": 2.123558521270752, "learning_rate": 8.694362017804155e-06, "loss": 0.47, "mean_token_accuracy": 0.8549266457557678, "num_tokens": 78085485.0, "step": 2052 }, { "epoch": 0.26116270194631724, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.307710886001587, "learning_rate": 8.698601102161932e-06, "loss": 0.5051, "mean_token_accuracy": 0.8452775478363037, "num_tokens": 78128212.0, "step": 2053 }, { "epoch": 0.26128991222490777, "ewc_loss": 0.0341796875, "ewc_loss_diag": 5.2928924560546875e-05, "ewc_loss_parallel": 2.8967857360839844e-05, "grad_norm": 2.1110901832580566, "learning_rate": 8.702840186519711e-06, "loss": 0.4567, "mean_token_accuracy": 0.8610056638717651, "num_tokens": 78170959.0, "step": 2054 }, { "epoch": 0.2614171225034983, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.321056604385376, "learning_rate": 8.70707927087749e-06, "loss": 0.4578, "mean_token_accuracy": 0.860851526260376, "num_tokens": 78211181.0, "step": 2055 }, { "epoch": 0.2615443327820888, "ewc_loss": 0.034423828125, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 2.9087066650390625e-05, "grad_norm": 2.148212194442749, "learning_rate": 8.711318355235269e-06, "loss": 0.5418, "mean_token_accuracy": 0.830090343952179, "num_tokens": 78246281.0, "step": 2056 }, { "epoch": 0.2616715430606793, "ewc_loss": 0.036865234375, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.334458827972412, "learning_rate": 8.715557439593048e-06, "loss": 0.4616, "mean_token_accuracy": 0.8568254709243774, "num_tokens": 78279604.0, "step": 2057 }, { "epoch": 0.26179875333926983, "ewc_loss": 0.03466796875, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 2.944469451904297e-05, "grad_norm": 2.1383965015411377, "learning_rate": 8.719796523950827e-06, "loss": 0.5438, "mean_token_accuracy": 0.8313683271408081, "num_tokens": 78317534.0, "step": 2058 }, { "epoch": 0.2619259636178603, "ewc_loss": 0.037353515625, "ewc_loss_diag": 5.340576171875e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.3346407413482666, "learning_rate": 8.724035608308606e-06, "loss": 0.426, "mean_token_accuracy": 0.8686760663986206, "num_tokens": 78349744.0, "step": 2059 }, { "epoch": 0.26205317389645083, "ewc_loss": 0.03466796875, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 2.944469451904297e-05, "grad_norm": 2.1331405639648438, "learning_rate": 8.728274692666385e-06, "loss": 0.4949, "mean_token_accuracy": 0.846828818321228, "num_tokens": 78390841.0, "step": 2060 }, { "epoch": 0.26218038417504136, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.3567957878112793, "learning_rate": 8.732513777024163e-06, "loss": 0.4936, "mean_token_accuracy": 0.8489432334899902, "num_tokens": 78425604.0, "step": 2061 }, { "epoch": 0.26230759445363183, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 2.1335859298706055, "learning_rate": 8.736752861381942e-06, "loss": 0.5383, "mean_token_accuracy": 0.8337217569351196, "num_tokens": 78469444.0, "step": 2062 }, { "epoch": 0.26243480473222236, "ewc_loss": 0.037841796875, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 2.353191614151001, "learning_rate": 8.740991945739721e-06, "loss": 0.49, "mean_token_accuracy": 0.84963059425354, "num_tokens": 78507931.0, "step": 2063 }, { "epoch": 0.2625620150108129, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 2.1362571716308594, "learning_rate": 8.7452310300975e-06, "loss": 0.483, "mean_token_accuracy": 0.8529418110847473, "num_tokens": 78550572.0, "step": 2064 }, { "epoch": 0.26268922528940336, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.3613221645355225, "learning_rate": 8.749470114455279e-06, "loss": 0.4632, "mean_token_accuracy": 0.8584471344947815, "num_tokens": 78586975.0, "step": 2065 }, { "epoch": 0.2628164355679939, "ewc_loss": 0.035400390625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 3.0040740966796875e-05, "grad_norm": 2.1443662643432617, "learning_rate": 8.753709198813058e-06, "loss": 0.5585, "mean_token_accuracy": 0.8297072052955627, "num_tokens": 78630535.0, "step": 2066 }, { "epoch": 0.2629436458465844, "ewc_loss": 0.037841796875, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 2.36248517036438, "learning_rate": 8.757948283170837e-06, "loss": 0.4651, "mean_token_accuracy": 0.8612661957740784, "num_tokens": 78672486.0, "step": 2067 }, { "epoch": 0.2630708561251749, "ewc_loss": 0.03515625, "ewc_loss_diag": 5.364418029785156e-05, "ewc_loss_parallel": 2.9802322387695312e-05, "grad_norm": 2.1485862731933594, "learning_rate": 8.762187367528614e-06, "loss": 0.455, "mean_token_accuracy": 0.859498143196106, "num_tokens": 78710208.0, "step": 2068 }, { "epoch": 0.2631980664037654, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.463411569595337, "learning_rate": 8.766426451886393e-06, "loss": 0.5081, "mean_token_accuracy": 0.8437525033950806, "num_tokens": 78754233.0, "step": 2069 }, { "epoch": 0.26332527668235595, "ewc_loss": 0.03564453125, "ewc_loss_diag": 5.412101745605469e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 2.1610827445983887, "learning_rate": 8.770665536244172e-06, "loss": 0.5402, "mean_token_accuracy": 0.8391519784927368, "num_tokens": 78794530.0, "step": 2070 }, { "epoch": 0.2634524869609464, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.3995823860168457, "learning_rate": 8.77490462060195e-06, "loss": 0.4599, "mean_token_accuracy": 0.8595524430274963, "num_tokens": 78828620.0, "step": 2071 }, { "epoch": 0.26357969723953695, "ewc_loss": 0.03564453125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 2.159390687942505, "learning_rate": 8.77914370495973e-06, "loss": 0.4572, "mean_token_accuracy": 0.858407199382782, "num_tokens": 78867658.0, "step": 2072 }, { "epoch": 0.2637069075181275, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.4128191471099854, "learning_rate": 8.783382789317508e-06, "loss": 0.5046, "mean_token_accuracy": 0.8471987247467041, "num_tokens": 78903744.0, "step": 2073 }, { "epoch": 0.26383411779671795, "ewc_loss": 0.036376953125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 2.199697494506836, "learning_rate": 8.787621873675287e-06, "loss": 0.5246, "mean_token_accuracy": 0.8374277353286743, "num_tokens": 78934098.0, "step": 2074 }, { "epoch": 0.2639613280753085, "ewc_loss": 0.038330078125, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.375767946243286, "learning_rate": 8.791860958033066e-06, "loss": 0.4643, "mean_token_accuracy": 0.8572537899017334, "num_tokens": 78969473.0, "step": 2075 }, { "epoch": 0.264088538353899, "ewc_loss": 0.035888671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.18922758102417, "learning_rate": 8.796100042390843e-06, "loss": 0.4593, "mean_token_accuracy": 0.8608297109603882, "num_tokens": 79008799.0, "step": 2076 }, { "epoch": 0.2642157486324895, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.3725950717926025, "learning_rate": 8.800339126748622e-06, "loss": 0.4683, "mean_token_accuracy": 0.8559244871139526, "num_tokens": 79042502.0, "step": 2077 }, { "epoch": 0.26434295891108, "ewc_loss": 0.035888671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.1925718784332275, "learning_rate": 8.804578211106401e-06, "loss": 0.5366, "mean_token_accuracy": 0.8333635926246643, "num_tokens": 79079641.0, "step": 2078 }, { "epoch": 0.26447016918967053, "ewc_loss": 0.037353515625, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.3469736576080322, "learning_rate": 8.80881729546418e-06, "loss": 0.4863, "mean_token_accuracy": 0.8533764481544495, "num_tokens": 79116120.0, "step": 2079 }, { "epoch": 0.26459737946826106, "ewc_loss": 0.035888671875, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.194772720336914, "learning_rate": 8.813056379821959e-06, "loss": 0.4671, "mean_token_accuracy": 0.8601491451263428, "num_tokens": 79150322.0, "step": 2080 }, { "epoch": 0.26472458974685154, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.435943603515625e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.342825412750244, "learning_rate": 8.817295464179738e-06, "loss": 0.4943, "mean_token_accuracy": 0.8473258018493652, "num_tokens": 79184944.0, "step": 2081 }, { "epoch": 0.26485180002544206, "ewc_loss": 0.03564453125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 2.179161310195923, "learning_rate": 8.821534548537517e-06, "loss": 0.5594, "mean_token_accuracy": 0.8256255388259888, "num_tokens": 79226137.0, "step": 2082 }, { "epoch": 0.2649790103040326, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.314861536026001, "learning_rate": 8.825773632895296e-06, "loss": 0.4683, "mean_token_accuracy": 0.8562135696411133, "num_tokens": 79267721.0, "step": 2083 }, { "epoch": 0.26510622058262306, "ewc_loss": 0.035888671875, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.039836883544922e-05, "grad_norm": 2.1801676750183105, "learning_rate": 8.830012717253074e-06, "loss": 0.4712, "mean_token_accuracy": 0.856116533279419, "num_tokens": 79305355.0, "step": 2084 }, { "epoch": 0.2652334308612136, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.3064141273498535, "learning_rate": 8.834251801610853e-06, "loss": 0.5063, "mean_token_accuracy": 0.8443623781204224, "num_tokens": 79349714.0, "step": 2085 }, { "epoch": 0.2653606411398041, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 2.1788718700408936, "learning_rate": 8.838490885968632e-06, "loss": 0.4567, "mean_token_accuracy": 0.8611088991165161, "num_tokens": 79387208.0, "step": 2086 }, { "epoch": 0.2654878514183946, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.3037703037261963, "learning_rate": 8.842729970326411e-06, "loss": 0.4488, "mean_token_accuracy": 0.8667595386505127, "num_tokens": 79424396.0, "step": 2087 }, { "epoch": 0.2656150616969851, "ewc_loss": 0.035888671875, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.039836883544922e-05, "grad_norm": 2.1880316734313965, "learning_rate": 8.846969054684188e-06, "loss": 0.5676, "mean_token_accuracy": 0.8305679559707642, "num_tokens": 79465105.0, "step": 2088 }, { "epoch": 0.26574227197557565, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.3025624752044678, "learning_rate": 8.851208139041967e-06, "loss": 0.4578, "mean_token_accuracy": 0.8591738939285278, "num_tokens": 79504000.0, "step": 2089 }, { "epoch": 0.2658694822541661, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.2091920375823975, "learning_rate": 8.855447223399746e-06, "loss": 0.5164, "mean_token_accuracy": 0.8405475616455078, "num_tokens": 79538954.0, "step": 2090 }, { "epoch": 0.26599669253275665, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.285351514816284, "learning_rate": 8.859686307757525e-06, "loss": 0.5376, "mean_token_accuracy": 0.8362392783164978, "num_tokens": 79577264.0, "step": 2091 }, { "epoch": 0.2661239028113472, "ewc_loss": 0.035888671875, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.039836883544922e-05, "grad_norm": 2.1906280517578125, "learning_rate": 8.863925392115304e-06, "loss": 0.4818, "mean_token_accuracy": 0.8518678545951843, "num_tokens": 79612803.0, "step": 2092 }, { "epoch": 0.26625111308993765, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.258444309234619, "learning_rate": 8.868164476473083e-06, "loss": 0.4367, "mean_token_accuracy": 0.866183876991272, "num_tokens": 79651595.0, "step": 2093 }, { "epoch": 0.2663783233685282, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.1885228157043457, "learning_rate": 8.872403560830862e-06, "loss": 0.4739, "mean_token_accuracy": 0.8565741777420044, "num_tokens": 79691587.0, "step": 2094 }, { "epoch": 0.2665055336471187, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.507469177246094e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.2700421810150146, "learning_rate": 8.87664264518864e-06, "loss": 0.5266, "mean_token_accuracy": 0.8395030498504639, "num_tokens": 79729803.0, "step": 2095 }, { "epoch": 0.2666327439257092, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.2025389671325684, "learning_rate": 8.88088172954642e-06, "loss": 0.4446, "mean_token_accuracy": 0.8638450503349304, "num_tokens": 79765866.0, "step": 2096 }, { "epoch": 0.2667599542042997, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.53131103515625e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.244196653366089, "learning_rate": 8.885120813904198e-06, "loss": 0.4554, "mean_token_accuracy": 0.8599114418029785, "num_tokens": 79808973.0, "step": 2097 }, { "epoch": 0.26688716448289024, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.2144930362701416, "learning_rate": 8.889359898261977e-06, "loss": 0.555, "mean_token_accuracy": 0.8324068784713745, "num_tokens": 79843677.0, "step": 2098 }, { "epoch": 0.2670143747614807, "ewc_loss": 0.03662109375, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 2.243706464767456, "learning_rate": 8.893598982619754e-06, "loss": 0.4701, "mean_token_accuracy": 0.8577079772949219, "num_tokens": 79882680.0, "step": 2099 }, { "epoch": 0.26714158504007124, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.194735527038574, "learning_rate": 8.897838066977533e-06, "loss": 0.484, "mean_token_accuracy": 0.8525426387786865, "num_tokens": 79929691.0, "step": 2100 }, { "epoch": 0.26726879531866177, "ewc_loss": 0.03662109375, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 2.238359212875366, "learning_rate": 8.902077151335312e-06, "loss": 0.4822, "mean_token_accuracy": 0.8516595363616943, "num_tokens": 79970515.0, "step": 2101 }, { "epoch": 0.26739600559725224, "ewc_loss": 0.035888671875, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 3.039836883544922e-05, "grad_norm": 2.2175803184509277, "learning_rate": 8.906316235693091e-06, "loss": 0.5741, "mean_token_accuracy": 0.8195424675941467, "num_tokens": 80007550.0, "step": 2102 }, { "epoch": 0.26752321587584277, "ewc_loss": 0.03662109375, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 2.2413294315338135, "learning_rate": 8.91055532005087e-06, "loss": 0.4899, "mean_token_accuracy": 0.8476901650428772, "num_tokens": 80044106.0, "step": 2103 }, { "epoch": 0.2676504261544333, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.20754075050354, "learning_rate": 8.914794404408649e-06, "loss": 0.4894, "mean_token_accuracy": 0.8510953187942505, "num_tokens": 80084429.0, "step": 2104 }, { "epoch": 0.26777763643302377, "ewc_loss": 0.03662109375, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 2.225693464279175, "learning_rate": 8.919033488766428e-06, "loss": 0.4269, "mean_token_accuracy": 0.8674908876419067, "num_tokens": 80123317.0, "step": 2105 }, { "epoch": 0.2679048467116143, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.227041244506836, "learning_rate": 8.923272573124205e-06, "loss": 0.5161, "mean_token_accuracy": 0.8467725515365601, "num_tokens": 80157658.0, "step": 2106 }, { "epoch": 0.2680320569902048, "ewc_loss": 0.03662109375, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 2.2301323413848877, "learning_rate": 8.927511657481984e-06, "loss": 0.5253, "mean_token_accuracy": 0.8383527398109436, "num_tokens": 80200482.0, "step": 2107 }, { "epoch": 0.2681592672687953, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.237288475036621, "learning_rate": 8.931750741839763e-06, "loss": 0.4513, "mean_token_accuracy": 0.8610032796859741, "num_tokens": 80231844.0, "step": 2108 }, { "epoch": 0.2682864775473858, "ewc_loss": 0.036376953125, "ewc_loss_diag": 5.555152893066406e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 2.2291979789733887, "learning_rate": 8.935989826197541e-06, "loss": 0.4727, "mean_token_accuracy": 0.8556811809539795, "num_tokens": 80269734.0, "step": 2109 }, { "epoch": 0.26841368782597635, "ewc_loss": 0.036376953125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 2.350005626678467, "learning_rate": 8.94022891055532e-06, "loss": 0.4433, "mean_token_accuracy": 0.8672896027565002, "num_tokens": 80306309.0, "step": 2110 }, { "epoch": 0.2685408981045668, "ewc_loss": 0.035888671875, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 3.039836883544922e-05, "grad_norm": 2.1791157722473145, "learning_rate": 8.9444679949131e-06, "loss": 0.4977, "mean_token_accuracy": 0.8447951078414917, "num_tokens": 80345390.0, "step": 2111 }, { "epoch": 0.26866810838315736, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.2851336002349854, "learning_rate": 8.948707079270878e-06, "loss": 0.4177, "mean_token_accuracy": 0.8727319240570068, "num_tokens": 80384118.0, "step": 2112 }, { "epoch": 0.2687953186617479, "ewc_loss": 0.035400390625, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 2.9921531677246094e-05, "grad_norm": 2.177985906600952, "learning_rate": 8.952946163628657e-06, "loss": 0.4735, "mean_token_accuracy": 0.8548781871795654, "num_tokens": 80416703.0, "step": 2113 }, { "epoch": 0.26892252894033836, "ewc_loss": 0.037353515625, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.3030011653900146, "learning_rate": 8.957185247986436e-06, "loss": 0.426, "mean_token_accuracy": 0.8703120350837708, "num_tokens": 80453732.0, "step": 2114 }, { "epoch": 0.2690497392189289, "ewc_loss": 0.03564453125, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 3.0159950256347656e-05, "grad_norm": 2.170146942138672, "learning_rate": 8.961424332344215e-06, "loss": 0.4791, "mean_token_accuracy": 0.8537455797195435, "num_tokens": 80495265.0, "step": 2115 }, { "epoch": 0.2691769494975194, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.5789947509765625e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.3221864700317383, "learning_rate": 8.965663416701994e-06, "loss": 0.4723, "mean_token_accuracy": 0.8539634346961975, "num_tokens": 80533789.0, "step": 2116 }, { "epoch": 0.2693041597761099, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 3.039836883544922e-05, "grad_norm": 2.172860622406006, "learning_rate": 8.969902501059773e-06, "loss": 0.4471, "mean_token_accuracy": 0.8644811511039734, "num_tokens": 80570802.0, "step": 2117 }, { "epoch": 0.2694313700547004, "ewc_loss": 0.037841796875, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.3277227878570557, "learning_rate": 8.974141585417551e-06, "loss": 0.4191, "mean_token_accuracy": 0.8710547685623169, "num_tokens": 80608031.0, "step": 2118 }, { "epoch": 0.26955858033329094, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.1702895164489746, "learning_rate": 8.97838066977533e-06, "loss": 0.5056, "mean_token_accuracy": 0.8451735973358154, "num_tokens": 80646271.0, "step": 2119 }, { "epoch": 0.2696857906118814, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 2.351062774658203, "learning_rate": 8.98261975413311e-06, "loss": 0.5051, "mean_token_accuracy": 0.8454908132553101, "num_tokens": 80683325.0, "step": 2120 }, { "epoch": 0.26981300089047194, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 3.039836883544922e-05, "grad_norm": 2.1563291549682617, "learning_rate": 8.986858838490888e-06, "loss": 0.5116, "mean_token_accuracy": 0.8454457521438599, "num_tokens": 80726403.0, "step": 2121 }, { "epoch": 0.26994021116906247, "ewc_loss": 0.038330078125, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.3650150299072266, "learning_rate": 8.991097922848665e-06, "loss": 0.4736, "mean_token_accuracy": 0.8559752702713013, "num_tokens": 80767293.0, "step": 2122 }, { "epoch": 0.27006742144765294, "ewc_loss": 0.035888671875, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 3.0279159545898438e-05, "grad_norm": 2.1692001819610596, "learning_rate": 8.995337007206444e-06, "loss": 0.5007, "mean_token_accuracy": 0.8459383249282837, "num_tokens": 80806218.0, "step": 2123 }, { "epoch": 0.2701946317262435, "ewc_loss": 0.038818359375, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.3897287845611572, "learning_rate": 8.999576091564223e-06, "loss": 0.4702, "mean_token_accuracy": 0.8583725094795227, "num_tokens": 80841694.0, "step": 2124 }, { "epoch": 0.270321842004834, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.169080972671509, "learning_rate": 9.003815175922002e-06, "loss": 0.473, "mean_token_accuracy": 0.853020191192627, "num_tokens": 80874879.0, "step": 2125 }, { "epoch": 0.2704490522834245, "ewc_loss": 0.038818359375, "ewc_loss_diag": 5.602836608886719e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.3950819969177246, "learning_rate": 9.00805426027978e-06, "loss": 0.4915, "mean_token_accuracy": 0.8512575626373291, "num_tokens": 80916317.0, "step": 2126 }, { "epoch": 0.270576262562015, "ewc_loss": 0.0361328125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 3.0517578125e-05, "grad_norm": 2.1794636249542236, "learning_rate": 9.01229334463756e-06, "loss": 0.5122, "mean_token_accuracy": 0.8446028232574463, "num_tokens": 80952304.0, "step": 2127 }, { "epoch": 0.27070347284060553, "ewc_loss": 0.038818359375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.3977179527282715, "learning_rate": 9.016532428995339e-06, "loss": 0.4561, "mean_token_accuracy": 0.8611776828765869, "num_tokens": 80989886.0, "step": 2128 }, { "epoch": 0.27083068311919606, "ewc_loss": 0.036376953125, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 3.075599670410156e-05, "grad_norm": 2.1724600791931152, "learning_rate": 9.020771513353116e-06, "loss": 0.5073, "mean_token_accuracy": 0.8434985876083374, "num_tokens": 81031041.0, "step": 2129 }, { "epoch": 0.27095789339778653, "ewc_loss": 0.038818359375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.4171736240386963, "learning_rate": 9.025010597710895e-06, "loss": 0.5022, "mean_token_accuracy": 0.8463405966758728, "num_tokens": 81068915.0, "step": 2130 }, { "epoch": 0.27108510367637706, "ewc_loss": 0.03662109375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 3.0994415283203125e-05, "grad_norm": 2.174440622329712, "learning_rate": 9.029249682068674e-06, "loss": 0.4648, "mean_token_accuracy": 0.8598382472991943, "num_tokens": 81106044.0, "step": 2131 }, { "epoch": 0.2712123139549676, "ewc_loss": 0.0390625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.4230453968048096, "learning_rate": 9.033488766426452e-06, "loss": 0.4764, "mean_token_accuracy": 0.8541147112846375, "num_tokens": 81142589.0, "step": 2132 }, { "epoch": 0.27133952423355806, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.183201313018799, "learning_rate": 9.037727850784231e-06, "loss": 0.4524, "mean_token_accuracy": 0.8615013360977173, "num_tokens": 81178415.0, "step": 2133 }, { "epoch": 0.2714667345121486, "ewc_loss": 0.0390625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.4447402954101562, "learning_rate": 9.04196693514201e-06, "loss": 0.4972, "mean_token_accuracy": 0.8497787714004517, "num_tokens": 81216952.0, "step": 2134 }, { "epoch": 0.2715939447907391, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.1872968673706055, "learning_rate": 9.046206019499789e-06, "loss": 0.4434, "mean_token_accuracy": 0.8659665584564209, "num_tokens": 81257879.0, "step": 2135 }, { "epoch": 0.2717211550693296, "ewc_loss": 0.039306640625, "ewc_loss_diag": 5.626678466796875e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.4523284435272217, "learning_rate": 9.050445103857568e-06, "loss": 0.5207, "mean_token_accuracy": 0.8448755145072937, "num_tokens": 81298045.0, "step": 2136 }, { "epoch": 0.2718483653479201, "ewc_loss": 0.037353515625, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.208822250366211, "learning_rate": 9.054684188215345e-06, "loss": 0.4335, "mean_token_accuracy": 0.869620680809021, "num_tokens": 81332342.0, "step": 2137 }, { "epoch": 0.27197557562651065, "ewc_loss": 0.039306640625, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.4561591148376465, "learning_rate": 9.058923272573124e-06, "loss": 0.5108, "mean_token_accuracy": 0.8433405160903931, "num_tokens": 81367270.0, "step": 2138 }, { "epoch": 0.2721027859051011, "ewc_loss": 0.037353515625, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.1935722827911377, "learning_rate": 9.063162356930903e-06, "loss": 0.4586, "mean_token_accuracy": 0.8614330887794495, "num_tokens": 81406644.0, "step": 2139 }, { "epoch": 0.27222999618369165, "ewc_loss": 0.03955078125, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.4542460441589355, "learning_rate": 9.067401441288682e-06, "loss": 0.4942, "mean_token_accuracy": 0.8526386022567749, "num_tokens": 81443225.0, "step": 2140 }, { "epoch": 0.2723572064622822, "ewc_loss": 0.037353515625, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.205481767654419, "learning_rate": 9.07164052564646e-06, "loss": 0.4604, "mean_token_accuracy": 0.8546544909477234, "num_tokens": 81482168.0, "step": 2141 }, { "epoch": 0.27248441674087265, "ewc_loss": 0.03955078125, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.4563894271850586, "learning_rate": 9.07587961000424e-06, "loss": 0.5051, "mean_token_accuracy": 0.8453013896942139, "num_tokens": 81518127.0, "step": 2142 }, { "epoch": 0.2726116270194632, "ewc_loss": 0.037353515625, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.238417148590088, "learning_rate": 9.080118694362018e-06, "loss": 0.5039, "mean_token_accuracy": 0.8432029485702515, "num_tokens": 81559857.0, "step": 2143 }, { "epoch": 0.2727388372980537, "ewc_loss": 0.03955078125, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.471421480178833, "learning_rate": 9.084357778719797e-06, "loss": 0.4984, "mean_token_accuracy": 0.8465894460678101, "num_tokens": 81596509.0, "step": 2144 }, { "epoch": 0.2728660475766442, "ewc_loss": 0.037353515625, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.2311387062072754, "learning_rate": 9.088596863077576e-06, "loss": 0.475, "mean_token_accuracy": 0.8584515452384949, "num_tokens": 81632693.0, "step": 2145 }, { "epoch": 0.2729932578552347, "ewc_loss": 0.039306640625, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.4352622032165527, "learning_rate": 9.092835947435355e-06, "loss": 0.4887, "mean_token_accuracy": 0.8500818610191345, "num_tokens": 81670067.0, "step": 2146 }, { "epoch": 0.27312046813382523, "ewc_loss": 0.037353515625, "ewc_loss_diag": 5.650520324707031e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.242901086807251, "learning_rate": 9.097075031793134e-06, "loss": 0.5288, "mean_token_accuracy": 0.8380274772644043, "num_tokens": 81706902.0, "step": 2147 }, { "epoch": 0.2732476784124157, "ewc_loss": 0.0390625, "ewc_loss_diag": 5.6743621826171875e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.407543182373047, "learning_rate": 9.101314116150913e-06, "loss": 0.4428, "mean_token_accuracy": 0.8677439093589783, "num_tokens": 81743340.0, "step": 2148 }, { "epoch": 0.27337488869100623, "ewc_loss": 0.037353515625, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.2459335327148438, "learning_rate": 9.105553200508692e-06, "loss": 0.4571, "mean_token_accuracy": 0.861394464969635, "num_tokens": 81779573.0, "step": 2149 }, { "epoch": 0.27350209896959676, "ewc_loss": 0.038818359375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.3790199756622314, "learning_rate": 9.10979228486647e-06, "loss": 0.4856, "mean_token_accuracy": 0.8533850908279419, "num_tokens": 81822004.0, "step": 2150 }, { "epoch": 0.27362930924818724, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.2343642711639404, "learning_rate": 9.11403136922425e-06, "loss": 0.4863, "mean_token_accuracy": 0.8519635200500488, "num_tokens": 81860472.0, "step": 2151 }, { "epoch": 0.27375651952677776, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.381401300430298, "learning_rate": 9.118270453582027e-06, "loss": 0.5467, "mean_token_accuracy": 0.8385040760040283, "num_tokens": 81899151.0, "step": 2152 }, { "epoch": 0.2738837298053683, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.218402624130249, "learning_rate": 9.122509537939806e-06, "loss": 0.4708, "mean_token_accuracy": 0.8521524667739868, "num_tokens": 81941213.0, "step": 2153 }, { "epoch": 0.27401094008395877, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.3868026733398438, "learning_rate": 9.126748622297584e-06, "loss": 0.519, "mean_token_accuracy": 0.8366348743438721, "num_tokens": 81978055.0, "step": 2154 }, { "epoch": 0.2741381503625493, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.230748176574707, "learning_rate": 9.130987706655363e-06, "loss": 0.5074, "mean_token_accuracy": 0.8446528911590576, "num_tokens": 82015599.0, "step": 2155 }, { "epoch": 0.2742653606411398, "ewc_loss": 0.038330078125, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.373967170715332, "learning_rate": 9.135226791013142e-06, "loss": 0.5339, "mean_token_accuracy": 0.8441746234893799, "num_tokens": 82055113.0, "step": 2156 }, { "epoch": 0.2743925709197303, "ewc_loss": 0.036865234375, "ewc_loss_diag": 5.698204040527344e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 2.23724627494812, "learning_rate": 9.139465875370921e-06, "loss": 0.4774, "mean_token_accuracy": 0.8515714406967163, "num_tokens": 82083052.0, "step": 2157 }, { "epoch": 0.2745197811983208, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.37099289894104, "learning_rate": 9.1437049597287e-06, "loss": 0.4962, "mean_token_accuracy": 0.8460484743118286, "num_tokens": 82118850.0, "step": 2158 }, { "epoch": 0.27464699147691135, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 2.2430953979492188, "learning_rate": 9.147944044086477e-06, "loss": 0.5154, "mean_token_accuracy": 0.8417200446128845, "num_tokens": 82151261.0, "step": 2159 }, { "epoch": 0.2747742017555018, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.3560750484466553, "learning_rate": 9.152183128444256e-06, "loss": 0.5199, "mean_token_accuracy": 0.8479429483413696, "num_tokens": 82192018.0, "step": 2160 }, { "epoch": 0.27490141203409235, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.2247016429901123, "learning_rate": 9.156422212802035e-06, "loss": 0.5274, "mean_token_accuracy": 0.8343837261199951, "num_tokens": 82231044.0, "step": 2161 }, { "epoch": 0.2750286223126829, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.3616678714752197, "learning_rate": 9.160661297159814e-06, "loss": 0.4586, "mean_token_accuracy": 0.8599470257759094, "num_tokens": 82265334.0, "step": 2162 }, { "epoch": 0.27515583259127335, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.7220458984375e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.2202138900756836, "learning_rate": 9.164900381517593e-06, "loss": 0.4725, "mean_token_accuracy": 0.8529521226882935, "num_tokens": 82303995.0, "step": 2163 }, { "epoch": 0.2752830428698639, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.3542263507843018, "learning_rate": 9.169139465875372e-06, "loss": 0.4831, "mean_token_accuracy": 0.8517647981643677, "num_tokens": 82345157.0, "step": 2164 }, { "epoch": 0.2754102531484544, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.230442762374878, "learning_rate": 9.17337855023315e-06, "loss": 0.488, "mean_token_accuracy": 0.8495588302612305, "num_tokens": 82380723.0, "step": 2165 }, { "epoch": 0.2755374634270449, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.35263991355896, "learning_rate": 9.17761763459093e-06, "loss": 0.4479, "mean_token_accuracy": 0.862844705581665, "num_tokens": 82418985.0, "step": 2166 }, { "epoch": 0.2756646737056354, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.2173211574554443, "learning_rate": 9.181856718948707e-06, "loss": 0.4261, "mean_token_accuracy": 0.8686418533325195, "num_tokens": 82455498.0, "step": 2167 }, { "epoch": 0.27579188398422594, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.3646490573883057, "learning_rate": 9.186095803306485e-06, "loss": 0.4706, "mean_token_accuracy": 0.8549133539199829, "num_tokens": 82491117.0, "step": 2168 }, { "epoch": 0.2759190942628164, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.2211318016052246, "learning_rate": 9.190334887664264e-06, "loss": 0.5171, "mean_token_accuracy": 0.8436141610145569, "num_tokens": 82533640.0, "step": 2169 }, { "epoch": 0.27604630454140694, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.3882429599761963, "learning_rate": 9.194573972022043e-06, "loss": 0.5727, "mean_token_accuracy": 0.8326842784881592, "num_tokens": 82568812.0, "step": 2170 }, { "epoch": 0.27617351481999747, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 2.2061548233032227, "learning_rate": 9.198813056379822e-06, "loss": 0.5598, "mean_token_accuracy": 0.8294222354888916, "num_tokens": 82611777.0, "step": 2171 }, { "epoch": 0.27630072509858794, "ewc_loss": 0.0390625, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.403095245361328, "learning_rate": 9.203052140737601e-06, "loss": 0.5074, "mean_token_accuracy": 0.8464998006820679, "num_tokens": 82646938.0, "step": 2172 }, { "epoch": 0.27642793537717847, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.2029924392700195, "learning_rate": 9.20729122509538e-06, "loss": 0.4461, "mean_token_accuracy": 0.864292323589325, "num_tokens": 82684316.0, "step": 2173 }, { "epoch": 0.276555145655769, "ewc_loss": 0.0390625, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.41084885597229, "learning_rate": 9.211530309453159e-06, "loss": 0.4914, "mean_token_accuracy": 0.8473008871078491, "num_tokens": 82723204.0, "step": 2174 }, { "epoch": 0.27668235593435947, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.205110549926758, "learning_rate": 9.215769393810938e-06, "loss": 0.4849, "mean_token_accuracy": 0.8531083464622498, "num_tokens": 82759634.0, "step": 2175 }, { "epoch": 0.27680956621295, "ewc_loss": 0.0390625, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.431291103363037, "learning_rate": 9.220008478168717e-06, "loss": 0.5767, "mean_token_accuracy": 0.8243556022644043, "num_tokens": 82800446.0, "step": 2176 }, { "epoch": 0.2769367764915405, "ewc_loss": 0.037109375, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.123283386230469e-05, "grad_norm": 2.195096969604492, "learning_rate": 9.224247562526495e-06, "loss": 0.49, "mean_token_accuracy": 0.8488646745681763, "num_tokens": 82840754.0, "step": 2177 }, { "epoch": 0.277063986770131, "ewc_loss": 0.03955078125, "ewc_loss_diag": 5.745887756347656e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.445866823196411, "learning_rate": 9.228486646884274e-06, "loss": 0.4917, "mean_token_accuracy": 0.8512554168701172, "num_tokens": 82881859.0, "step": 2178 }, { "epoch": 0.2771911970487215, "ewc_loss": 0.037353515625, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 3.147125244140625e-05, "grad_norm": 2.20210337638855, "learning_rate": 9.232725731242053e-06, "loss": 0.5318, "mean_token_accuracy": 0.8364838361740112, "num_tokens": 82922146.0, "step": 2179 }, { "epoch": 0.27731840732731206, "ewc_loss": 0.039794921875, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.476067066192627, "learning_rate": 9.236964815599832e-06, "loss": 0.525, "mean_token_accuracy": 0.8407266139984131, "num_tokens": 82960017.0, "step": 2180 }, { "epoch": 0.2774456176059026, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.1893880367279053, "learning_rate": 9.241203899957611e-06, "loss": 0.4612, "mean_token_accuracy": 0.8586641550064087, "num_tokens": 83001778.0, "step": 2181 }, { "epoch": 0.27757282788449306, "ewc_loss": 0.040283203125, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.503373622894287, "learning_rate": 9.245442984315388e-06, "loss": 0.4771, "mean_token_accuracy": 0.8547168970108032, "num_tokens": 83037023.0, "step": 2182 }, { "epoch": 0.2777000381630836, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.2239272594451904, "learning_rate": 9.249682068673167e-06, "loss": 0.5123, "mean_token_accuracy": 0.8430241942405701, "num_tokens": 83075355.0, "step": 2183 }, { "epoch": 0.2778272484416741, "ewc_loss": 0.040283203125, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.503282070159912, "learning_rate": 9.253921153030946e-06, "loss": 0.5498, "mean_token_accuracy": 0.834187388420105, "num_tokens": 83116233.0, "step": 2184 }, { "epoch": 0.2779544587202646, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.7697296142578125e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.232379198074341, "learning_rate": 9.258160237388725e-06, "loss": 0.4645, "mean_token_accuracy": 0.861378014087677, "num_tokens": 83152238.0, "step": 2185 }, { "epoch": 0.2780816689988551, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.50260066986084, "learning_rate": 9.262399321746504e-06, "loss": 0.4727, "mean_token_accuracy": 0.8544561862945557, "num_tokens": 83186779.0, "step": 2186 }, { "epoch": 0.27820887927744564, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.2529714107513428, "learning_rate": 9.266638406104283e-06, "loss": 0.5322, "mean_token_accuracy": 0.8364776372909546, "num_tokens": 83226001.0, "step": 2187 }, { "epoch": 0.2783360895560361, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.793571472167969e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.5008413791656494, "learning_rate": 9.270877490462061e-06, "loss": 0.4945, "mean_token_accuracy": 0.8496766686439514, "num_tokens": 83257196.0, "step": 2188 }, { "epoch": 0.27846329983462664, "ewc_loss": 0.038330078125, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 2.273422956466675, "learning_rate": 9.27511657481984e-06, "loss": 0.497, "mean_token_accuracy": 0.8501578569412231, "num_tokens": 83293261.0, "step": 2189 }, { "epoch": 0.27859051011321717, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.4886138439178467, "learning_rate": 9.279355659177618e-06, "loss": 0.5119, "mean_token_accuracy": 0.8492439985275269, "num_tokens": 83327096.0, "step": 2190 }, { "epoch": 0.27871772039180764, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.2741715908050537, "learning_rate": 9.283594743535396e-06, "loss": 0.5274, "mean_token_accuracy": 0.8413830995559692, "num_tokens": 83360558.0, "step": 2191 }, { "epoch": 0.2788449306703982, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.4561049938201904, "learning_rate": 9.287833827893175e-06, "loss": 0.4937, "mean_token_accuracy": 0.8473902344703674, "num_tokens": 83400563.0, "step": 2192 }, { "epoch": 0.2789721409489887, "ewc_loss": 0.037841796875, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.272094488143921, "learning_rate": 9.292072912250954e-06, "loss": 0.5021, "mean_token_accuracy": 0.8454942107200623, "num_tokens": 83432961.0, "step": 2193 }, { "epoch": 0.2790993512275792, "ewc_loss": 0.039306640625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.4233312606811523, "learning_rate": 9.296311996608733e-06, "loss": 0.4384, "mean_token_accuracy": 0.863582968711853, "num_tokens": 83472183.0, "step": 2194 }, { "epoch": 0.2792265615061697, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.817413330078125e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.276441812515259, "learning_rate": 9.300551080966512e-06, "loss": 0.56, "mean_token_accuracy": 0.8272430896759033, "num_tokens": 83512376.0, "step": 2195 }, { "epoch": 0.27935377178476023, "ewc_loss": 0.0390625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.4204864501953125, "learning_rate": 9.304790165324291e-06, "loss": 0.4827, "mean_token_accuracy": 0.8499619364738464, "num_tokens": 83546141.0, "step": 2196 }, { "epoch": 0.2794809820633507, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.2565560340881348, "learning_rate": 9.30902924968207e-06, "loss": 0.4818, "mean_token_accuracy": 0.8541561365127563, "num_tokens": 83584223.0, "step": 2197 }, { "epoch": 0.27960819234194123, "ewc_loss": 0.0390625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.416210412979126, "learning_rate": 9.313268334039847e-06, "loss": 0.505, "mean_token_accuracy": 0.8455461263656616, "num_tokens": 83621773.0, "step": 2198 }, { "epoch": 0.27973540262053176, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.2402946949005127, "learning_rate": 9.317507418397626e-06, "loss": 0.4263, "mean_token_accuracy": 0.8709064722061157, "num_tokens": 83659129.0, "step": 2199 }, { "epoch": 0.27986261289912223, "ewc_loss": 0.03955078125, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.42549729347229, "learning_rate": 9.321746502755405e-06, "loss": 0.5352, "mean_token_accuracy": 0.839840829372406, "num_tokens": 83699146.0, "step": 2200 }, { "epoch": 0.27998982317771276, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.2605984210968018, "learning_rate": 9.325985587113184e-06, "loss": 0.4955, "mean_token_accuracy": 0.8497163653373718, "num_tokens": 83733266.0, "step": 2201 }, { "epoch": 0.2801170334563033, "ewc_loss": 0.039306640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.433253526687622, "learning_rate": 9.330224671470962e-06, "loss": 0.5017, "mean_token_accuracy": 0.8543287515640259, "num_tokens": 83766610.0, "step": 2202 }, { "epoch": 0.28024424373489376, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.2681374549865723, "learning_rate": 9.334463755828741e-06, "loss": 0.5179, "mean_token_accuracy": 0.8379625082015991, "num_tokens": 83797891.0, "step": 2203 }, { "epoch": 0.2803714540134843, "ewc_loss": 0.039306640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.408766508102417, "learning_rate": 9.33870284018652e-06, "loss": 0.519, "mean_token_accuracy": 0.8386151790618896, "num_tokens": 83837046.0, "step": 2204 }, { "epoch": 0.2804986642920748, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.2459421157836914, "learning_rate": 9.342941924544299e-06, "loss": 0.5109, "mean_token_accuracy": 0.8435385227203369, "num_tokens": 83876457.0, "step": 2205 }, { "epoch": 0.2806258745706653, "ewc_loss": 0.03955078125, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.425867795944214, "learning_rate": 9.347181008902078e-06, "loss": 0.4516, "mean_token_accuracy": 0.8654850125312805, "num_tokens": 83912729.0, "step": 2206 }, { "epoch": 0.2807530848492558, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.255431652069092, "learning_rate": 9.351420093259857e-06, "loss": 0.4498, "mean_token_accuracy": 0.8626507520675659, "num_tokens": 83945958.0, "step": 2207 }, { "epoch": 0.28088029512784635, "ewc_loss": 0.039306640625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.4237172603607178, "learning_rate": 9.355659177617636e-06, "loss": 0.4393, "mean_token_accuracy": 0.8628315925598145, "num_tokens": 83976767.0, "step": 2208 }, { "epoch": 0.2810075054064368, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.2277424335479736, "learning_rate": 9.359898261975415e-06, "loss": 0.4332, "mean_token_accuracy": 0.8673996925354004, "num_tokens": 84017996.0, "step": 2209 }, { "epoch": 0.28113471568502735, "ewc_loss": 0.03955078125, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.4432499408721924, "learning_rate": 9.364137346333194e-06, "loss": 0.5644, "mean_token_accuracy": 0.8323777318000793, "num_tokens": 84057528.0, "step": 2210 }, { "epoch": 0.2812619259636179, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.236637830734253, "learning_rate": 9.368376430690972e-06, "loss": 0.4472, "mean_token_accuracy": 0.8620940446853638, "num_tokens": 84091009.0, "step": 2211 }, { "epoch": 0.28138913624220835, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.471266031265259, "learning_rate": 9.372615515048751e-06, "loss": 0.5255, "mean_token_accuracy": 0.8405264616012573, "num_tokens": 84125042.0, "step": 2212 }, { "epoch": 0.2815163465207989, "ewc_loss": 0.03759765625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.170967102050781e-05, "grad_norm": 2.226533889770508, "learning_rate": 9.376854599406528e-06, "loss": 0.5267, "mean_token_accuracy": 0.8401902318000793, "num_tokens": 84169340.0, "step": 2213 }, { "epoch": 0.2816435567993894, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.4836010932922363, "learning_rate": 9.381093683764307e-06, "loss": 0.4889, "mean_token_accuracy": 0.8487728834152222, "num_tokens": 84205347.0, "step": 2214 }, { "epoch": 0.2817707670779799, "ewc_loss": 0.037841796875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.2272536754608154, "learning_rate": 9.385332768122086e-06, "loss": 0.4746, "mean_token_accuracy": 0.851655125617981, "num_tokens": 84245788.0, "step": 2215 }, { "epoch": 0.2818979773565704, "ewc_loss": 0.040283203125, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.4939870834350586, "learning_rate": 9.389571852479865e-06, "loss": 0.4806, "mean_token_accuracy": 0.8566806316375732, "num_tokens": 84286744.0, "step": 2216 }, { "epoch": 0.28202518763516093, "ewc_loss": 0.037841796875, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.238389015197754, "learning_rate": 9.393810936837644e-06, "loss": 0.4848, "mean_token_accuracy": 0.8494547605514526, "num_tokens": 84326031.0, "step": 2217 }, { "epoch": 0.2821523979137514, "ewc_loss": 0.04052734375, "ewc_loss_diag": 5.841255187988281e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.504138946533203, "learning_rate": 9.398050021195423e-06, "loss": 0.5153, "mean_token_accuracy": 0.8454068899154663, "num_tokens": 84367697.0, "step": 2218 }, { "epoch": 0.28227960819234194, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.2422757148742676, "learning_rate": 9.402289105553202e-06, "loss": 0.4597, "mean_token_accuracy": 0.8575769662857056, "num_tokens": 84410647.0, "step": 2219 }, { "epoch": 0.28240681847093246, "ewc_loss": 0.04052734375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.5186939239501953, "learning_rate": 9.406528189910979e-06, "loss": 0.4873, "mean_token_accuracy": 0.8528348207473755, "num_tokens": 84446839.0, "step": 2220 }, { "epoch": 0.28253402874952294, "ewc_loss": 0.038330078125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 2.2561962604522705, "learning_rate": 9.410767274268758e-06, "loss": 0.4599, "mean_token_accuracy": 0.8577489852905273, "num_tokens": 84485017.0, "step": 2221 }, { "epoch": 0.28266123902811346, "ewc_loss": 0.040771484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.530101776123047, "learning_rate": 9.415006358626537e-06, "loss": 0.5637, "mean_token_accuracy": 0.826097846031189, "num_tokens": 84524518.0, "step": 2222 }, { "epoch": 0.282788449306704, "ewc_loss": 0.038330078125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 2.275839328765869, "learning_rate": 9.419245442984316e-06, "loss": 0.5438, "mean_token_accuracy": 0.83460533618927, "num_tokens": 84559850.0, "step": 2223 }, { "epoch": 0.28291565958529447, "ewc_loss": 0.040771484375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.5204193592071533, "learning_rate": 9.423484527342095e-06, "loss": 0.4851, "mean_token_accuracy": 0.8501133918762207, "num_tokens": 84594332.0, "step": 2224 }, { "epoch": 0.283042869863885, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.282461404800415, "learning_rate": 9.427723611699873e-06, "loss": 0.4896, "mean_token_accuracy": 0.8506044149398804, "num_tokens": 84628759.0, "step": 2225 }, { "epoch": 0.2831700801424755, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.4913530349731445, "learning_rate": 9.431962696057652e-06, "loss": 0.5097, "mean_token_accuracy": 0.8452638387680054, "num_tokens": 84672022.0, "step": 2226 }, { "epoch": 0.283297290421066, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.2785701751708984, "learning_rate": 9.436201780415431e-06, "loss": 0.4539, "mean_token_accuracy": 0.8640917539596558, "num_tokens": 84709266.0, "step": 2227 }, { "epoch": 0.2834245006996565, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.48405122756958, "learning_rate": 9.44044086477321e-06, "loss": 0.5058, "mean_token_accuracy": 0.8482953310012817, "num_tokens": 84745502.0, "step": 2228 }, { "epoch": 0.28355171097824705, "ewc_loss": 0.038330078125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 2.2823803424835205, "learning_rate": 9.444679949130989e-06, "loss": 0.5084, "mean_token_accuracy": 0.8444318771362305, "num_tokens": 84786061.0, "step": 2229 }, { "epoch": 0.2836789212568376, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.4834482669830322, "learning_rate": 9.448919033488768e-06, "loss": 0.521, "mean_token_accuracy": 0.8419445157051086, "num_tokens": 84822550.0, "step": 2230 }, { "epoch": 0.28380613153542805, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.27553653717041, "learning_rate": 9.453158117846547e-06, "loss": 0.5088, "mean_token_accuracy": 0.8447429537773132, "num_tokens": 84860608.0, "step": 2231 }, { "epoch": 0.2839333418140186, "ewc_loss": 0.039794921875, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.470287799835205, "learning_rate": 9.457397202204326e-06, "loss": 0.5368, "mean_token_accuracy": 0.8376107215881348, "num_tokens": 84903065.0, "step": 2232 }, { "epoch": 0.2840605520926091, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.2814035415649414, "learning_rate": 9.461636286562103e-06, "loss": 0.4593, "mean_token_accuracy": 0.8611350059509277, "num_tokens": 84938583.0, "step": 2233 }, { "epoch": 0.2841877623711996, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.47320294380188, "learning_rate": 9.465875370919882e-06, "loss": 0.5441, "mean_token_accuracy": 0.8362080454826355, "num_tokens": 84975939.0, "step": 2234 }, { "epoch": 0.2843149726497901, "ewc_loss": 0.038330078125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 2.2820513248443604, "learning_rate": 9.47011445527766e-06, "loss": 0.5105, "mean_token_accuracy": 0.8427730202674866, "num_tokens": 85011304.0, "step": 2235 }, { "epoch": 0.28444218292838064, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.451416015625, "learning_rate": 9.47435353963544e-06, "loss": 0.4784, "mean_token_accuracy": 0.8523802757263184, "num_tokens": 85053163.0, "step": 2236 }, { "epoch": 0.2845693932069711, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.3022775650024414, "learning_rate": 9.478592623993218e-06, "loss": 0.4769, "mean_token_accuracy": 0.8509673476219177, "num_tokens": 85085919.0, "step": 2237 }, { "epoch": 0.28469660348556164, "ewc_loss": 0.03955078125, "ewc_loss_diag": 5.8650970458984375e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.43190860748291, "learning_rate": 9.482831708350997e-06, "loss": 0.481, "mean_token_accuracy": 0.8537862300872803, "num_tokens": 85124323.0, "step": 2238 }, { "epoch": 0.28482381376415217, "ewc_loss": 0.038330078125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 2.2851052284240723, "learning_rate": 9.487070792708776e-06, "loss": 0.5208, "mean_token_accuracy": 0.8446779251098633, "num_tokens": 85163735.0, "step": 2239 }, { "epoch": 0.28495102404274264, "ewc_loss": 0.03955078125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.414865016937256, "learning_rate": 9.491309877066555e-06, "loss": 0.5101, "mean_token_accuracy": 0.8443934917449951, "num_tokens": 85206533.0, "step": 2240 }, { "epoch": 0.28507823432133317, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.286000967025757, "learning_rate": 9.495548961424334e-06, "loss": 0.5308, "mean_token_accuracy": 0.8374101519584656, "num_tokens": 85248435.0, "step": 2241 }, { "epoch": 0.2852054445999237, "ewc_loss": 0.039306640625, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.417903423309326, "learning_rate": 9.499788045782113e-06, "loss": 0.4941, "mean_token_accuracy": 0.8507439494132996, "num_tokens": 85285686.0, "step": 2242 }, { "epoch": 0.28533265487851417, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.272670269012451, "learning_rate": 9.50402713013989e-06, "loss": 0.5063, "mean_token_accuracy": 0.8461191058158875, "num_tokens": 85327518.0, "step": 2243 }, { "epoch": 0.2854598651571047, "ewc_loss": 0.03955078125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.419232130050659, "learning_rate": 9.508266214497669e-06, "loss": 0.4604, "mean_token_accuracy": 0.8601134419441223, "num_tokens": 85364309.0, "step": 2244 }, { "epoch": 0.2855870754356952, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.2655036449432373, "learning_rate": 9.512505298855448e-06, "loss": 0.4935, "mean_token_accuracy": 0.8535271286964417, "num_tokens": 85407231.0, "step": 2245 }, { "epoch": 0.2857142857142857, "ewc_loss": 0.039794921875, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.429117441177368, "learning_rate": 9.516744383213227e-06, "loss": 0.4366, "mean_token_accuracy": 0.8660986423492432, "num_tokens": 85443753.0, "step": 2246 }, { "epoch": 0.2858414959928762, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.262958288192749, "learning_rate": 9.520983467571005e-06, "loss": 0.5158, "mean_token_accuracy": 0.8425299525260925, "num_tokens": 85482874.0, "step": 2247 }, { "epoch": 0.28596870627146675, "ewc_loss": 0.039794921875, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.456000804901123, "learning_rate": 9.525222551928784e-06, "loss": 0.5538, "mean_token_accuracy": 0.8396703600883484, "num_tokens": 85521648.0, "step": 2248 }, { "epoch": 0.2860959165500572, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.2614645957946777, "learning_rate": 9.529461636286563e-06, "loss": 0.4658, "mean_token_accuracy": 0.8579589128494263, "num_tokens": 85553078.0, "step": 2249 }, { "epoch": 0.28622312682864776, "ewc_loss": 0.039794921875, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.458251476287842, "learning_rate": 9.533700720644342e-06, "loss": 0.4459, "mean_token_accuracy": 0.8630850911140442, "num_tokens": 85589034.0, "step": 2250 }, { "epoch": 0.2863503371072383, "ewc_loss": 0.0380859375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.218650817871094e-05, "grad_norm": 2.2596921920776367, "learning_rate": 9.53793980500212e-06, "loss": 0.5055, "mean_token_accuracy": 0.8492625951766968, "num_tokens": 85631541.0, "step": 2251 }, { "epoch": 0.28647754738582876, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.4738245010375977, "learning_rate": 9.542178889359898e-06, "loss": 0.4776, "mean_token_accuracy": 0.8548774719238281, "num_tokens": 85668498.0, "step": 2252 }, { "epoch": 0.2866047576644193, "ewc_loss": 0.037841796875, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.257693290710449, "learning_rate": 9.546417973717677e-06, "loss": 0.5091, "mean_token_accuracy": 0.8469999432563782, "num_tokens": 85708419.0, "step": 2253 }, { "epoch": 0.2867319679430098, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.472635507583618, "learning_rate": 9.550657058075456e-06, "loss": 0.5194, "mean_token_accuracy": 0.8437877893447876, "num_tokens": 85751564.0, "step": 2254 }, { "epoch": 0.2868591782216003, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.2673513889312744, "learning_rate": 9.554896142433235e-06, "loss": 0.5203, "mean_token_accuracy": 0.8390709161758423, "num_tokens": 85789776.0, "step": 2255 }, { "epoch": 0.2869863885001908, "ewc_loss": 0.0400390625, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.4921154975891113, "learning_rate": 9.559135226791014e-06, "loss": 0.5339, "mean_token_accuracy": 0.8349185585975647, "num_tokens": 85828427.0, "step": 2256 }, { "epoch": 0.28711359877878134, "ewc_loss": 0.038330078125, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.24249267578125e-05, "grad_norm": 2.244365930557251, "learning_rate": 9.563374311148793e-06, "loss": 0.5259, "mean_token_accuracy": 0.8405232429504395, "num_tokens": 85877598.0, "step": 2257 }, { "epoch": 0.2872408090573718, "ewc_loss": 0.04052734375, "ewc_loss_diag": 5.91278076171875e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.5330252647399902, "learning_rate": 9.567613395506572e-06, "loss": 0.4748, "mean_token_accuracy": 0.8572700619697571, "num_tokens": 85908395.0, "step": 2258 }, { "epoch": 0.28736801933596234, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.9604644775390625e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.2481300830841064, "learning_rate": 9.57185247986435e-06, "loss": 0.461, "mean_token_accuracy": 0.8594180941581726, "num_tokens": 85945344.0, "step": 2259 }, { "epoch": 0.28749522961455287, "ewc_loss": 0.041015625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.543466567993164, "learning_rate": 9.57609156422213e-06, "loss": 0.4312, "mean_token_accuracy": 0.8678506016731262, "num_tokens": 85982484.0, "step": 2260 }, { "epoch": 0.28762243989314334, "ewc_loss": 0.03857421875, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 3.266334533691406e-05, "grad_norm": 2.2706522941589355, "learning_rate": 9.580330648579908e-06, "loss": 0.5134, "mean_token_accuracy": 0.8410812616348267, "num_tokens": 86019496.0, "step": 2261 }, { "epoch": 0.2877496501717339, "ewc_loss": 0.04150390625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.557321071624756, "learning_rate": 9.584569732937687e-06, "loss": 0.5088, "mean_token_accuracy": 0.8477023839950562, "num_tokens": 86059344.0, "step": 2262 }, { "epoch": 0.2878768604503244, "ewc_loss": 0.0390625, "ewc_loss_diag": 5.984306335449219e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.2948248386383057, "learning_rate": 9.588808817295466e-06, "loss": 0.4927, "mean_token_accuracy": 0.8469408750534058, "num_tokens": 86093090.0, "step": 2263 }, { "epoch": 0.2880040707289149, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.008148193359375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.5544967651367188, "learning_rate": 9.593047901653245e-06, "loss": 0.4913, "mean_token_accuracy": 0.8500645756721497, "num_tokens": 86127149.0, "step": 2264 }, { "epoch": 0.2881312810075054, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.3002829551696777, "learning_rate": 9.597286986011024e-06, "loss": 0.4498, "mean_token_accuracy": 0.8618689179420471, "num_tokens": 86163541.0, "step": 2265 }, { "epoch": 0.28825849128609593, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.548677682876587, "learning_rate": 9.601526070368801e-06, "loss": 0.5345, "mean_token_accuracy": 0.8395720720291138, "num_tokens": 86200124.0, "step": 2266 }, { "epoch": 0.2883857015646864, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.301905393600464, "learning_rate": 9.60576515472658e-06, "loss": 0.4855, "mean_token_accuracy": 0.8523037433624268, "num_tokens": 86236464.0, "step": 2267 }, { "epoch": 0.28851291184327693, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.5365452766418457, "learning_rate": 9.610004239084359e-06, "loss": 0.4996, "mean_token_accuracy": 0.8477230072021484, "num_tokens": 86277504.0, "step": 2268 }, { "epoch": 0.28864012212186746, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.3174209594726562, "learning_rate": 9.614243323442138e-06, "loss": 0.5616, "mean_token_accuracy": 0.8284275531768799, "num_tokens": 86314404.0, "step": 2269 }, { "epoch": 0.28876733240045793, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.5189952850341797, "learning_rate": 9.618482407799916e-06, "loss": 0.4732, "mean_token_accuracy": 0.8533497452735901, "num_tokens": 86350690.0, "step": 2270 }, { "epoch": 0.28889454267904846, "ewc_loss": 0.039306640625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.2981133460998535, "learning_rate": 9.622721492157695e-06, "loss": 0.4366, "mean_token_accuracy": 0.8674622774124146, "num_tokens": 86389294.0, "step": 2271 }, { "epoch": 0.289021752957639, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.5351319313049316, "learning_rate": 9.626960576515474e-06, "loss": 0.4931, "mean_token_accuracy": 0.8485755920410156, "num_tokens": 86420234.0, "step": 2272 }, { "epoch": 0.28914896323622946, "ewc_loss": 0.039306640625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.298832654953003, "learning_rate": 9.631199660873251e-06, "loss": 0.4695, "mean_token_accuracy": 0.8541333675384521, "num_tokens": 86458435.0, "step": 2273 }, { "epoch": 0.28927617351482, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.5160727500915527, "learning_rate": 9.63543874523103e-06, "loss": 0.477, "mean_token_accuracy": 0.855944037437439, "num_tokens": 86499082.0, "step": 2274 }, { "epoch": 0.2894033837934105, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.3027381896972656, "learning_rate": 9.63967782958881e-06, "loss": 0.5071, "mean_token_accuracy": 0.841322660446167, "num_tokens": 86536989.0, "step": 2275 }, { "epoch": 0.289530594072001, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.5252766609191895, "learning_rate": 9.643916913946588e-06, "loss": 0.4641, "mean_token_accuracy": 0.8587747812271118, "num_tokens": 86572154.0, "step": 2276 }, { "epoch": 0.2896578043505915, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.2923576831817627, "learning_rate": 9.648155998304367e-06, "loss": 0.4401, "mean_token_accuracy": 0.865435779094696, "num_tokens": 86610908.0, "step": 2277 }, { "epoch": 0.28978501462918205, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.5339670181274414, "learning_rate": 9.652395082662146e-06, "loss": 0.5595, "mean_token_accuracy": 0.8278103470802307, "num_tokens": 86653585.0, "step": 2278 }, { "epoch": 0.2899122249077726, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.2952592372894287, "learning_rate": 9.656634167019925e-06, "loss": 0.4951, "mean_token_accuracy": 0.8481771945953369, "num_tokens": 86690794.0, "step": 2279 }, { "epoch": 0.29003943518636305, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.5314009189605713, "learning_rate": 9.660873251377704e-06, "loss": 0.5149, "mean_token_accuracy": 0.8444822430610657, "num_tokens": 86729742.0, "step": 2280 }, { "epoch": 0.2901666454649536, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.303987503051758, "learning_rate": 9.66511233573548e-06, "loss": 0.5241, "mean_token_accuracy": 0.844983696937561, "num_tokens": 86765837.0, "step": 2281 }, { "epoch": 0.2902938557435441, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.031990051269531e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.522528648376465, "learning_rate": 9.66935142009326e-06, "loss": 0.4633, "mean_token_accuracy": 0.8587510585784912, "num_tokens": 86799697.0, "step": 2282 }, { "epoch": 0.2904210660221346, "ewc_loss": 0.039306640625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.3189163208007812, "learning_rate": 9.673590504451039e-06, "loss": 0.4846, "mean_token_accuracy": 0.8505973815917969, "num_tokens": 86832264.0, "step": 2283 }, { "epoch": 0.2905482763007251, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.5195817947387695, "learning_rate": 9.677829588808817e-06, "loss": 0.4715, "mean_token_accuracy": 0.8508239984512329, "num_tokens": 86864492.0, "step": 2284 }, { "epoch": 0.29067548657931563, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.318026304244995, "learning_rate": 9.682068673166596e-06, "loss": 0.5421, "mean_token_accuracy": 0.8392031192779541, "num_tokens": 86898186.0, "step": 2285 }, { "epoch": 0.2908026968579061, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.5164997577667236, "learning_rate": 9.686307757524375e-06, "loss": 0.4953, "mean_token_accuracy": 0.8488215208053589, "num_tokens": 86935850.0, "step": 2286 }, { "epoch": 0.29092990713649663, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.2937753200531006, "learning_rate": 9.690546841882154e-06, "loss": 0.4202, "mean_token_accuracy": 0.8724335432052612, "num_tokens": 86972732.0, "step": 2287 }, { "epoch": 0.29105711741508716, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.5138816833496094, "learning_rate": 9.694785926239933e-06, "loss": 0.5005, "mean_token_accuracy": 0.8483247756958008, "num_tokens": 87012538.0, "step": 2288 }, { "epoch": 0.29118432769367764, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.4036242961883545, "learning_rate": 9.699025010597712e-06, "loss": 0.5419, "mean_token_accuracy": 0.8358193635940552, "num_tokens": 87046917.0, "step": 2289 }, { "epoch": 0.29131153797226816, "ewc_loss": 0.0400390625, "ewc_loss_diag": 6.079673767089844e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.5572128295898438, "learning_rate": 9.70326409495549e-06, "loss": 0.5057, "mean_token_accuracy": 0.844873309135437, "num_tokens": 87086002.0, "step": 2290 }, { "epoch": 0.2914387482508587, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.3078558444976807, "learning_rate": 9.70750317931327e-06, "loss": 0.4864, "mean_token_accuracy": 0.8533928394317627, "num_tokens": 87126138.0, "step": 2291 }, { "epoch": 0.29156595852944917, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.4912354946136475, "learning_rate": 9.711742263671049e-06, "loss": 0.4805, "mean_token_accuracy": 0.8537206053733826, "num_tokens": 87163062.0, "step": 2292 }, { "epoch": 0.2916931688080397, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.287621259689331, "learning_rate": 9.715981348028827e-06, "loss": 0.47, "mean_token_accuracy": 0.8563292026519775, "num_tokens": 87203836.0, "step": 2293 }, { "epoch": 0.2918203790866302, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.5176031589508057, "learning_rate": 9.720220432386606e-06, "loss": 0.4762, "mean_token_accuracy": 0.8569762110710144, "num_tokens": 87236854.0, "step": 2294 }, { "epoch": 0.2919475893652207, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.2845230102539062, "learning_rate": 9.724459516744385e-06, "loss": 0.4608, "mean_token_accuracy": 0.8614703416824341, "num_tokens": 87276316.0, "step": 2295 }, { "epoch": 0.2920747996438112, "ewc_loss": 0.04150390625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.5261070728302, "learning_rate": 9.728698601102162e-06, "loss": 0.4585, "mean_token_accuracy": 0.8617305159568787, "num_tokens": 87314504.0, "step": 2296 }, { "epoch": 0.29220200992240175, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.3111732006073, "learning_rate": 9.732937685459941e-06, "loss": 0.4859, "mean_token_accuracy": 0.8499029278755188, "num_tokens": 87348403.0, "step": 2297 }, { "epoch": 0.2923292202009922, "ewc_loss": 0.04150390625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.527012586593628, "learning_rate": 9.73717676981772e-06, "loss": 0.5278, "mean_token_accuracy": 0.8434265851974487, "num_tokens": 87393044.0, "step": 2298 }, { "epoch": 0.29245643047958275, "ewc_loss": 0.039306640625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.311924934387207, "learning_rate": 9.741415854175499e-06, "loss": 0.548, "mean_token_accuracy": 0.8370615839958191, "num_tokens": 87430930.0, "step": 2299 }, { "epoch": 0.2925836407581733, "ewc_loss": 0.04150390625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.5379209518432617, "learning_rate": 9.745654938533278e-06, "loss": 0.5129, "mean_token_accuracy": 0.8425272703170776, "num_tokens": 87467941.0, "step": 2300 }, { "epoch": 0.29271085103676375, "ewc_loss": 0.039306640625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.3075761795043945, "learning_rate": 9.749894022891057e-06, "loss": 0.5327, "mean_token_accuracy": 0.8399831056594849, "num_tokens": 87507453.0, "step": 2301 }, { "epoch": 0.2928380613153543, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.5259969234466553, "learning_rate": 9.754133107248836e-06, "loss": 0.4505, "mean_token_accuracy": 0.8625757694244385, "num_tokens": 87545983.0, "step": 2302 }, { "epoch": 0.2929652715939448, "ewc_loss": 0.039306640625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.312716007232666, "learning_rate": 9.758372191606615e-06, "loss": 0.534, "mean_token_accuracy": 0.8357081413269043, "num_tokens": 87582352.0, "step": 2303 }, { "epoch": 0.2930924818725353, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5275676250457764, "learning_rate": 9.762611275964392e-06, "loss": 0.4476, "mean_token_accuracy": 0.8653054237365723, "num_tokens": 87616668.0, "step": 2304 }, { "epoch": 0.2932196921511258, "ewc_loss": 0.03955078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.3189914226531982, "learning_rate": 9.76685036032217e-06, "loss": 0.5146, "mean_token_accuracy": 0.8425691723823547, "num_tokens": 87653008.0, "step": 2305 }, { "epoch": 0.29334690242971634, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.509507417678833, "learning_rate": 9.77108944467995e-06, "loss": 0.4656, "mean_token_accuracy": 0.8577518463134766, "num_tokens": 87694985.0, "step": 2306 }, { "epoch": 0.2934741127083068, "ewc_loss": 0.039306640625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.325829029083252, "learning_rate": 9.775328529037728e-06, "loss": 0.5071, "mean_token_accuracy": 0.8436159491539001, "num_tokens": 87730841.0, "step": 2307 }, { "epoch": 0.29360132298689734, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.5343079566955566, "learning_rate": 9.779567613395507e-06, "loss": 0.5417, "mean_token_accuracy": 0.8407772779464722, "num_tokens": 87762215.0, "step": 2308 }, { "epoch": 0.29372853326548787, "ewc_loss": 0.039306640625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.3113579750061035, "learning_rate": 9.783806697753286e-06, "loss": 0.482, "mean_token_accuracy": 0.8534024357795715, "num_tokens": 87799150.0, "step": 2309 }, { "epoch": 0.29385574354407834, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.501542806625366, "learning_rate": 9.788045782111065e-06, "loss": 0.45, "mean_token_accuracy": 0.8614401817321777, "num_tokens": 87842463.0, "step": 2310 }, { "epoch": 0.29398295382266887, "ewc_loss": 0.039306640625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.305215358734131, "learning_rate": 9.792284866468842e-06, "loss": 0.4413, "mean_token_accuracy": 0.8632267713546753, "num_tokens": 87880093.0, "step": 2311 }, { "epoch": 0.2941101641012594, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.5309841632843018, "learning_rate": 9.796523950826621e-06, "loss": 0.4482, "mean_token_accuracy": 0.861441969871521, "num_tokens": 87912713.0, "step": 2312 }, { "epoch": 0.29423737437984987, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.3260977268218994, "learning_rate": 9.8007630351844e-06, "loss": 0.5196, "mean_token_accuracy": 0.8442461490631104, "num_tokens": 87946357.0, "step": 2313 }, { "epoch": 0.2943645846584404, "ewc_loss": 0.04150390625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.5383098125457764, "learning_rate": 9.805002119542179e-06, "loss": 0.5524, "mean_token_accuracy": 0.8321632146835327, "num_tokens": 87981102.0, "step": 2314 }, { "epoch": 0.2944917949370309, "ewc_loss": 0.039306640625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.2989184856414795, "learning_rate": 9.809241203899958e-06, "loss": 0.4855, "mean_token_accuracy": 0.8545794486999512, "num_tokens": 88022325.0, "step": 2315 }, { "epoch": 0.2946190052156214, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5519497394561768, "learning_rate": 9.813480288257737e-06, "loss": 0.5096, "mean_token_accuracy": 0.84087735414505, "num_tokens": 88057254.0, "step": 2316 }, { "epoch": 0.2947462154942119, "ewc_loss": 0.039306640625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.2962241172790527, "learning_rate": 9.817719372615516e-06, "loss": 0.4629, "mean_token_accuracy": 0.8592562675476074, "num_tokens": 88095667.0, "step": 2317 }, { "epoch": 0.29487342577280246, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5755887031555176, "learning_rate": 9.821958456973294e-06, "loss": 0.4348, "mean_token_accuracy": 0.8675005435943604, "num_tokens": 88129634.0, "step": 2318 }, { "epoch": 0.29500063605139293, "ewc_loss": 0.039306640625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.314018249511719e-05, "grad_norm": 2.3078525066375732, "learning_rate": 9.826197541331073e-06, "loss": 0.448, "mean_token_accuracy": 0.8620961308479309, "num_tokens": 88164592.0, "step": 2319 }, { "epoch": 0.29512784632998346, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.585843563079834, "learning_rate": 9.830436625688852e-06, "loss": 0.5068, "mean_token_accuracy": 0.8471459150314331, "num_tokens": 88202390.0, "step": 2320 }, { "epoch": 0.295255056608574, "ewc_loss": 0.03955078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.3072848320007324, "learning_rate": 9.834675710046631e-06, "loss": 0.4601, "mean_token_accuracy": 0.8593083620071411, "num_tokens": 88243104.0, "step": 2321 }, { "epoch": 0.29538226688716446, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 3.544149398803711, "learning_rate": 9.83891479440441e-06, "loss": 0.4514, "mean_token_accuracy": 0.8652856349945068, "num_tokens": 88284038.0, "step": 2322 }, { "epoch": 0.295509477165755, "ewc_loss": 0.048095703125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6416759490966797, "learning_rate": 9.843153878762189e-06, "loss": 0.5329, "mean_token_accuracy": 0.8442984223365784, "num_tokens": 88320705.0, "step": 2323 }, { "epoch": 0.2956366874443455, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.6472153663635254, "learning_rate": 9.847392963119968e-06, "loss": 0.4745, "mean_token_accuracy": 0.8565487861633301, "num_tokens": 88359636.0, "step": 2324 }, { "epoch": 0.295763897722936, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.460365056991577, "learning_rate": 9.851632047477747e-06, "loss": 0.4962, "mean_token_accuracy": 0.847180962562561, "num_tokens": 88402029.0, "step": 2325 }, { "epoch": 0.2958911080015265, "ewc_loss": 0.044677734375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6566078662872314, "learning_rate": 9.855871131835524e-06, "loss": 0.4209, "mean_token_accuracy": 0.8718474507331848, "num_tokens": 88438245.0, "step": 2326 }, { "epoch": 0.29601831828011704, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4458816051483154, "learning_rate": 9.860110216193303e-06, "loss": 0.4561, "mean_token_accuracy": 0.8618000745773315, "num_tokens": 88477987.0, "step": 2327 }, { "epoch": 0.2961455285587075, "ewc_loss": 0.042724609375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6058108806610107, "learning_rate": 9.864349300551082e-06, "loss": 0.4482, "mean_token_accuracy": 0.8638469576835632, "num_tokens": 88510759.0, "step": 2328 }, { "epoch": 0.29627273883729804, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4403090476989746, "learning_rate": 9.86858838490886e-06, "loss": 0.4615, "mean_token_accuracy": 0.8580456376075745, "num_tokens": 88544819.0, "step": 2329 }, { "epoch": 0.2963999491158886, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5563461780548096, "learning_rate": 9.87282746926664e-06, "loss": 0.4659, "mean_token_accuracy": 0.8576679229736328, "num_tokens": 88584560.0, "step": 2330 }, { "epoch": 0.2965271593944791, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.398085594177246, "learning_rate": 9.877066553624418e-06, "loss": 0.4606, "mean_token_accuracy": 0.8586320281028748, "num_tokens": 88620733.0, "step": 2331 }, { "epoch": 0.2966543696730696, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5328848361968994, "learning_rate": 9.881305637982197e-06, "loss": 0.4786, "mean_token_accuracy": 0.8541520833969116, "num_tokens": 88661306.0, "step": 2332 }, { "epoch": 0.2967815799516601, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3761990070343018, "learning_rate": 9.885544722339976e-06, "loss": 0.4576, "mean_token_accuracy": 0.8593248128890991, "num_tokens": 88696886.0, "step": 2333 }, { "epoch": 0.29690879023025063, "ewc_loss": 0.04150390625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.524651288986206, "learning_rate": 9.889783806697753e-06, "loss": 0.4361, "mean_token_accuracy": 0.8657915592193604, "num_tokens": 88733242.0, "step": 2334 }, { "epoch": 0.2970360005088411, "ewc_loss": 0.039794921875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3528506755828857, "learning_rate": 9.894022891055532e-06, "loss": 0.4373, "mean_token_accuracy": 0.8675006628036499, "num_tokens": 88769599.0, "step": 2335 }, { "epoch": 0.29716321078743163, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5338470935821533, "learning_rate": 9.898261975413311e-06, "loss": 0.4219, "mean_token_accuracy": 0.8722171783447266, "num_tokens": 88802885.0, "step": 2336 }, { "epoch": 0.29729042106602216, "ewc_loss": 0.039794921875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3439276218414307, "learning_rate": 9.90250105977109e-06, "loss": 0.4769, "mean_token_accuracy": 0.8548654317855835, "num_tokens": 88837061.0, "step": 2337 }, { "epoch": 0.29741763134461263, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5427327156066895, "learning_rate": 9.906740144128869e-06, "loss": 0.4883, "mean_token_accuracy": 0.8481442928314209, "num_tokens": 88874672.0, "step": 2338 }, { "epoch": 0.29754484162320316, "ewc_loss": 0.03955078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.3320231437683105, "learning_rate": 9.910979228486648e-06, "loss": 0.4366, "mean_token_accuracy": 0.8651371598243713, "num_tokens": 88910520.0, "step": 2339 }, { "epoch": 0.2976720519017937, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.560701370239258, "learning_rate": 9.915218312844427e-06, "loss": 0.5113, "mean_token_accuracy": 0.8480079174041748, "num_tokens": 88949495.0, "step": 2340 }, { "epoch": 0.29779926218038416, "ewc_loss": 0.0390625, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.2901763916015625e-05, "grad_norm": 2.3133983612060547, "learning_rate": 9.919457397202205e-06, "loss": 0.5048, "mean_token_accuracy": 0.8451906442642212, "num_tokens": 88988776.0, "step": 2341 }, { "epoch": 0.2979264724589747, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.596726655960083, "learning_rate": 9.923696481559984e-06, "loss": 0.458, "mean_token_accuracy": 0.8624445796012878, "num_tokens": 89018983.0, "step": 2342 }, { "epoch": 0.2980536827375652, "ewc_loss": 0.03955078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.322477102279663, "learning_rate": 9.927935565917761e-06, "loss": 0.5087, "mean_token_accuracy": 0.8447844982147217, "num_tokens": 89055834.0, "step": 2343 }, { "epoch": 0.2981808930161557, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5951552391052246, "learning_rate": 9.93217465027554e-06, "loss": 0.5001, "mean_token_accuracy": 0.8526912927627563, "num_tokens": 89098612.0, "step": 2344 }, { "epoch": 0.2983081032947462, "ewc_loss": 0.039794921875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3463985919952393, "learning_rate": 9.93641373463332e-06, "loss": 0.5588, "mean_token_accuracy": 0.8352242708206177, "num_tokens": 89130169.0, "step": 2345 }, { "epoch": 0.29843531357333675, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5849788188934326, "learning_rate": 9.940652818991098e-06, "loss": 0.449, "mean_token_accuracy": 0.8660982847213745, "num_tokens": 89170346.0, "step": 2346 }, { "epoch": 0.2985625238519272, "ewc_loss": 0.03955078125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.3283824920654297, "learning_rate": 9.944891903348877e-06, "loss": 0.5123, "mean_token_accuracy": 0.8430777192115784, "num_tokens": 89211713.0, "step": 2347 }, { "epoch": 0.29868973413051775, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5910286903381348, "learning_rate": 9.949130987706656e-06, "loss": 0.465, "mean_token_accuracy": 0.8601582646369934, "num_tokens": 89252842.0, "step": 2348 }, { "epoch": 0.2988169444091083, "ewc_loss": 0.039794921875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3268678188323975, "learning_rate": 9.953370072064435e-06, "loss": 0.4448, "mean_token_accuracy": 0.8629378080368042, "num_tokens": 89291914.0, "step": 2349 }, { "epoch": 0.29894415468769875, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.103515625e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.6065568923950195, "learning_rate": 9.957609156422214e-06, "loss": 0.5328, "mean_token_accuracy": 0.8395408391952515, "num_tokens": 89333912.0, "step": 2350 }, { "epoch": 0.2990713649662893, "ewc_loss": 0.0400390625, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.360836982727051, "learning_rate": 9.961848240779993e-06, "loss": 0.4734, "mean_token_accuracy": 0.8566151857376099, "num_tokens": 89363603.0, "step": 2351 }, { "epoch": 0.2991985752448798, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5888705253601074, "learning_rate": 9.966087325137771e-06, "loss": 0.4541, "mean_token_accuracy": 0.86363285779953, "num_tokens": 89404392.0, "step": 2352 }, { "epoch": 0.2993257855234703, "ewc_loss": 0.039794921875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.342686176300049, "learning_rate": 9.97032640949555e-06, "loss": 0.5182, "mean_token_accuracy": 0.8384481072425842, "num_tokens": 89444180.0, "step": 2353 }, { "epoch": 0.2994529958020608, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5953102111816406, "learning_rate": 9.97456549385333e-06, "loss": 0.5268, "mean_token_accuracy": 0.8375846147537231, "num_tokens": 89483875.0, "step": 2354 }, { "epoch": 0.29958020608065133, "ewc_loss": 0.039794921875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3593242168426514, "learning_rate": 9.978804578211108e-06, "loss": 0.5154, "mean_token_accuracy": 0.8442119359970093, "num_tokens": 89519544.0, "step": 2355 }, { "epoch": 0.2997074163592418, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.198883056640625e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.606231689453125, "learning_rate": 9.983043662568887e-06, "loss": 0.5065, "mean_token_accuracy": 0.8490967154502869, "num_tokens": 89552635.0, "step": 2356 }, { "epoch": 0.29983462663783234, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3516685962677, "learning_rate": 9.987282746926664e-06, "loss": 0.4677, "mean_token_accuracy": 0.8601732850074768, "num_tokens": 89588514.0, "step": 2357 }, { "epoch": 0.29996183691642286, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.588196039199829, "learning_rate": 9.991521831284443e-06, "loss": 0.4767, "mean_token_accuracy": 0.8566280007362366, "num_tokens": 89629029.0, "step": 2358 }, { "epoch": 0.30008904719501334, "ewc_loss": 0.039794921875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.35524845123291, "learning_rate": 9.995760915642222e-06, "loss": 0.4187, "mean_token_accuracy": 0.8711795806884766, "num_tokens": 89659214.0, "step": 2359 }, { "epoch": 0.30021625747360386, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.585376739501953, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8612909317016602, "num_tokens": 89695342.0, "step": 2360 }, { "epoch": 0.3003434677521944, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3599417209625244, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8519028425216675, "num_tokens": 89731209.0, "step": 2361 }, { "epoch": 0.30047067803078487, "ewc_loss": 0.042724609375, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5748796463012695, "learning_rate": 1e-05, "loss": 0.5344, "mean_token_accuracy": 0.8361498713493347, "num_tokens": 89771516.0, "step": 2362 }, { "epoch": 0.3005978883093754, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.353044271469116, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8406541347503662, "num_tokens": 89817929.0, "step": 2363 }, { "epoch": 0.3007250985879659, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.6092636585235596, "learning_rate": 1e-05, "loss": 0.601, "mean_token_accuracy": 0.8252404928207397, "num_tokens": 89850714.0, "step": 2364 }, { "epoch": 0.3008523088665564, "ewc_loss": 0.039794921875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.3535139560699463, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8500891923904419, "num_tokens": 89885686.0, "step": 2365 }, { "epoch": 0.3009795191451469, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5780794620513916, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8592206835746765, "num_tokens": 89926549.0, "step": 2366 }, { "epoch": 0.30110672942373745, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.246566772460938e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.350149393081665, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.839293897151947, "num_tokens": 89969031.0, "step": 2367 }, { "epoch": 0.3012339397023279, "ewc_loss": 0.042724609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.602163076400757, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8464337587356567, "num_tokens": 90000348.0, "step": 2368 }, { "epoch": 0.30136114998091845, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.364300489425659, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8511936068534851, "num_tokens": 90033945.0, "step": 2369 }, { "epoch": 0.301488360259509, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5952539443969727, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.860974133014679, "num_tokens": 90069537.0, "step": 2370 }, { "epoch": 0.30161557053809945, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3559749126434326, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8626810908317566, "num_tokens": 90107923.0, "step": 2371 }, { "epoch": 0.30174278081669, "ewc_loss": 0.042724609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.583549976348877, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8410559296607971, "num_tokens": 90152211.0, "step": 2372 }, { "epoch": 0.3018699910952805, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3600027561187744, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8479876518249512, "num_tokens": 90193687.0, "step": 2373 }, { "epoch": 0.301997201373871, "ewc_loss": 0.042724609375, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5727314949035645, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.868363618850708, "num_tokens": 90232639.0, "step": 2374 }, { "epoch": 0.3021244116524615, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3562533855438232, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8459551930427551, "num_tokens": 90273344.0, "step": 2375 }, { "epoch": 0.30225162193105204, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.573162794113159, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8510386943817139, "num_tokens": 90309999.0, "step": 2376 }, { "epoch": 0.3023788322096425, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.29425048828125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3473942279815674, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8554219007492065, "num_tokens": 90349049.0, "step": 2377 }, { "epoch": 0.30250604248823304, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5737056732177734, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.845832884311676, "num_tokens": 90388795.0, "step": 2378 }, { "epoch": 0.30263325276682357, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.366847276687622, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8347100615501404, "num_tokens": 90423785.0, "step": 2379 }, { "epoch": 0.3027604630454141, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5687506198883057, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.850482165813446, "num_tokens": 90461625.0, "step": 2380 }, { "epoch": 0.30288767332400457, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3418426513671875, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8590874075889587, "num_tokens": 90500286.0, "step": 2381 }, { "epoch": 0.3030148836025951, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.578787326812744, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8545056581497192, "num_tokens": 90533463.0, "step": 2382 }, { "epoch": 0.3031420938811856, "ewc_loss": 0.0400390625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3422834873199463, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8587543964385986, "num_tokens": 90570536.0, "step": 2383 }, { "epoch": 0.3032693041597761, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5996251106262207, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8376882672309875, "num_tokens": 90603030.0, "step": 2384 }, { "epoch": 0.3033965144383666, "ewc_loss": 0.039794921875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.3466475009918213, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.842557966709137, "num_tokens": 90638459.0, "step": 2385 }, { "epoch": 0.30352372471695716, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.60518217086792, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8519953489303589, "num_tokens": 90671894.0, "step": 2386 }, { "epoch": 0.30365093499554763, "ewc_loss": 0.0400390625, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3390016555786133, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8593123555183411, "num_tokens": 90708583.0, "step": 2387 }, { "epoch": 0.30377814527413816, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5979361534118652, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8427632451057434, "num_tokens": 90749914.0, "step": 2388 }, { "epoch": 0.3039053555527287, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3339943885803223, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8499721884727478, "num_tokens": 90791955.0, "step": 2389 }, { "epoch": 0.30403256583131916, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.341934204101562e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6122450828552246, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8693144917488098, "num_tokens": 90827666.0, "step": 2390 }, { "epoch": 0.3041597761099097, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.358940362930298, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8464900255203247, "num_tokens": 90860647.0, "step": 2391 }, { "epoch": 0.3042869863885002, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6305580139160156, "learning_rate": 1e-05, "loss": 0.5301, "mean_token_accuracy": 0.837494969367981, "num_tokens": 90899737.0, "step": 2392 }, { "epoch": 0.3044141966670907, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.342752695083618, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8588851690292358, "num_tokens": 90939934.0, "step": 2393 }, { "epoch": 0.3045414069456812, "ewc_loss": 0.04345703125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6441962718963623, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8435364365577698, "num_tokens": 90971538.0, "step": 2394 }, { "epoch": 0.30466861722427174, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.345156192779541, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8544541001319885, "num_tokens": 91012119.0, "step": 2395 }, { "epoch": 0.3047958275028622, "ewc_loss": 0.043701171875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.627861499786377, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8493776321411133, "num_tokens": 91051342.0, "step": 2396 }, { "epoch": 0.30492303778145274, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3580667972564697, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8568102121353149, "num_tokens": 91087650.0, "step": 2397 }, { "epoch": 0.30505024806004327, "ewc_loss": 0.04345703125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.619727849960327, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8536503911018372, "num_tokens": 91129786.0, "step": 2398 }, { "epoch": 0.30517745833863374, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3742356300354004, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.850534975528717, "num_tokens": 91165273.0, "step": 2399 }, { "epoch": 0.3053046686172243, "ewc_loss": 0.043212890625, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6152803897857666, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8665440082550049, "num_tokens": 91198034.0, "step": 2400 }, { "epoch": 0.3054318788958148, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3728299140930176, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8554863929748535, "num_tokens": 91241415.0, "step": 2401 }, { "epoch": 0.3055590891744053, "ewc_loss": 0.042724609375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5961461067199707, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8463007211685181, "num_tokens": 91284713.0, "step": 2402 }, { "epoch": 0.3056862994529958, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3742733001708984, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8471937775611877, "num_tokens": 91322143.0, "step": 2403 }, { "epoch": 0.30581350973158633, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.581916332244873, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8463606834411621, "num_tokens": 91366825.0, "step": 2404 }, { "epoch": 0.3059407200101768, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.376466751098633, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.8450700044631958, "num_tokens": 91404469.0, "step": 2405 }, { "epoch": 0.30606793028876733, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5711607933044434, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8419308662414551, "num_tokens": 91444799.0, "step": 2406 }, { "epoch": 0.30619514056735786, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3655495643615723, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8491349220275879, "num_tokens": 91483843.0, "step": 2407 }, { "epoch": 0.30632235084594833, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.437301635742188e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.562519073486328, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8507669568061829, "num_tokens": 91523922.0, "step": 2408 }, { "epoch": 0.30644956112453886, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.374760150909424, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8545135259628296, "num_tokens": 91561155.0, "step": 2409 }, { "epoch": 0.3065767714031294, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5689125061035156, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8518132567405701, "num_tokens": 91596072.0, "step": 2410 }, { "epoch": 0.30670398168171986, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.346578598022461, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8495166897773743, "num_tokens": 91640369.0, "step": 2411 }, { "epoch": 0.3068311919603104, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5646305084228516, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.852366030216217, "num_tokens": 91674941.0, "step": 2412 }, { "epoch": 0.3069584022389009, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.349416732788086, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8528255224227905, "num_tokens": 91709023.0, "step": 2413 }, { "epoch": 0.3070856125174914, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5617876052856445, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8535504341125488, "num_tokens": 91750684.0, "step": 2414 }, { "epoch": 0.3072128227960819, "ewc_loss": 0.0400390625, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.343435049057007, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8488244414329529, "num_tokens": 91789369.0, "step": 2415 }, { "epoch": 0.30734003307467245, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5728888511657715, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8402898907661438, "num_tokens": 91828228.0, "step": 2416 }, { "epoch": 0.3074672433532629, "ewc_loss": 0.0400390625, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3359315395355225, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8606293201446533, "num_tokens": 91866191.0, "step": 2417 }, { "epoch": 0.30759445363185345, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5751004219055176, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8531850576400757, "num_tokens": 91907945.0, "step": 2418 }, { "epoch": 0.307721663910444, "ewc_loss": 0.0400390625, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.350602626800537, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8584593534469604, "num_tokens": 91938651.0, "step": 2419 }, { "epoch": 0.30784887418903445, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5793094635009766, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8514851331710815, "num_tokens": 91977557.0, "step": 2420 }, { "epoch": 0.307976084467625, "ewc_loss": 0.0400390625, "ewc_loss_diag": 6.4849853515625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3534963130950928, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8426984548568726, "num_tokens": 92012316.0, "step": 2421 }, { "epoch": 0.3081032947462155, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5854623317718506, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8506628274917603, "num_tokens": 92049262.0, "step": 2422 }, { "epoch": 0.308230505024806, "ewc_loss": 0.0400390625, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.391183614730835, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8465626835823059, "num_tokens": 92086048.0, "step": 2423 }, { "epoch": 0.3083577153033965, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5455586910247803, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8601250648498535, "num_tokens": 92123540.0, "step": 2424 }, { "epoch": 0.30848492558198704, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.379176139831543, "learning_rate": 1e-05, "loss": 0.5499, "mean_token_accuracy": 0.8302397727966309, "num_tokens": 92158275.0, "step": 2425 }, { "epoch": 0.3086121358605775, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5169296264648438, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8691853284835815, "num_tokens": 92196444.0, "step": 2426 }, { "epoch": 0.30873934613916804, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.532669067382812e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3542990684509277, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8584747910499573, "num_tokens": 92233988.0, "step": 2427 }, { "epoch": 0.30886655641775856, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5173566341400146, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8605228662490845, "num_tokens": 92271102.0, "step": 2428 }, { "epoch": 0.3089937666963491, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3664114475250244, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8505603671073914, "num_tokens": 92306112.0, "step": 2429 }, { "epoch": 0.30912097697493957, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.514719009399414, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8580412864685059, "num_tokens": 92347796.0, "step": 2430 }, { "epoch": 0.3092481872535301, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.370814085006714, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8450140953063965, "num_tokens": 92382018.0, "step": 2431 }, { "epoch": 0.3093753975321206, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4903476238250732, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8482295274734497, "num_tokens": 92428352.0, "step": 2432 }, { "epoch": 0.3095026078107111, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.368082284927368, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8470339775085449, "num_tokens": 92462826.0, "step": 2433 }, { "epoch": 0.3096298180893016, "ewc_loss": 0.04150390625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4871368408203125, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8574818968772888, "num_tokens": 92498371.0, "step": 2434 }, { "epoch": 0.30975702836789215, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.36179780960083, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8511373400688171, "num_tokens": 92536921.0, "step": 2435 }, { "epoch": 0.3098842386464826, "ewc_loss": 0.04150390625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.513307571411133, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8412978649139404, "num_tokens": 92570627.0, "step": 2436 }, { "epoch": 0.31001144892507315, "ewc_loss": 0.0400390625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.346740484237671, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8570107221603394, "num_tokens": 92608683.0, "step": 2437 }, { "epoch": 0.3101386592036637, "ewc_loss": 0.04150390625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.5104806423187256, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8596034049987793, "num_tokens": 92644689.0, "step": 2438 }, { "epoch": 0.31026586948225415, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.337066888809204, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8514076471328735, "num_tokens": 92684471.0, "step": 2439 }, { "epoch": 0.3103930797608447, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5388894081115723, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.8422123789787292, "num_tokens": 92724991.0, "step": 2440 }, { "epoch": 0.3105202900394352, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3356378078460693, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8445353507995605, "num_tokens": 92759302.0, "step": 2441 }, { "epoch": 0.3106475003180257, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.556548595428467, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8463112115859985, "num_tokens": 92794585.0, "step": 2442 }, { "epoch": 0.3107747105966162, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.330557346343994, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8453246355056763, "num_tokens": 92832405.0, "step": 2443 }, { "epoch": 0.31090192087520674, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.565434217453003, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8431133031845093, "num_tokens": 92877960.0, "step": 2444 }, { "epoch": 0.3110291311537972, "ewc_loss": 0.0400390625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.3185720443725586, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8629127740859985, "num_tokens": 92914421.0, "step": 2445 }, { "epoch": 0.31115634143238774, "ewc_loss": 0.042724609375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.58974027633667, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8489154577255249, "num_tokens": 92954844.0, "step": 2446 }, { "epoch": 0.31128355171097827, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.338082790374756, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8646132946014404, "num_tokens": 92987719.0, "step": 2447 }, { "epoch": 0.31141076198956874, "ewc_loss": 0.043212890625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5933327674865723, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8500248193740845, "num_tokens": 93027335.0, "step": 2448 }, { "epoch": 0.31153797226815927, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3251454830169678, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8520904779434204, "num_tokens": 93063325.0, "step": 2449 }, { "epoch": 0.3116651825467498, "ewc_loss": 0.043701171875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.605055570602417, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8418566584587097, "num_tokens": 93104057.0, "step": 2450 }, { "epoch": 0.31179239282534027, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3410134315490723, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.845099687576294, "num_tokens": 93138912.0, "step": 2451 }, { "epoch": 0.3119196031039308, "ewc_loss": 0.04345703125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6127593517303467, "learning_rate": 1e-05, "loss": 0.5485, "mean_token_accuracy": 0.8396745920181274, "num_tokens": 93178658.0, "step": 2452 }, { "epoch": 0.3120468133825213, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.335386037826538, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.85316002368927, "num_tokens": 93217723.0, "step": 2453 }, { "epoch": 0.3121740236611118, "ewc_loss": 0.0439453125, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.626607656478882, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8591405749320984, "num_tokens": 93253917.0, "step": 2454 }, { "epoch": 0.3123012339397023, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3366546630859375, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8590635061264038, "num_tokens": 93289735.0, "step": 2455 }, { "epoch": 0.31242844421829286, "ewc_loss": 0.04443359375, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6312167644500732, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8506080508232117, "num_tokens": 93331549.0, "step": 2456 }, { "epoch": 0.31255565449688333, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.353440046310425, "learning_rate": 1e-05, "loss": 0.5342, "mean_token_accuracy": 0.8377199172973633, "num_tokens": 93372977.0, "step": 2457 }, { "epoch": 0.31268286477547386, "ewc_loss": 0.043701171875, "ewc_loss_diag": 6.580352783203125e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.645233392715454, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8631533980369568, "num_tokens": 93406842.0, "step": 2458 }, { "epoch": 0.3128100750540644, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3532514572143555, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8530778884887695, "num_tokens": 93444113.0, "step": 2459 }, { "epoch": 0.31293728533265486, "ewc_loss": 0.043701171875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.630443572998047, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8400593996047974, "num_tokens": 93490618.0, "step": 2460 }, { "epoch": 0.3130644956112454, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3696389198303223, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8440078496932983, "num_tokens": 93524711.0, "step": 2461 }, { "epoch": 0.3131917058898359, "ewc_loss": 0.04345703125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.631112575531006, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8514559268951416, "num_tokens": 93559169.0, "step": 2462 }, { "epoch": 0.3133189161684264, "ewc_loss": 0.04150390625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.362398862838745, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8533161878585815, "num_tokens": 93598941.0, "step": 2463 }, { "epoch": 0.3134461264470169, "ewc_loss": 0.043701171875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6195058822631836, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8559097051620483, "num_tokens": 93634979.0, "step": 2464 }, { "epoch": 0.31357333672560744, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.367269992828369, "learning_rate": 1e-05, "loss": 0.5451, "mean_token_accuracy": 0.8298642635345459, "num_tokens": 93676736.0, "step": 2465 }, { "epoch": 0.3137005470041979, "ewc_loss": 0.043212890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6035585403442383, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8621112108230591, "num_tokens": 93720118.0, "step": 2466 }, { "epoch": 0.31382775728278844, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.363776445388794, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8580553531646729, "num_tokens": 93760814.0, "step": 2467 }, { "epoch": 0.313954967561379, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.6050002574920654, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8692046403884888, "num_tokens": 93795326.0, "step": 2468 }, { "epoch": 0.31408217783996945, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.372545003890991, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8525545597076416, "num_tokens": 93832993.0, "step": 2469 }, { "epoch": 0.31420938811856, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5837552547454834, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.864483118057251, "num_tokens": 93871282.0, "step": 2470 }, { "epoch": 0.3143365983971505, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.372616767883301, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.8476879596710205, "num_tokens": 93912698.0, "step": 2471 }, { "epoch": 0.314463808675741, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.6077115535736084, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8551893830299377, "num_tokens": 93952499.0, "step": 2472 }, { "epoch": 0.3145910189543315, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.361483097076416, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.8388307690620422, "num_tokens": 93992277.0, "step": 2473 }, { "epoch": 0.31471822923292203, "ewc_loss": 0.042724609375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5967843532562256, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8542578220367432, "num_tokens": 94032012.0, "step": 2474 }, { "epoch": 0.3148454395115125, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3502933979034424, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8593800663948059, "num_tokens": 94072210.0, "step": 2475 }, { "epoch": 0.31497264979010303, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5980939865112305, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8480715155601501, "num_tokens": 94112269.0, "step": 2476 }, { "epoch": 0.31509986006869356, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3692827224731445, "learning_rate": 1e-05, "loss": 0.5729, "mean_token_accuracy": 0.8269909620285034, "num_tokens": 94153596.0, "step": 2477 }, { "epoch": 0.31522707034728403, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5849876403808594, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8463221192359924, "num_tokens": 94195146.0, "step": 2478 }, { "epoch": 0.31535428062587456, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3610727787017822, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8602696657180786, "num_tokens": 94233196.0, "step": 2479 }, { "epoch": 0.3154814909044651, "ewc_loss": 0.043212890625, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5710482597351074, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8489061594009399, "num_tokens": 94275314.0, "step": 2480 }, { "epoch": 0.3156087011830556, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.356240749359131, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8547793030738831, "num_tokens": 94316673.0, "step": 2481 }, { "epoch": 0.3157359114616461, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5714352130889893, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8564642667770386, "num_tokens": 94354118.0, "step": 2482 }, { "epoch": 0.3158631217402366, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.373560667037964, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8462814688682556, "num_tokens": 94386476.0, "step": 2483 }, { "epoch": 0.31599033201882715, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.628036499023438e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.569988489151001, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8513826727867126, "num_tokens": 94420146.0, "step": 2484 }, { "epoch": 0.3161175422974176, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3746094703674316, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8487464785575867, "num_tokens": 94455057.0, "step": 2485 }, { "epoch": 0.31624475257600815, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5355563163757324, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8591829538345337, "num_tokens": 94500666.0, "step": 2486 }, { "epoch": 0.3163719628545987, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.352428913116455, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8656332492828369, "num_tokens": 94537623.0, "step": 2487 }, { "epoch": 0.31649917313318915, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5535244941711426, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8642795085906982, "num_tokens": 94568563.0, "step": 2488 }, { "epoch": 0.3166263834117797, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3536734580993652, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8660885691642761, "num_tokens": 94600931.0, "step": 2489 }, { "epoch": 0.3167535936903702, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.67572021484375e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.568429470062256, "learning_rate": 1e-05, "loss": 0.5514, "mean_token_accuracy": 0.8343077898025513, "num_tokens": 94634345.0, "step": 2490 }, { "epoch": 0.3168808039689607, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.333794355392456, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8644734025001526, "num_tokens": 94673157.0, "step": 2491 }, { "epoch": 0.3170080142475512, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.581195831298828, "learning_rate": 1e-05, "loss": 0.5291, "mean_token_accuracy": 0.8368008136749268, "num_tokens": 94712727.0, "step": 2492 }, { "epoch": 0.31713522452614173, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.34074330329895, "learning_rate": 1e-05, "loss": 0.5563, "mean_token_accuracy": 0.8338608145713806, "num_tokens": 94752341.0, "step": 2493 }, { "epoch": 0.3172624348047322, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5892889499664307, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.8438907861709595, "num_tokens": 94791287.0, "step": 2494 }, { "epoch": 0.31738964508332274, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.340559720993042, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.846501350402832, "num_tokens": 94833029.0, "step": 2495 }, { "epoch": 0.31751685536191326, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5882375240325928, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8507488965988159, "num_tokens": 94873400.0, "step": 2496 }, { "epoch": 0.31764406564050374, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.4290859699249268, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8332195281982422, "num_tokens": 94914603.0, "step": 2497 }, { "epoch": 0.31777127591909426, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.525761842727661, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8722125887870789, "num_tokens": 94955925.0, "step": 2498 }, { "epoch": 0.3178984861976848, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3701796531677246, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8696577548980713, "num_tokens": 94989420.0, "step": 2499 }, { "epoch": 0.31802569647627527, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.5130927562713623, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8455950021743774, "num_tokens": 95028835.0, "step": 2500 }, { "epoch": 0.3181529067548658, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3740267753601074, "learning_rate": 1e-05, "loss": 0.5579, "mean_token_accuracy": 0.8394989371299744, "num_tokens": 95069246.0, "step": 2501 }, { "epoch": 0.3182801170334563, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.5330331325531006, "learning_rate": 1e-05, "loss": 0.5992, "mean_token_accuracy": 0.8190516233444214, "num_tokens": 95110184.0, "step": 2502 }, { "epoch": 0.3184073273120468, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3552184104919434, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.856016993522644, "num_tokens": 95149118.0, "step": 2503 }, { "epoch": 0.3185345375906373, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5093865394592285, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8526110649108887, "num_tokens": 95195605.0, "step": 2504 }, { "epoch": 0.31866174786922785, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3560855388641357, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8453505039215088, "num_tokens": 95234602.0, "step": 2505 }, { "epoch": 0.3187889581478183, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.723403930664062e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.507826805114746, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8603084087371826, "num_tokens": 95275646.0, "step": 2506 }, { "epoch": 0.31891616842640885, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3588571548461914, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8490608930587769, "num_tokens": 95309948.0, "step": 2507 }, { "epoch": 0.3190433787049994, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.519641876220703, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8516067266464233, "num_tokens": 95347506.0, "step": 2508 }, { "epoch": 0.31917058898358985, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3479344844818115, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8509052395820618, "num_tokens": 95386809.0, "step": 2509 }, { "epoch": 0.3192977992621804, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5362231731414795, "learning_rate": 1e-05, "loss": 0.5503, "mean_token_accuracy": 0.8442923426628113, "num_tokens": 95429287.0, "step": 2510 }, { "epoch": 0.3194250095407709, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.3361027240753174, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8721363544464111, "num_tokens": 95466443.0, "step": 2511 }, { "epoch": 0.3195522198193614, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5346288681030273, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8575257658958435, "num_tokens": 95511593.0, "step": 2512 }, { "epoch": 0.3196794300979519, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.33467435836792, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8526665568351746, "num_tokens": 95546756.0, "step": 2513 }, { "epoch": 0.31980664037654244, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5449070930480957, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8521714806556702, "num_tokens": 95586434.0, "step": 2514 }, { "epoch": 0.3199338506551329, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3306446075439453, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8704110383987427, "num_tokens": 95616980.0, "step": 2515 }, { "epoch": 0.32006106093372344, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5699589252471924, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8518900871276855, "num_tokens": 95656919.0, "step": 2516 }, { "epoch": 0.32018827121231397, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3265717029571533, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8541947603225708, "num_tokens": 95691785.0, "step": 2517 }, { "epoch": 0.32031548149090444, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5837087631225586, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8496382236480713, "num_tokens": 95731055.0, "step": 2518 }, { "epoch": 0.32044269176949497, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3122916221618652, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8717653751373291, "num_tokens": 95767609.0, "step": 2519 }, { "epoch": 0.3205699020480855, "ewc_loss": 0.043701171875, "ewc_loss_diag": 6.771087646484375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6107122898101807, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8630102872848511, "num_tokens": 95805156.0, "step": 2520 }, { "epoch": 0.32069711232667597, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3133959770202637, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8628289103507996, "num_tokens": 95848809.0, "step": 2521 }, { "epoch": 0.3208243226052665, "ewc_loss": 0.044189453125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6519649028778076, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8467321991920471, "num_tokens": 95882961.0, "step": 2522 }, { "epoch": 0.320951532883857, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.353604555130005, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8646787405014038, "num_tokens": 95916360.0, "step": 2523 }, { "epoch": 0.3210787431624475, "ewc_loss": 0.044189453125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.652294397354126, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.854266881942749, "num_tokens": 95952064.0, "step": 2524 }, { "epoch": 0.32120595344103803, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3359646797180176, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8527308702468872, "num_tokens": 95993073.0, "step": 2525 }, { "epoch": 0.32133316371962856, "ewc_loss": 0.044677734375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6652798652648926, "learning_rate": 1e-05, "loss": 0.5191, "mean_token_accuracy": 0.8406528234481812, "num_tokens": 96028659.0, "step": 2526 }, { "epoch": 0.32146037399821903, "ewc_loss": 0.04150390625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.373893976211548, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8471285700798035, "num_tokens": 96064802.0, "step": 2527 }, { "epoch": 0.32158758427680956, "ewc_loss": 0.044677734375, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.638622999191284, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8433979153633118, "num_tokens": 96112980.0, "step": 2528 }, { "epoch": 0.3217147945554001, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3914988040924072, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8412091732025146, "num_tokens": 96144114.0, "step": 2529 }, { "epoch": 0.3218420048339906, "ewc_loss": 0.044189453125, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.628567934036255, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8565974831581116, "num_tokens": 96177383.0, "step": 2530 }, { "epoch": 0.3219692151125811, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3864331245422363, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8566693663597107, "num_tokens": 96216513.0, "step": 2531 }, { "epoch": 0.3220964253911716, "ewc_loss": 0.043212890625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5948352813720703, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8592046499252319, "num_tokens": 96256588.0, "step": 2532 }, { "epoch": 0.32222363566976214, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3853702545166016, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8506456017494202, "num_tokens": 96295365.0, "step": 2533 }, { "epoch": 0.3223508459483526, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.6034066677093506, "learning_rate": 1e-05, "loss": 0.5187, "mean_token_accuracy": 0.8482161164283752, "num_tokens": 96328048.0, "step": 2534 }, { "epoch": 0.32247805622694314, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3876121044158936, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8529223203659058, "num_tokens": 96362088.0, "step": 2535 }, { "epoch": 0.32260526650553367, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5930230617523193, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8481205701828003, "num_tokens": 96396161.0, "step": 2536 }, { "epoch": 0.32273247678412414, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3640835285186768, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8535548448562622, "num_tokens": 96433402.0, "step": 2537 }, { "epoch": 0.3228596870627147, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.578213691711426, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.845052182674408, "num_tokens": 96473011.0, "step": 2538 }, { "epoch": 0.3229868973413052, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.355816125869751, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8598412275314331, "num_tokens": 96511595.0, "step": 2539 }, { "epoch": 0.3231141076198957, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.60090708732605, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8478496074676514, "num_tokens": 96546192.0, "step": 2540 }, { "epoch": 0.3232413178984862, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3423471450805664, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8778947591781616, "num_tokens": 96582613.0, "step": 2541 }, { "epoch": 0.32336852817707673, "ewc_loss": 0.043212890625, "ewc_loss_diag": 6.818771362304688e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.61234974861145, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.847259521484375, "num_tokens": 96620219.0, "step": 2542 }, { "epoch": 0.3234957384556672, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3475170135498047, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8487591743469238, "num_tokens": 96661126.0, "step": 2543 }, { "epoch": 0.32362294873425773, "ewc_loss": 0.04345703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6224615573883057, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8444911241531372, "num_tokens": 96697995.0, "step": 2544 }, { "epoch": 0.32375015901284826, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3667054176330566, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8549262881278992, "num_tokens": 96732542.0, "step": 2545 }, { "epoch": 0.32387736929143873, "ewc_loss": 0.043701171875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6438374519348145, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8418710827827454, "num_tokens": 96761787.0, "step": 2546 }, { "epoch": 0.32400457957002926, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.355132579803467, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8478248119354248, "num_tokens": 96804918.0, "step": 2547 }, { "epoch": 0.3241317898486198, "ewc_loss": 0.0439453125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.629176616668701, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8534057140350342, "num_tokens": 96843646.0, "step": 2548 }, { "epoch": 0.32425900012721026, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3898746967315674, "learning_rate": 1e-05, "loss": 0.545, "mean_token_accuracy": 0.832888126373291, "num_tokens": 96874796.0, "step": 2549 }, { "epoch": 0.3243862104058008, "ewc_loss": 0.0439453125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.601693630218506, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.85910964012146, "num_tokens": 96912777.0, "step": 2550 }, { "epoch": 0.3245134206843913, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.374544620513916, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8463089466094971, "num_tokens": 96952045.0, "step": 2551 }, { "epoch": 0.3246406309629818, "ewc_loss": 0.04345703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.591984987258911, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8570079207420349, "num_tokens": 96984625.0, "step": 2552 }, { "epoch": 0.3247678412415723, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3754262924194336, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8594750165939331, "num_tokens": 97022409.0, "step": 2553 }, { "epoch": 0.32489505152016285, "ewc_loss": 0.043212890625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.577275514602661, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8467626571655273, "num_tokens": 97062461.0, "step": 2554 }, { "epoch": 0.3250222617987533, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3773293495178223, "learning_rate": 1e-05, "loss": 0.5642, "mean_token_accuracy": 0.8286082744598389, "num_tokens": 97102010.0, "step": 2555 }, { "epoch": 0.32514947207734385, "ewc_loss": 0.043212890625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5590364933013916, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8560899496078491, "num_tokens": 97140304.0, "step": 2556 }, { "epoch": 0.3252766823559344, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.355614185333252, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8726387023925781, "num_tokens": 97180460.0, "step": 2557 }, { "epoch": 0.32540389263452485, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5658462047576904, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8441990613937378, "num_tokens": 97223439.0, "step": 2558 }, { "epoch": 0.3255311029131154, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.387892484664917, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8382048010826111, "num_tokens": 97257050.0, "step": 2559 }, { "epoch": 0.3256583131917059, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5704116821289062, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8461006283760071, "num_tokens": 97296790.0, "step": 2560 }, { "epoch": 0.3257855234702964, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.353550672531128, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8497412800788879, "num_tokens": 97336511.0, "step": 2561 }, { "epoch": 0.3259127337488869, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5607657432556152, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8513407707214355, "num_tokens": 97373589.0, "step": 2562 }, { "epoch": 0.32603994402747744, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3440322875976562, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8679125905036926, "num_tokens": 97413986.0, "step": 2563 }, { "epoch": 0.3261671543060679, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5654425621032715, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8468616008758545, "num_tokens": 97457878.0, "step": 2564 }, { "epoch": 0.32629436458465844, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.341181993484497, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.856126070022583, "num_tokens": 97495463.0, "step": 2565 }, { "epoch": 0.32642157486324896, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5865423679351807, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8456076979637146, "num_tokens": 97533748.0, "step": 2566 }, { "epoch": 0.32654878514183944, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3402750492095947, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8543636798858643, "num_tokens": 97572476.0, "step": 2567 }, { "epoch": 0.32667599542042997, "ewc_loss": 0.04296875, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5954389572143555, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8575416803359985, "num_tokens": 97606658.0, "step": 2568 }, { "epoch": 0.3268032056990205, "ewc_loss": 0.040771484375, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.337851047515869, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.858217179775238, "num_tokens": 97645083.0, "step": 2569 }, { "epoch": 0.32693041597761097, "ewc_loss": 0.04345703125, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6130523681640625, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8505997061729431, "num_tokens": 97679440.0, "step": 2570 }, { "epoch": 0.3270576262562015, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.866455078125e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3528552055358887, "learning_rate": 1e-05, "loss": 0.5597, "mean_token_accuracy": 0.83282470703125, "num_tokens": 97719295.0, "step": 2571 }, { "epoch": 0.327184836534792, "ewc_loss": 0.04345703125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.60225772857666, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8608638644218445, "num_tokens": 97759417.0, "step": 2572 }, { "epoch": 0.3273120468133825, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3395118713378906, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8407434821128845, "num_tokens": 97804004.0, "step": 2573 }, { "epoch": 0.327439257091973, "ewc_loss": 0.0439453125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.620083808898926, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.8372113108634949, "num_tokens": 97845496.0, "step": 2574 }, { "epoch": 0.32756646737056355, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3526134490966797, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8507781624794006, "num_tokens": 97887184.0, "step": 2575 }, { "epoch": 0.327693677649154, "ewc_loss": 0.0439453125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.632864236831665, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.84108567237854, "num_tokens": 97924498.0, "step": 2576 }, { "epoch": 0.32782088792774455, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3607470989227295, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8583366870880127, "num_tokens": 97960225.0, "step": 2577 }, { "epoch": 0.3279480982063351, "ewc_loss": 0.043701171875, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6179375648498535, "learning_rate": 1e-05, "loss": 0.527, "mean_token_accuracy": 0.8359518647193909, "num_tokens": 98001746.0, "step": 2578 }, { "epoch": 0.3280753084849256, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.375221014022827, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8575619459152222, "num_tokens": 98038812.0, "step": 2579 }, { "epoch": 0.3282025187635161, "ewc_loss": 0.043212890625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.6103076934814453, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8497947454452515, "num_tokens": 98073782.0, "step": 2580 }, { "epoch": 0.3283297290421066, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3677263259887695, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8645508289337158, "num_tokens": 98109379.0, "step": 2581 }, { "epoch": 0.32845693932069714, "ewc_loss": 0.043212890625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.6024320125579834, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8639948964118958, "num_tokens": 98141661.0, "step": 2582 }, { "epoch": 0.3285841495992876, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.4488677978515625, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8407381176948547, "num_tokens": 98182634.0, "step": 2583 }, { "epoch": 0.32871135987787814, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.529390811920166, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8550624847412109, "num_tokens": 98225356.0, "step": 2584 }, { "epoch": 0.32883857015646867, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.396308183670044, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8532295823097229, "num_tokens": 98263575.0, "step": 2585 }, { "epoch": 0.32896578043505914, "ewc_loss": 0.042236328125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4989352226257324, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8486077189445496, "num_tokens": 98304993.0, "step": 2586 }, { "epoch": 0.32909299071364967, "ewc_loss": 0.041259765625, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3902108669281006, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8566323518753052, "num_tokens": 98339963.0, "step": 2587 }, { "epoch": 0.3292202009922402, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.492940902709961, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8580007553100586, "num_tokens": 98375507.0, "step": 2588 }, { "epoch": 0.32934741127083067, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.379436731338501, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8542436361312866, "num_tokens": 98413577.0, "step": 2589 }, { "epoch": 0.3294746215494212, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.486546754837036, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8429747223854065, "num_tokens": 98451615.0, "step": 2590 }, { "epoch": 0.3296018318280117, "ewc_loss": 0.040283203125, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.3837623596191406, "learning_rate": 1e-05, "loss": 0.5741, "mean_token_accuracy": 0.828118085861206, "num_tokens": 98490728.0, "step": 2591 }, { "epoch": 0.3297290421066022, "ewc_loss": 0.041748046875, "ewc_loss_diag": 6.914138793945312e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.48156476020813, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8666394948959351, "num_tokens": 98526645.0, "step": 2592 }, { "epoch": 0.3298562523851927, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.360525608062744, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8441196084022522, "num_tokens": 98568447.0, "step": 2593 }, { "epoch": 0.32998346266378326, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.490591287612915, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.852897047996521, "num_tokens": 98608952.0, "step": 2594 }, { "epoch": 0.33011067294237373, "ewc_loss": 0.041015625, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.342026948928833, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8711251020431519, "num_tokens": 98649180.0, "step": 2595 }, { "epoch": 0.33023788322096426, "ewc_loss": 0.0419921875, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.5082974433898926, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8515905737876892, "num_tokens": 98692506.0, "step": 2596 }, { "epoch": 0.3303650934995548, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.347600221633911, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8481954336166382, "num_tokens": 98727817.0, "step": 2597 }, { "epoch": 0.33049230377814526, "ewc_loss": 0.04248046875, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.504845142364502, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8559874892234802, "num_tokens": 98769714.0, "step": 2598 }, { "epoch": 0.3306195140567358, "ewc_loss": 0.04052734375, "ewc_loss_diag": 6.961822509765625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.328430652618408, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.862076997756958, "num_tokens": 98811894.0, "step": 2599 }, { "epoch": 0.3307467243353263, "ewc_loss": 0.042724609375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.550184965133667, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8452538251876831, "num_tokens": 98847117.0, "step": 2600 }, { "epoch": 0.3308739346139168, "ewc_loss": 0.040771484375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.32968807220459, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8543634414672852, "num_tokens": 98882745.0, "step": 2601 }, { "epoch": 0.3310011448925073, "ewc_loss": 0.042724609375, "ewc_loss_diag": 7.009506225585938e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.568844795227051, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8628479242324829, "num_tokens": 98918039.0, "step": 2602 }, { "epoch": 0.33112835517109784, "ewc_loss": 0.04052734375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.3072574138641357, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8539834022521973, "num_tokens": 98958571.0, "step": 2603 }, { "epoch": 0.3312555654496883, "ewc_loss": 0.043701171875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.584702730178833, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8546383380889893, "num_tokens": 99004970.0, "step": 2604 }, { "epoch": 0.33138277572827884, "ewc_loss": 0.04052734375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.331447124481201, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8418671488761902, "num_tokens": 99041112.0, "step": 2605 }, { "epoch": 0.3315099860068694, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6127374172210693, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8549880981445312, "num_tokens": 99076771.0, "step": 2606 }, { "epoch": 0.33163719628545985, "ewc_loss": 0.04052734375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.334749698638916, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8505063056945801, "num_tokens": 99112183.0, "step": 2607 }, { "epoch": 0.3317644065640504, "ewc_loss": 0.044189453125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6231179237365723, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8535821437835693, "num_tokens": 99148789.0, "step": 2608 }, { "epoch": 0.3318916168426409, "ewc_loss": 0.040771484375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3400373458862305, "learning_rate": 1e-05, "loss": 0.5585, "mean_token_accuracy": 0.8300089836120605, "num_tokens": 99189697.0, "step": 2609 }, { "epoch": 0.3320188271212314, "ewc_loss": 0.044189453125, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6409566402435303, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8488311767578125, "num_tokens": 99228269.0, "step": 2610 }, { "epoch": 0.3321460373998219, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3406903743743896, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8522570133209229, "num_tokens": 99264275.0, "step": 2611 }, { "epoch": 0.33227324767841243, "ewc_loss": 0.044677734375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.66117525100708, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8557517528533936, "num_tokens": 99294598.0, "step": 2612 }, { "epoch": 0.3324004579570029, "ewc_loss": 0.041259765625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.34151554107666, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8576185703277588, "num_tokens": 99336307.0, "step": 2613 }, { "epoch": 0.33252766823559343, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6583707332611084, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8641457557678223, "num_tokens": 99373983.0, "step": 2614 }, { "epoch": 0.33265487851418396, "ewc_loss": 0.041748046875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3632044792175293, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8461414575576782, "num_tokens": 99415794.0, "step": 2615 }, { "epoch": 0.33278208879277443, "ewc_loss": 0.04541015625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6618764400482178, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8458822965621948, "num_tokens": 99457572.0, "step": 2616 }, { "epoch": 0.33290929907136496, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.378221273422241, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8529147505760193, "num_tokens": 99496074.0, "step": 2617 }, { "epoch": 0.3330365093499555, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.636683464050293, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8446959257125854, "num_tokens": 99539655.0, "step": 2618 }, { "epoch": 0.33316371962854596, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.411869764328003, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8456884622573853, "num_tokens": 99574696.0, "step": 2619 }, { "epoch": 0.3332909299071365, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.604381799697876, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.855157732963562, "num_tokens": 99612960.0, "step": 2620 }, { "epoch": 0.333418140185727, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.4219236373901367, "learning_rate": 1e-05, "loss": 0.563, "mean_token_accuracy": 0.8280571103096008, "num_tokens": 99644107.0, "step": 2621 }, { "epoch": 0.3335453504643175, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5589184761047363, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8503284454345703, "num_tokens": 99682014.0, "step": 2622 }, { "epoch": 0.333672560742908, "ewc_loss": 0.041748046875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.410036563873291, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8459632396697998, "num_tokens": 99718969.0, "step": 2623 }, { "epoch": 0.33379977102149855, "ewc_loss": 0.042724609375, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5359389781951904, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8377265334129333, "num_tokens": 99756955.0, "step": 2624 }, { "epoch": 0.333926981300089, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3961329460144043, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8591892719268799, "num_tokens": 99792216.0, "step": 2625 }, { "epoch": 0.33405419157867955, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.525953769683838, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8436784744262695, "num_tokens": 99823973.0, "step": 2626 }, { "epoch": 0.3341814018572701, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.407921552658081, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8465874195098877, "num_tokens": 99857206.0, "step": 2627 }, { "epoch": 0.33430861213586055, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.524439573287964, "learning_rate": 1e-05, "loss": 0.5535, "mean_token_accuracy": 0.8351101279258728, "num_tokens": 99891146.0, "step": 2628 }, { "epoch": 0.3344358224144511, "ewc_loss": 0.041259765625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3769495487213135, "learning_rate": 1e-05, "loss": 0.5496, "mean_token_accuracy": 0.8338615298271179, "num_tokens": 99931456.0, "step": 2629 }, { "epoch": 0.3345630326930416, "ewc_loss": 0.042724609375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.526207685470581, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8464431762695312, "num_tokens": 99965968.0, "step": 2630 }, { "epoch": 0.33469024297163213, "ewc_loss": 0.041259765625, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.352025270462036, "learning_rate": 1e-05, "loss": 0.5246, "mean_token_accuracy": 0.8404240608215332, "num_tokens": 100010874.0, "step": 2631 }, { "epoch": 0.3348174532502226, "ewc_loss": 0.042724609375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5198607444763184, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8576540946960449, "num_tokens": 100052166.0, "step": 2632 }, { "epoch": 0.33494466352881314, "ewc_loss": 0.040771484375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3580567836761475, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8567318320274353, "num_tokens": 100093927.0, "step": 2633 }, { "epoch": 0.33507187380740366, "ewc_loss": 0.042724609375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5432939529418945, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8459756970405579, "num_tokens": 100128414.0, "step": 2634 }, { "epoch": 0.33519908408599414, "ewc_loss": 0.040771484375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.356374979019165, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8594903349876404, "num_tokens": 100160334.0, "step": 2635 }, { "epoch": 0.33532629436458466, "ewc_loss": 0.042724609375, "ewc_loss_diag": 7.05718994140625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5453832149505615, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8523952960968018, "num_tokens": 100195188.0, "step": 2636 }, { "epoch": 0.3354535046431752, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.104873657226562e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3367104530334473, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8569031953811646, "num_tokens": 100235338.0, "step": 2637 }, { "epoch": 0.33558071492176567, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5598092079162598, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8547475337982178, "num_tokens": 100276267.0, "step": 2638 }, { "epoch": 0.3357079252003562, "ewc_loss": 0.040771484375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3527297973632812, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8324310779571533, "num_tokens": 100316201.0, "step": 2639 }, { "epoch": 0.3358351354789467, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.572049856185913, "learning_rate": 1e-05, "loss": 0.5435, "mean_token_accuracy": 0.8384194374084473, "num_tokens": 100353278.0, "step": 2640 }, { "epoch": 0.3359623457575372, "ewc_loss": 0.040771484375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3714599609375, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8472989797592163, "num_tokens": 100384185.0, "step": 2641 }, { "epoch": 0.3360895560361277, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5629703998565674, "learning_rate": 1e-05, "loss": 0.5347, "mean_token_accuracy": 0.8345388770103455, "num_tokens": 100423232.0, "step": 2642 }, { "epoch": 0.33621676631471825, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3874247074127197, "learning_rate": 1e-05, "loss": 0.5535, "mean_token_accuracy": 0.8294274806976318, "num_tokens": 100457912.0, "step": 2643 }, { "epoch": 0.3363439765933087, "ewc_loss": 0.042724609375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5374269485473633, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8546526432037354, "num_tokens": 100499894.0, "step": 2644 }, { "epoch": 0.33647118687189925, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.344684600830078, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8652693033218384, "num_tokens": 100540961.0, "step": 2645 }, { "epoch": 0.3365983971504898, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.552319288253784, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8459700345993042, "num_tokens": 100581302.0, "step": 2646 }, { "epoch": 0.33672560742908025, "ewc_loss": 0.040771484375, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.353522539138794, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.856826901435852, "num_tokens": 100615289.0, "step": 2647 }, { "epoch": 0.3368528177076708, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5393896102905273, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8612719774246216, "num_tokens": 100659911.0, "step": 2648 }, { "epoch": 0.3369800279862613, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.152557373046875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3394131660461426, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8537978529930115, "num_tokens": 100700861.0, "step": 2649 }, { "epoch": 0.3371072382648518, "ewc_loss": 0.04345703125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.566650629043579, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8506591320037842, "num_tokens": 100737329.0, "step": 2650 }, { "epoch": 0.3372344485434423, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3360509872436523, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8687434196472168, "num_tokens": 100774278.0, "step": 2651 }, { "epoch": 0.33736165882203284, "ewc_loss": 0.04345703125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5679948329925537, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.848778486251831, "num_tokens": 100819083.0, "step": 2652 }, { "epoch": 0.3374888691006233, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3413166999816895, "learning_rate": 1e-05, "loss": 0.5493, "mean_token_accuracy": 0.8379952311515808, "num_tokens": 100859711.0, "step": 2653 }, { "epoch": 0.33761607937921384, "ewc_loss": 0.04345703125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.6074764728546143, "learning_rate": 1e-05, "loss": 0.5566, "mean_token_accuracy": 0.8286476731300354, "num_tokens": 100893915.0, "step": 2654 }, { "epoch": 0.33774328965780437, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3283839225769043, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.859619140625, "num_tokens": 100932609.0, "step": 2655 }, { "epoch": 0.33787049993639484, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.625783920288086, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8547396659851074, "num_tokens": 100966539.0, "step": 2656 }, { "epoch": 0.33799771021498537, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3499205112457275, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.839322566986084, "num_tokens": 100999979.0, "step": 2657 }, { "epoch": 0.3381249204935759, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6324734687805176, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.8442889451980591, "num_tokens": 101036546.0, "step": 2658 }, { "epoch": 0.33825213077216637, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.354862689971924, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8504189252853394, "num_tokens": 101067861.0, "step": 2659 }, { "epoch": 0.3383793410507569, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6206579208374023, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.871846616268158, "num_tokens": 101104268.0, "step": 2660 }, { "epoch": 0.3385065513293474, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3453049659729004, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8693498373031616, "num_tokens": 101143364.0, "step": 2661 }, { "epoch": 0.3386337616079379, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.7317707538604736, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8405982851982117, "num_tokens": 101178359.0, "step": 2662 }, { "epoch": 0.33876097188652843, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3537023067474365, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8641656637191772, "num_tokens": 101215993.0, "step": 2663 }, { "epoch": 0.33888818216511896, "ewc_loss": 0.0458984375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.675034999847412, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8645703196525574, "num_tokens": 101254924.0, "step": 2664 }, { "epoch": 0.33901539244370943, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3542628288269043, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8621695041656494, "num_tokens": 101295632.0, "step": 2665 }, { "epoch": 0.33914260272229996, "ewc_loss": 0.04638671875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6874139308929443, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8588660955429077, "num_tokens": 101333295.0, "step": 2666 }, { "epoch": 0.3392698130008905, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3821098804473877, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8563694953918457, "num_tokens": 101372319.0, "step": 2667 }, { "epoch": 0.33939702327948096, "ewc_loss": 0.04638671875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.684309482574463, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8538888096809387, "num_tokens": 101405300.0, "step": 2668 }, { "epoch": 0.3395242335580715, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4061484336853027, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8387171030044556, "num_tokens": 101445576.0, "step": 2669 }, { "epoch": 0.339651443836662, "ewc_loss": 0.04541015625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6693613529205322, "learning_rate": 1e-05, "loss": 0.5615, "mean_token_accuracy": 0.8336908221244812, "num_tokens": 101481840.0, "step": 2670 }, { "epoch": 0.3397786541152525, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.437394618988037, "learning_rate": 1e-05, "loss": 0.5233, "mean_token_accuracy": 0.8424704074859619, "num_tokens": 101511968.0, "step": 2671 }, { "epoch": 0.339905864393843, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.603367567062378, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8556069135665894, "num_tokens": 101549909.0, "step": 2672 }, { "epoch": 0.34003307467243354, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.433474063873291, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8441800475120544, "num_tokens": 101580892.0, "step": 2673 }, { "epoch": 0.340160284951024, "ewc_loss": 0.04345703125, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5703818798065186, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8568722009658813, "num_tokens": 101614589.0, "step": 2674 }, { "epoch": 0.34028749522961454, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.394084930419922, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8562583327293396, "num_tokens": 101658988.0, "step": 2675 }, { "epoch": 0.3404147055082051, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5551578998565674, "learning_rate": 1e-05, "loss": 0.5579, "mean_token_accuracy": 0.8358694314956665, "num_tokens": 101697838.0, "step": 2676 }, { "epoch": 0.34054191578679555, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.387207269668579, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8606473207473755, "num_tokens": 101733831.0, "step": 2677 }, { "epoch": 0.3406691260653861, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5418310165405273, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.854121208190918, "num_tokens": 101772697.0, "step": 2678 }, { "epoch": 0.3407963363439766, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.387664556503296, "learning_rate": 1e-05, "loss": 0.5566, "mean_token_accuracy": 0.8347400426864624, "num_tokens": 101810107.0, "step": 2679 }, { "epoch": 0.34092354662256713, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5455660820007324, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8563501834869385, "num_tokens": 101847239.0, "step": 2680 }, { "epoch": 0.3410507569011576, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.381143569946289, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8414961099624634, "num_tokens": 101884508.0, "step": 2681 }, { "epoch": 0.34117796717974813, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5473246574401855, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8761783838272095, "num_tokens": 101925351.0, "step": 2682 }, { "epoch": 0.34130517745833866, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3541605472564697, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8549160361289978, "num_tokens": 101966086.0, "step": 2683 }, { "epoch": 0.34143238773692913, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.200241088867188e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.54421067237854, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8576422929763794, "num_tokens": 102004935.0, "step": 2684 }, { "epoch": 0.34155959801551966, "ewc_loss": 0.041259765625, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.364971399307251, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8503581285476685, "num_tokens": 102039026.0, "step": 2685 }, { "epoch": 0.3416868082941102, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.54826283454895, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8497141003608704, "num_tokens": 102078990.0, "step": 2686 }, { "epoch": 0.34181401857270066, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3647847175598145, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8563241958618164, "num_tokens": 102114631.0, "step": 2687 }, { "epoch": 0.3419412288512912, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5307209491729736, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8530874252319336, "num_tokens": 102161583.0, "step": 2688 }, { "epoch": 0.3420684391298817, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.345161199569702, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8575140237808228, "num_tokens": 102202183.0, "step": 2689 }, { "epoch": 0.3421956494084722, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5446977615356445, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8594487905502319, "num_tokens": 102235288.0, "step": 2690 }, { "epoch": 0.3423228596870627, "ewc_loss": 0.041259765625, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3678102493286133, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8472236394882202, "num_tokens": 102272814.0, "step": 2691 }, { "epoch": 0.34245006996565325, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.543084144592285, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8501202464103699, "num_tokens": 102312763.0, "step": 2692 }, { "epoch": 0.3425772802442437, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3488550186157227, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8567671775817871, "num_tokens": 102351490.0, "step": 2693 }, { "epoch": 0.34270449052283425, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.539348602294922, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8509316444396973, "num_tokens": 102394495.0, "step": 2694 }, { "epoch": 0.3428317008014248, "ewc_loss": 0.041259765625, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3549644947052, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8459848165512085, "num_tokens": 102431689.0, "step": 2695 }, { "epoch": 0.34295891108001525, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5498745441436768, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8608524203300476, "num_tokens": 102468653.0, "step": 2696 }, { "epoch": 0.3430861213586058, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3635337352752686, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8667213916778564, "num_tokens": 102504442.0, "step": 2697 }, { "epoch": 0.3432133316371963, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5415306091308594, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8600239753723145, "num_tokens": 102539401.0, "step": 2698 }, { "epoch": 0.3433405419157868, "ewc_loss": 0.041259765625, "ewc_loss_diag": 7.2479248046875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3417651653289795, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8585357666015625, "num_tokens": 102579874.0, "step": 2699 }, { "epoch": 0.3434677521943773, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.557818651199341, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8475285768508911, "num_tokens": 102617906.0, "step": 2700 }, { "epoch": 0.34359496247296784, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3391010761260986, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8570475578308105, "num_tokens": 102657744.0, "step": 2701 }, { "epoch": 0.3437221727515583, "ewc_loss": 0.04345703125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.579348087310791, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8477341532707214, "num_tokens": 102696650.0, "step": 2702 }, { "epoch": 0.34384938303014884, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.342384099960327, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8505210280418396, "num_tokens": 102729114.0, "step": 2703 }, { "epoch": 0.34397659330873936, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.600404977798462, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8526557683944702, "num_tokens": 102763562.0, "step": 2704 }, { "epoch": 0.34410380358732984, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.332615852355957, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8436075448989868, "num_tokens": 102803680.0, "step": 2705 }, { "epoch": 0.34423101386592037, "ewc_loss": 0.044677734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6052701473236084, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8416793346405029, "num_tokens": 102847932.0, "step": 2706 }, { "epoch": 0.3443582241445109, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3303112983703613, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8585489392280579, "num_tokens": 102888371.0, "step": 2707 }, { "epoch": 0.34448543442310137, "ewc_loss": 0.044677734375, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.635915994644165, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.8343538045883179, "num_tokens": 102928943.0, "step": 2708 }, { "epoch": 0.3446126447016919, "ewc_loss": 0.041748046875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3399972915649414, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8510276675224304, "num_tokens": 102969773.0, "step": 2709 }, { "epoch": 0.3447398549802824, "ewc_loss": 0.045166015625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.648325204849243, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8402506113052368, "num_tokens": 103012496.0, "step": 2710 }, { "epoch": 0.3448670652588729, "ewc_loss": 0.041748046875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3634300231933594, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8473517298698425, "num_tokens": 103047259.0, "step": 2711 }, { "epoch": 0.3449942755374634, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6282689571380615, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8611844182014465, "num_tokens": 103087534.0, "step": 2712 }, { "epoch": 0.34512148581605395, "ewc_loss": 0.041748046875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3513216972351074, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.861376941204071, "num_tokens": 103128782.0, "step": 2713 }, { "epoch": 0.3452486960946444, "ewc_loss": 0.045166015625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6489715576171875, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8555821180343628, "num_tokens": 103161252.0, "step": 2714 }, { "epoch": 0.34537590637323495, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.364373207092285, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8598044514656067, "num_tokens": 103201524.0, "step": 2715 }, { "epoch": 0.3455031166518255, "ewc_loss": 0.045166015625, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.649840831756592, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8429949283599854, "num_tokens": 103235444.0, "step": 2716 }, { "epoch": 0.34563032693041595, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.381765127182007, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.83882737159729, "num_tokens": 103270927.0, "step": 2717 }, { "epoch": 0.3457575372090065, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.622903823852539, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8612804412841797, "num_tokens": 103310518.0, "step": 2718 }, { "epoch": 0.345884747487597, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3690876960754395, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8714879155158997, "num_tokens": 103347940.0, "step": 2719 }, { "epoch": 0.3460119577661875, "ewc_loss": 0.044677734375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6219067573547363, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8633847236633301, "num_tokens": 103383995.0, "step": 2720 }, { "epoch": 0.346139168044778, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.381350040435791, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8712859749794006, "num_tokens": 103416411.0, "step": 2721 }, { "epoch": 0.34626637832336854, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6171538829803467, "learning_rate": 1e-05, "loss": 0.5175, "mean_token_accuracy": 0.844908595085144, "num_tokens": 103454199.0, "step": 2722 }, { "epoch": 0.346393588601959, "ewc_loss": 0.041748046875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.39224910736084, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8432281017303467, "num_tokens": 103490996.0, "step": 2723 }, { "epoch": 0.34652079888054954, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5928561687469482, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8588734269142151, "num_tokens": 103525987.0, "step": 2724 }, { "epoch": 0.34664800915914007, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.295608520507812e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3816351890563965, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8412127494812012, "num_tokens": 103565702.0, "step": 2725 }, { "epoch": 0.34677521943773054, "ewc_loss": 0.043701171875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.590311050415039, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8534291982650757, "num_tokens": 103601764.0, "step": 2726 }, { "epoch": 0.34690242971632107, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.380594491958618, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8485420942306519, "num_tokens": 103640293.0, "step": 2727 }, { "epoch": 0.3470296399949116, "ewc_loss": 0.043212890625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5752317905426025, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.849166750907898, "num_tokens": 103680464.0, "step": 2728 }, { "epoch": 0.3471568502735021, "ewc_loss": 0.041748046875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3765182495117188, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8533450365066528, "num_tokens": 103717434.0, "step": 2729 }, { "epoch": 0.3472840605520926, "ewc_loss": 0.04345703125, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.569573402404785, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8552557229995728, "num_tokens": 103755186.0, "step": 2730 }, { "epoch": 0.3474112708306831, "ewc_loss": 0.041748046875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3699896335601807, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8669345378875732, "num_tokens": 103789516.0, "step": 2731 }, { "epoch": 0.34753848110927366, "ewc_loss": 0.043212890625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5666720867156982, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8461368083953857, "num_tokens": 103826103.0, "step": 2732 }, { "epoch": 0.34766569138786413, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.375574827194214, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8537123799324036, "num_tokens": 103863059.0, "step": 2733 }, { "epoch": 0.34779290166645466, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5592966079711914, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8592624664306641, "num_tokens": 103897976.0, "step": 2734 }, { "epoch": 0.3479201119450452, "ewc_loss": 0.041748046875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3719539642333984, "learning_rate": 1e-05, "loss": 0.5373, "mean_token_accuracy": 0.837279200553894, "num_tokens": 103935485.0, "step": 2735 }, { "epoch": 0.34804732222363566, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.556518316268921, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8556749820709229, "num_tokens": 103971917.0, "step": 2736 }, { "epoch": 0.3481745325022262, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3852477073669434, "learning_rate": 1e-05, "loss": 0.5275, "mean_token_accuracy": 0.8455027937889099, "num_tokens": 104008797.0, "step": 2737 }, { "epoch": 0.3483017427808167, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.541452646255493, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8448083996772766, "num_tokens": 104049551.0, "step": 2738 }, { "epoch": 0.3484289530594072, "ewc_loss": 0.041259765625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.368598461151123, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8350414633750916, "num_tokens": 104089341.0, "step": 2739 }, { "epoch": 0.3485561633379977, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5289266109466553, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8487824201583862, "num_tokens": 104133445.0, "step": 2740 }, { "epoch": 0.34868337361658824, "ewc_loss": 0.041259765625, "ewc_loss_diag": 7.343292236328125e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3587207794189453, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8537811040878296, "num_tokens": 104170435.0, "step": 2741 }, { "epoch": 0.3488105838951787, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5355873107910156, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8623827695846558, "num_tokens": 104205660.0, "step": 2742 }, { "epoch": 0.34893779417376924, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.343728542327881, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8624746799468994, "num_tokens": 104243749.0, "step": 2743 }, { "epoch": 0.3490650044523598, "ewc_loss": 0.04345703125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5493316650390625, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8524438142776489, "num_tokens": 104283411.0, "step": 2744 }, { "epoch": 0.34919221473095025, "ewc_loss": 0.041015625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3395440578460693, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8569011688232422, "num_tokens": 104324570.0, "step": 2745 }, { "epoch": 0.3493194250095408, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.567506790161133, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8473191261291504, "num_tokens": 104363619.0, "step": 2746 }, { "epoch": 0.3494466352881313, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.332622528076172, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8577282428741455, "num_tokens": 104405572.0, "step": 2747 }, { "epoch": 0.3495738455667218, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5871593952178955, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8474107980728149, "num_tokens": 104446772.0, "step": 2748 }, { "epoch": 0.3497010558453123, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3294644355773926, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8581658601760864, "num_tokens": 104490912.0, "step": 2749 }, { "epoch": 0.34982826612390283, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6128110885620117, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8598154187202454, "num_tokens": 104525307.0, "step": 2750 }, { "epoch": 0.3499554764024933, "ewc_loss": 0.04150390625, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3376123905181885, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8491858243942261, "num_tokens": 104562431.0, "step": 2751 }, { "epoch": 0.35008268668108383, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.626004457473755, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8698985576629639, "num_tokens": 104596440.0, "step": 2752 }, { "epoch": 0.35020989695967436, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3331565856933594, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8651426434516907, "num_tokens": 104630804.0, "step": 2753 }, { "epoch": 0.35033710723826483, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.639761447906494, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8618544340133667, "num_tokens": 104667997.0, "step": 2754 }, { "epoch": 0.35046431751685536, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.4039716720581055, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8591256141662598, "num_tokens": 104711605.0, "step": 2755 }, { "epoch": 0.3505915277954459, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.6111562252044678, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8482779860496521, "num_tokens": 104748427.0, "step": 2756 }, { "epoch": 0.35071873807403636, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.371858835220337, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8553131222724915, "num_tokens": 104784536.0, "step": 2757 }, { "epoch": 0.3508459483526269, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5963854789733887, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8577805757522583, "num_tokens": 104820096.0, "step": 2758 }, { "epoch": 0.3509731586312174, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.43865966796875e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3709754943847656, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8579115271568298, "num_tokens": 104857016.0, "step": 2759 }, { "epoch": 0.3511003689098079, "ewc_loss": 0.044189453125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5750033855438232, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8541918992996216, "num_tokens": 104903999.0, "step": 2760 }, { "epoch": 0.3512275791883984, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.375530481338501, "learning_rate": 1e-05, "loss": 0.5544, "mean_token_accuracy": 0.834830641746521, "num_tokens": 104943954.0, "step": 2761 }, { "epoch": 0.35135478946698895, "ewc_loss": 0.043701171875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5799713134765625, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8436580896377563, "num_tokens": 104980028.0, "step": 2762 }, { "epoch": 0.3514819997455794, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3677854537963867, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8572520017623901, "num_tokens": 105020471.0, "step": 2763 }, { "epoch": 0.35160921002416995, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 3.3463327884674072, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8550585508346558, "num_tokens": 105050419.0, "step": 2764 }, { "epoch": 0.3517364203027605, "ewc_loss": 0.046630859375, "ewc_loss_diag": 7.486343383789062e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.536515235900879, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8611819744110107, "num_tokens": 105079656.0, "step": 2765 }, { "epoch": 0.35186363058135095, "ewc_loss": 0.04638671875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.7172651290893555, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8570247888565063, "num_tokens": 105119715.0, "step": 2766 }, { "epoch": 0.3519908408599415, "ewc_loss": 0.043212890625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4398062229156494, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8466697931289673, "num_tokens": 105158281.0, "step": 2767 }, { "epoch": 0.352118051138532, "ewc_loss": 0.049072265625, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.731407403945923, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8467930555343628, "num_tokens": 105201372.0, "step": 2768 }, { "epoch": 0.3522452614171225, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.452713966369629, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8474252223968506, "num_tokens": 105240557.0, "step": 2769 }, { "epoch": 0.352372471695713, "ewc_loss": 0.047607421875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6958327293395996, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8518399596214294, "num_tokens": 105276701.0, "step": 2770 }, { "epoch": 0.35249968197430354, "ewc_loss": 0.043701171875, "ewc_loss_diag": 7.534027099609375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4533278942108154, "learning_rate": 1e-05, "loss": 0.5434, "mean_token_accuracy": 0.8415560722351074, "num_tokens": 105319185.0, "step": 2771 }, { "epoch": 0.352626892252894, "ewc_loss": 0.046875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6346054077148438, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8703897595405579, "num_tokens": 105362800.0, "step": 2772 }, { "epoch": 0.35275410253148454, "ewc_loss": 0.043212890625, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.44016170501709, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8488908410072327, "num_tokens": 105400731.0, "step": 2773 }, { "epoch": 0.35288131281007507, "ewc_loss": 0.045654296875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.610764980316162, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.8461277484893799, "num_tokens": 105439712.0, "step": 2774 }, { "epoch": 0.35300852308866554, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.440139055252075, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8559561967849731, "num_tokens": 105475159.0, "step": 2775 }, { "epoch": 0.35313573336725607, "ewc_loss": 0.044677734375, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5843584537506104, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8430356979370117, "num_tokens": 105511323.0, "step": 2776 }, { "epoch": 0.3532629436458466, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.420149564743042, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8539377450942993, "num_tokens": 105544056.0, "step": 2777 }, { "epoch": 0.35339015392443707, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.581710815429688e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5473272800445557, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.868116021156311, "num_tokens": 105582335.0, "step": 2778 }, { "epoch": 0.3535173642030276, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.4043545722961426, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8598587512969971, "num_tokens": 105614046.0, "step": 2779 }, { "epoch": 0.3536445744816181, "ewc_loss": 0.043212890625, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5483484268188477, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8609327077865601, "num_tokens": 105648294.0, "step": 2780 }, { "epoch": 0.35377178476020865, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.387754201889038, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8498402833938599, "num_tokens": 105685707.0, "step": 2781 }, { "epoch": 0.3538989950387991, "ewc_loss": 0.04345703125, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5388267040252686, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8544743061065674, "num_tokens": 105724662.0, "step": 2782 }, { "epoch": 0.35402620531738965, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.62939453125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.4048566818237305, "learning_rate": 1e-05, "loss": 0.5379, "mean_token_accuracy": 0.8412443995475769, "num_tokens": 105756781.0, "step": 2783 }, { "epoch": 0.3541534155959802, "ewc_loss": 0.04345703125, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5131936073303223, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8622202277183533, "num_tokens": 105798466.0, "step": 2784 }, { "epoch": 0.35428062587457065, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3748745918273926, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8576482534408569, "num_tokens": 105835085.0, "step": 2785 }, { "epoch": 0.3544078361531612, "ewc_loss": 0.04296875, "ewc_loss_diag": 7.677078247070312e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.517076253890991, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8630198240280151, "num_tokens": 105872268.0, "step": 2786 }, { "epoch": 0.3545350464317517, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3810524940490723, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8525872826576233, "num_tokens": 105905721.0, "step": 2787 }, { "epoch": 0.3546622567103422, "ewc_loss": 0.04345703125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.519089937210083, "learning_rate": 1e-05, "loss": 0.5364, "mean_token_accuracy": 0.8381004333496094, "num_tokens": 105945573.0, "step": 2788 }, { "epoch": 0.3547894669889327, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.374244451522827, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.851598858833313, "num_tokens": 105978800.0, "step": 2789 }, { "epoch": 0.35491667726752324, "ewc_loss": 0.04345703125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5218372344970703, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8532505035400391, "num_tokens": 106014035.0, "step": 2790 }, { "epoch": 0.3550438875461137, "ewc_loss": 0.041748046875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3723490238189697, "learning_rate": 1e-05, "loss": 0.5821, "mean_token_accuracy": 0.8278222680091858, "num_tokens": 106050721.0, "step": 2791 }, { "epoch": 0.35517109782470424, "ewc_loss": 0.04345703125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.579665422439575, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8571186065673828, "num_tokens": 106086802.0, "step": 2792 }, { "epoch": 0.35529830810329477, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.34401535987854, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.864234209060669, "num_tokens": 106124174.0, "step": 2793 }, { "epoch": 0.35542551838188524, "ewc_loss": 0.043701171875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.559537649154663, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8572334051132202, "num_tokens": 106166147.0, "step": 2794 }, { "epoch": 0.35555272866047577, "ewc_loss": 0.041748046875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3341734409332275, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8443505764007568, "num_tokens": 106206758.0, "step": 2795 }, { "epoch": 0.3556799389390663, "ewc_loss": 0.044677734375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5864760875701904, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.850562572479248, "num_tokens": 106241823.0, "step": 2796 }, { "epoch": 0.35580714921765677, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3448429107666016, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8648314476013184, "num_tokens": 106275716.0, "step": 2797 }, { "epoch": 0.3559343594962473, "ewc_loss": 0.044189453125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5962135791778564, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8448739051818848, "num_tokens": 106312595.0, "step": 2798 }, { "epoch": 0.3560615697748378, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3564393520355225, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8443930745124817, "num_tokens": 106346009.0, "step": 2799 }, { "epoch": 0.3561887800534283, "ewc_loss": 0.044189453125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.591871500015259, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8636341691017151, "num_tokens": 106384501.0, "step": 2800 }, { "epoch": 0.35631599033201883, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.353806257247925, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8544549345970154, "num_tokens": 106420099.0, "step": 2801 }, { "epoch": 0.35644320061060936, "ewc_loss": 0.044677734375, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.593337297439575, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8670039176940918, "num_tokens": 106459697.0, "step": 2802 }, { "epoch": 0.35657041088919983, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3652005195617676, "learning_rate": 1e-05, "loss": 0.5885, "mean_token_accuracy": 0.8278485536575317, "num_tokens": 106499054.0, "step": 2803 }, { "epoch": 0.35669762116779036, "ewc_loss": 0.044189453125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.582880973815918, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8571300506591797, "num_tokens": 106538704.0, "step": 2804 }, { "epoch": 0.3568248314463809, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3574538230895996, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8442667722702026, "num_tokens": 106580012.0, "step": 2805 }, { "epoch": 0.35695204172497136, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5934336185455322, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8558051586151123, "num_tokens": 106618160.0, "step": 2806 }, { "epoch": 0.3570792520035619, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.724761962890625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3750336170196533, "learning_rate": 1e-05, "loss": 0.5611, "mean_token_accuracy": 0.8284904956817627, "num_tokens": 106654387.0, "step": 2807 }, { "epoch": 0.3572064622821524, "ewc_loss": 0.044677734375, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5780539512634277, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8636099696159363, "num_tokens": 106693318.0, "step": 2808 }, { "epoch": 0.3573336725607429, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3781850337982178, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8459958434104919, "num_tokens": 106728779.0, "step": 2809 }, { "epoch": 0.3574608828393334, "ewc_loss": 0.043701171875, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.567162275314331, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8537247776985168, "num_tokens": 106771149.0, "step": 2810 }, { "epoch": 0.35758809311792394, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3771705627441406, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8413922786712646, "num_tokens": 106809132.0, "step": 2811 }, { "epoch": 0.3577153033965144, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.55997633934021, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8695740103721619, "num_tokens": 106842747.0, "step": 2812 }, { "epoch": 0.35784251367510495, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3628320693969727, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8403695225715637, "num_tokens": 106883519.0, "step": 2813 }, { "epoch": 0.3579697239536955, "ewc_loss": 0.044189453125, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5774285793304443, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.840900182723999, "num_tokens": 106914818.0, "step": 2814 }, { "epoch": 0.35809693423228595, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.772445678710938e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3471086025238037, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.863219141960144, "num_tokens": 106954387.0, "step": 2815 }, { "epoch": 0.3582241445108765, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5913712978363037, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8448120355606079, "num_tokens": 106994264.0, "step": 2816 }, { "epoch": 0.358351354789467, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.336232900619507, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8588376045227051, "num_tokens": 107035575.0, "step": 2817 }, { "epoch": 0.3584785650680575, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6165812015533447, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.843351423740387, "num_tokens": 107074644.0, "step": 2818 }, { "epoch": 0.358605775346648, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3506674766540527, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8562959432601929, "num_tokens": 107108758.0, "step": 2819 }, { "epoch": 0.35873298562523853, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.82012939453125e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.606182336807251, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8702616095542908, "num_tokens": 107147572.0, "step": 2820 }, { "epoch": 0.358860195903829, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.369921922683716, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8457282781600952, "num_tokens": 107184928.0, "step": 2821 }, { "epoch": 0.35898740618241953, "ewc_loss": 0.044677734375, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5963761806488037, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8658185601234436, "num_tokens": 107223462.0, "step": 2822 }, { "epoch": 0.35911461646101006, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.357944965362549, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8570092916488647, "num_tokens": 107263114.0, "step": 2823 }, { "epoch": 0.35924182673960053, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5974197387695312, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8724640607833862, "num_tokens": 107301234.0, "step": 2824 }, { "epoch": 0.35936903701819106, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.385725498199463, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8570960760116577, "num_tokens": 107334903.0, "step": 2825 }, { "epoch": 0.3594962472967816, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5730721950531006, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8723562955856323, "num_tokens": 107374989.0, "step": 2826 }, { "epoch": 0.35962345757537206, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3669044971466064, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8620509505271912, "num_tokens": 107412545.0, "step": 2827 }, { "epoch": 0.3597506678539626, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.597477436065674, "learning_rate": 1e-05, "loss": 0.5437, "mean_token_accuracy": 0.8356187343597412, "num_tokens": 107447642.0, "step": 2828 }, { "epoch": 0.3598778781325531, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3604214191436768, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8555788993835449, "num_tokens": 107486514.0, "step": 2829 }, { "epoch": 0.36000508841114365, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5915727615356445, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8468649387359619, "num_tokens": 107527351.0, "step": 2830 }, { "epoch": 0.3601322986897341, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.369581460952759, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8568894863128662, "num_tokens": 107558973.0, "step": 2831 }, { "epoch": 0.36025950896832465, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6007466316223145, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8495211601257324, "num_tokens": 107597659.0, "step": 2832 }, { "epoch": 0.3603867192469152, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.363039970397949, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8506706953048706, "num_tokens": 107635833.0, "step": 2833 }, { "epoch": 0.36051392952550565, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.593764305114746, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8622349500656128, "num_tokens": 107673347.0, "step": 2834 }, { "epoch": 0.3606411398040962, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.359266757965088, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8519341349601746, "num_tokens": 107713820.0, "step": 2835 }, { "epoch": 0.3607683500826867, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5993642807006836, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8592170476913452, "num_tokens": 107745586.0, "step": 2836 }, { "epoch": 0.3608955603612772, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.349649667739868, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8659960031509399, "num_tokens": 107787649.0, "step": 2837 }, { "epoch": 0.3610227706398677, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6146793365478516, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8495252132415771, "num_tokens": 107831545.0, "step": 2838 }, { "epoch": 0.36114998091845824, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.348315715789795, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.8388420343399048, "num_tokens": 107875252.0, "step": 2839 }, { "epoch": 0.3612771911970487, "ewc_loss": 0.045166015625, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.656994104385376, "learning_rate": 1e-05, "loss": 0.5567, "mean_token_accuracy": 0.8328679203987122, "num_tokens": 107912060.0, "step": 2840 }, { "epoch": 0.36140440147563924, "ewc_loss": 0.042236328125, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3928918838500977, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.849058210849762, "num_tokens": 107949778.0, "step": 2841 }, { "epoch": 0.36153161175422976, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.867813110351562e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.608856439590454, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8533673286437988, "num_tokens": 107991160.0, "step": 2842 }, { "epoch": 0.36165882203282024, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3756203651428223, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8412156105041504, "num_tokens": 108027059.0, "step": 2843 }, { "epoch": 0.36178603231141077, "ewc_loss": 0.04541015625, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.608975410461426, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8504648208618164, "num_tokens": 108066738.0, "step": 2844 }, { "epoch": 0.3619132425900013, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3792524337768555, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.841763973236084, "num_tokens": 108106498.0, "step": 2845 }, { "epoch": 0.36204045286859177, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6074023246765137, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8513808846473694, "num_tokens": 108139146.0, "step": 2846 }, { "epoch": 0.3621676631471823, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.35908842086792, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8669767379760742, "num_tokens": 108176070.0, "step": 2847 }, { "epoch": 0.3622948734257728, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5975115299224854, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8606668710708618, "num_tokens": 108214859.0, "step": 2848 }, { "epoch": 0.3624220837043633, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3795278072357178, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8527534008026123, "num_tokens": 108250201.0, "step": 2849 }, { "epoch": 0.3625492939829538, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5968270301818848, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8592867255210876, "num_tokens": 108282111.0, "step": 2850 }, { "epoch": 0.36267650426154435, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3829433917999268, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8387937545776367, "num_tokens": 108317873.0, "step": 2851 }, { "epoch": 0.3628037145401348, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.582834005355835, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8663365244865417, "num_tokens": 108352874.0, "step": 2852 }, { "epoch": 0.36293092481872535, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3882548809051514, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8440753817558289, "num_tokens": 108387620.0, "step": 2853 }, { "epoch": 0.3630581350973159, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5835092067718506, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.839857816696167, "num_tokens": 108428961.0, "step": 2854 }, { "epoch": 0.36318534537590635, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.915496826171875e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3683557510375977, "learning_rate": 1e-05, "loss": 0.5665, "mean_token_accuracy": 0.8315600156784058, "num_tokens": 108471505.0, "step": 2855 }, { "epoch": 0.3633125556544969, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5703961849212646, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8517836332321167, "num_tokens": 108515874.0, "step": 2856 }, { "epoch": 0.3634397659330874, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.363736391067505, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.854928731918335, "num_tokens": 108558302.0, "step": 2857 }, { "epoch": 0.3635669762116779, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5712966918945312, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8453129529953003, "num_tokens": 108597668.0, "step": 2858 }, { "epoch": 0.3636941864902684, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3572256565093994, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8516772389411926, "num_tokens": 108636623.0, "step": 2859 }, { "epoch": 0.36382139676885894, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.572134256362915, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8735635280609131, "num_tokens": 108673660.0, "step": 2860 }, { "epoch": 0.3639486070474494, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3651652336120605, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8642902374267578, "num_tokens": 108710350.0, "step": 2861 }, { "epoch": 0.36407581732603994, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.580883502960205, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8472808599472046, "num_tokens": 108747028.0, "step": 2862 }, { "epoch": 0.36420302760463047, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.378925085067749, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8425593376159668, "num_tokens": 108780315.0, "step": 2863 }, { "epoch": 0.36433023788322094, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.58772873878479, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8435735702514648, "num_tokens": 108811706.0, "step": 2864 }, { "epoch": 0.36445744816181147, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.367623805999756, "learning_rate": 1e-05, "loss": 0.5254, "mean_token_accuracy": 0.8428804278373718, "num_tokens": 108851545.0, "step": 2865 }, { "epoch": 0.364584658440402, "ewc_loss": 0.0439453125, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.568321466445923, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8664689660072327, "num_tokens": 108887013.0, "step": 2866 }, { "epoch": 0.36471186871899247, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.345841407775879, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8426628112792969, "num_tokens": 108934289.0, "step": 2867 }, { "epoch": 0.364839078997583, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5964624881744385, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.86241614818573, "num_tokens": 108973823.0, "step": 2868 }, { "epoch": 0.36496628927617353, "ewc_loss": 0.0419921875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3681399822235107, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8619144558906555, "num_tokens": 109006947.0, "step": 2869 }, { "epoch": 0.365093499554764, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6107826232910156, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8409788608551025, "num_tokens": 109039839.0, "step": 2870 }, { "epoch": 0.36522070983335453, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3673033714294434, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8625337481498718, "num_tokens": 109076348.0, "step": 2871 }, { "epoch": 0.36534792011194506, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5909006595611572, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8672347068786621, "num_tokens": 109117050.0, "step": 2872 }, { "epoch": 0.36547513039053553, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3634016513824463, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8513563871383667, "num_tokens": 109151131.0, "step": 2873 }, { "epoch": 0.36560234066912606, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5880067348480225, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.849130392074585, "num_tokens": 109194542.0, "step": 2874 }, { "epoch": 0.3657295509477166, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3777718544006348, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8512371182441711, "num_tokens": 109229636.0, "step": 2875 }, { "epoch": 0.36585676122630706, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.595445394515991, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8470116853713989, "num_tokens": 109264705.0, "step": 2876 }, { "epoch": 0.3659839715048976, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3564438819885254, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8618158102035522, "num_tokens": 109304220.0, "step": 2877 }, { "epoch": 0.3661111817834881, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.588721752166748, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8515940308570862, "num_tokens": 109347315.0, "step": 2878 }, { "epoch": 0.3662383920620786, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3693084716796875, "learning_rate": 1e-05, "loss": 0.5215, "mean_token_accuracy": 0.8413978815078735, "num_tokens": 109382679.0, "step": 2879 }, { "epoch": 0.3663656023406691, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.586341142654419, "learning_rate": 1e-05, "loss": 0.5588, "mean_token_accuracy": 0.8369885683059692, "num_tokens": 109428491.0, "step": 2880 }, { "epoch": 0.36649281261925964, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.363638401031494, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8429408073425293, "num_tokens": 109469700.0, "step": 2881 }, { "epoch": 0.3666200228978502, "ewc_loss": 0.044921875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5801703929901123, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8606386184692383, "num_tokens": 109510245.0, "step": 2882 }, { "epoch": 0.36674723317644065, "ewc_loss": 0.04248046875, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.371638774871826, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8437494039535522, "num_tokens": 109544932.0, "step": 2883 }, { "epoch": 0.3668744434550312, "ewc_loss": 0.04443359375, "ewc_loss_diag": 7.963180541992188e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5869863033294678, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8602231740951538, "num_tokens": 109579519.0, "step": 2884 }, { "epoch": 0.3670016537336217, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.369245767593384, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8610645532608032, "num_tokens": 109620313.0, "step": 2885 }, { "epoch": 0.3671288640122122, "ewc_loss": 0.044189453125, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5785536766052246, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8396292924880981, "num_tokens": 109663800.0, "step": 2886 }, { "epoch": 0.3672560742908027, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.363330841064453, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8626019358634949, "num_tokens": 109701118.0, "step": 2887 }, { "epoch": 0.36738328456939323, "ewc_loss": 0.044189453125, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5794434547424316, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.870227575302124, "num_tokens": 109739486.0, "step": 2888 }, { "epoch": 0.3675104948479837, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3613624572753906, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8663986921310425, "num_tokens": 109781097.0, "step": 2889 }, { "epoch": 0.36763770512657423, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5945816040039062, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8597812652587891, "num_tokens": 109813562.0, "step": 2890 }, { "epoch": 0.36776491540516476, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3598806858062744, "learning_rate": 1e-05, "loss": 0.5412, "mean_token_accuracy": 0.8346158266067505, "num_tokens": 109855326.0, "step": 2891 }, { "epoch": 0.36789212568375523, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5996875762939453, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8508591055870056, "num_tokens": 109890787.0, "step": 2892 }, { "epoch": 0.36801933596234576, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.350857973098755, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8508822917938232, "num_tokens": 109932199.0, "step": 2893 }, { "epoch": 0.3681465462409363, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6064236164093018, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8574745655059814, "num_tokens": 109967388.0, "step": 2894 }, { "epoch": 0.36827375651952676, "ewc_loss": 0.042236328125, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3476686477661133, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8649910688400269, "num_tokens": 110003437.0, "step": 2895 }, { "epoch": 0.3684009667981173, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6218652725219727, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8495635986328125, "num_tokens": 110040822.0, "step": 2896 }, { "epoch": 0.3685281770767078, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.354220390319824, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8584800362586975, "num_tokens": 110079628.0, "step": 2897 }, { "epoch": 0.3686553873552983, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.619194746017456, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8629448413848877, "num_tokens": 110115794.0, "step": 2898 }, { "epoch": 0.3687825976338888, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3690693378448486, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8484293222427368, "num_tokens": 110153227.0, "step": 2899 }, { "epoch": 0.36890980791247935, "ewc_loss": 0.045654296875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6198856830596924, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.8358039259910583, "num_tokens": 110193186.0, "step": 2900 }, { "epoch": 0.3690370181910698, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.368225574493408, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8443083167076111, "num_tokens": 110232130.0, "step": 2901 }, { "epoch": 0.36916422846966035, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6098275184631348, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8657056093215942, "num_tokens": 110271376.0, "step": 2902 }, { "epoch": 0.3692914387482509, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3786864280700684, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8478127717971802, "num_tokens": 110314220.0, "step": 2903 }, { "epoch": 0.36941864902684135, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.602837324142456, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8485586643218994, "num_tokens": 110357292.0, "step": 2904 }, { "epoch": 0.3695458593054319, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3774595260620117, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.8358107209205627, "num_tokens": 110397813.0, "step": 2905 }, { "epoch": 0.3696730695840224, "ewc_loss": 0.044677734375, "ewc_loss_diag": 8.0108642578125e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6492738723754883, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8642598986625671, "num_tokens": 110435773.0, "step": 2906 }, { "epoch": 0.3698002798626129, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.368046522140503, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8769351840019226, "num_tokens": 110469518.0, "step": 2907 }, { "epoch": 0.3699274901412034, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6365463733673096, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8513827323913574, "num_tokens": 110503156.0, "step": 2908 }, { "epoch": 0.37005470041979394, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.372591972351074, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8441236615180969, "num_tokens": 110543960.0, "step": 2909 }, { "epoch": 0.3701819106983844, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.62493634223938, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8416059017181396, "num_tokens": 110584811.0, "step": 2910 }, { "epoch": 0.37030912097697494, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.383478879928589, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.8439209461212158, "num_tokens": 110622433.0, "step": 2911 }, { "epoch": 0.37043633125556547, "ewc_loss": 0.045654296875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6111326217651367, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8482288122177124, "num_tokens": 110660994.0, "step": 2912 }, { "epoch": 0.37056354153415594, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4192352294921875, "learning_rate": 1e-05, "loss": 0.5673, "mean_token_accuracy": 0.8316991925239563, "num_tokens": 110703022.0, "step": 2913 }, { "epoch": 0.37069075181274647, "ewc_loss": 0.044189453125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.564642906188965, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8649140000343323, "num_tokens": 110743864.0, "step": 2914 }, { "epoch": 0.370817962091337, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.4005494117736816, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8483849763870239, "num_tokens": 110778591.0, "step": 2915 }, { "epoch": 0.37094517236992747, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.533047914505005, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8615995645523071, "num_tokens": 110818402.0, "step": 2916 }, { "epoch": 0.371072382648518, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3989551067352295, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8543838262557983, "num_tokens": 110855129.0, "step": 2917 }, { "epoch": 0.3711995929271085, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.518177032470703, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8560788631439209, "num_tokens": 110896307.0, "step": 2918 }, { "epoch": 0.371326803205699, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.383669853210449, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8571904301643372, "num_tokens": 110940639.0, "step": 2919 }, { "epoch": 0.3714540134842895, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.521282434463501, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8539755344390869, "num_tokens": 110979134.0, "step": 2920 }, { "epoch": 0.37158122376288005, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.384932518005371, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8655641078948975, "num_tokens": 111013799.0, "step": 2921 }, { "epoch": 0.3717084340414705, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.513519048690796, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.853118896484375, "num_tokens": 111055486.0, "step": 2922 }, { "epoch": 0.37183564432006105, "ewc_loss": 0.042236328125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3786182403564453, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8561559319496155, "num_tokens": 111101332.0, "step": 2923 }, { "epoch": 0.3719628545986516, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.519037961959839, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.855748176574707, "num_tokens": 111137285.0, "step": 2924 }, { "epoch": 0.37209006487724205, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.370030641555786, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8460497260093689, "num_tokens": 111177254.0, "step": 2925 }, { "epoch": 0.3722172751558326, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.058547973632812e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5148332118988037, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8534351587295532, "num_tokens": 111217633.0, "step": 2926 }, { "epoch": 0.3723444854344231, "ewc_loss": 0.0419921875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.369784116744995, "learning_rate": 1e-05, "loss": 0.5777, "mean_token_accuracy": 0.8232579827308655, "num_tokens": 111259273.0, "step": 2927 }, { "epoch": 0.3724716957130136, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5292418003082275, "learning_rate": 1e-05, "loss": 0.5251, "mean_token_accuracy": 0.8383791446685791, "num_tokens": 111295068.0, "step": 2928 }, { "epoch": 0.3725989059916041, "ewc_loss": 0.0419921875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.355809211730957, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8409690856933594, "num_tokens": 111336863.0, "step": 2929 }, { "epoch": 0.37272611627019464, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5367937088012695, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.8399286270141602, "num_tokens": 111378432.0, "step": 2930 }, { "epoch": 0.37285332654878517, "ewc_loss": 0.0419921875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.3548176288604736, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.846946120262146, "num_tokens": 111412481.0, "step": 2931 }, { "epoch": 0.37298053682737564, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5516469478607178, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8365986347198486, "num_tokens": 111450534.0, "step": 2932 }, { "epoch": 0.37310774710596617, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3458118438720703, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.86139315366745, "num_tokens": 111484278.0, "step": 2933 }, { "epoch": 0.3732349573845567, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5546534061431885, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.847741961479187, "num_tokens": 111526797.0, "step": 2934 }, { "epoch": 0.37336216766314717, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.153915405273438e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.334449052810669, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8418369889259338, "num_tokens": 111568139.0, "step": 2935 }, { "epoch": 0.3734893779417377, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5808844566345215, "learning_rate": 1e-05, "loss": 0.556, "mean_token_accuracy": 0.8287699818611145, "num_tokens": 111614519.0, "step": 2936 }, { "epoch": 0.3736165882203282, "ewc_loss": 0.0419921875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3460781574249268, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8585817217826843, "num_tokens": 111647965.0, "step": 2937 }, { "epoch": 0.3737437984989187, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5941483974456787, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8522976636886597, "num_tokens": 111682492.0, "step": 2938 }, { "epoch": 0.37387100877750923, "ewc_loss": 0.0419921875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.373183012008667, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8436999320983887, "num_tokens": 111717550.0, "step": 2939 }, { "epoch": 0.37399821905609976, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5805587768554688, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8631782531738281, "num_tokens": 111752173.0, "step": 2940 }, { "epoch": 0.37412542933469023, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3507604598999023, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8523909449577332, "num_tokens": 111790845.0, "step": 2941 }, { "epoch": 0.37425263961328076, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5875906944274902, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8496394157409668, "num_tokens": 111827837.0, "step": 2942 }, { "epoch": 0.3743798498918713, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3537657260894775, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8430789709091187, "num_tokens": 111868785.0, "step": 2943 }, { "epoch": 0.37450706017046176, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5958330631256104, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8489711284637451, "num_tokens": 111907296.0, "step": 2944 }, { "epoch": 0.3746342704490523, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.354649782180786, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8668887615203857, "num_tokens": 111947619.0, "step": 2945 }, { "epoch": 0.3747614807276428, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.615623712539673, "learning_rate": 1e-05, "loss": 0.5439, "mean_token_accuracy": 0.8351684808731079, "num_tokens": 111987255.0, "step": 2946 }, { "epoch": 0.3748886910062333, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3667171001434326, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.849827766418457, "num_tokens": 112023003.0, "step": 2947 }, { "epoch": 0.3750159012848238, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.616335153579712, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8438957333564758, "num_tokens": 112056292.0, "step": 2948 }, { "epoch": 0.37514311156341434, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.371035099029541, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8499938249588013, "num_tokens": 112089947.0, "step": 2949 }, { "epoch": 0.3752703218420048, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.62929630279541, "learning_rate": 1e-05, "loss": 0.5717, "mean_token_accuracy": 0.8293587565422058, "num_tokens": 112125415.0, "step": 2950 }, { "epoch": 0.37539753212059535, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.36069655418396, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.846805214881897, "num_tokens": 112164364.0, "step": 2951 }, { "epoch": 0.3755247423991859, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6089322566986084, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.859906792640686, "num_tokens": 112208994.0, "step": 2952 }, { "epoch": 0.37565195267777635, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3582892417907715, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8547133803367615, "num_tokens": 112249902.0, "step": 2953 }, { "epoch": 0.3757791629563669, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.612787961959839, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8612697124481201, "num_tokens": 112290446.0, "step": 2954 }, { "epoch": 0.3759063732349574, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.387847661972046, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.850167989730835, "num_tokens": 112323217.0, "step": 2955 }, { "epoch": 0.3760335835135479, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.590392827987671, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8551337718963623, "num_tokens": 112365863.0, "step": 2956 }, { "epoch": 0.3761607937921384, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3854997158050537, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8522749543190002, "num_tokens": 112400708.0, "step": 2957 }, { "epoch": 0.37628800407072893, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5794949531555176, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.865668535232544, "num_tokens": 112440063.0, "step": 2958 }, { "epoch": 0.3764152143493194, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3635380268096924, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8782511949539185, "num_tokens": 112478856.0, "step": 2959 }, { "epoch": 0.37654242462790993, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5950231552124023, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8528308868408203, "num_tokens": 112515707.0, "step": 2960 }, { "epoch": 0.37666963490650046, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.38014817237854, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.845854640007019, "num_tokens": 112551302.0, "step": 2961 }, { "epoch": 0.37679684518509093, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5838756561279297, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8648302555084229, "num_tokens": 112585821.0, "step": 2962 }, { "epoch": 0.37692405546368146, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.374196767807007, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8543482422828674, "num_tokens": 112624261.0, "step": 2963 }, { "epoch": 0.377051265742272, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.583533763885498, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8686026334762573, "num_tokens": 112656944.0, "step": 2964 }, { "epoch": 0.37717847602086246, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3817195892333984, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8608072996139526, "num_tokens": 112694386.0, "step": 2965 }, { "epoch": 0.377305686299453, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.575725793838501, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8545201420783997, "num_tokens": 112731597.0, "step": 2966 }, { "epoch": 0.3774328965780435, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.20159912109375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.368302345275879, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8787540197372437, "num_tokens": 112768012.0, "step": 2967 }, { "epoch": 0.377560106856634, "ewc_loss": 0.044677734375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.6007239818573, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.8405231833457947, "num_tokens": 112798089.0, "step": 2968 }, { "epoch": 0.3776873171352245, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.356975555419922, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8572180271148682, "num_tokens": 112842332.0, "step": 2969 }, { "epoch": 0.37781452741381505, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5987625122070312, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8475549221038818, "num_tokens": 112882382.0, "step": 2970 }, { "epoch": 0.3779417376924055, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3650200366973877, "learning_rate": 1e-05, "loss": 0.5553, "mean_token_accuracy": 0.8311638832092285, "num_tokens": 112922627.0, "step": 2971 }, { "epoch": 0.37806894797099605, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.624737024307251, "learning_rate": 1e-05, "loss": 0.5489, "mean_token_accuracy": 0.8314311504364014, "num_tokens": 112957571.0, "step": 2972 }, { "epoch": 0.3781961582495866, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3456928730010986, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8518381714820862, "num_tokens": 113001128.0, "step": 2973 }, { "epoch": 0.37832336852817705, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6320738792419434, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8589181900024414, "num_tokens": 113034515.0, "step": 2974 }, { "epoch": 0.3784505788067676, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.360935926437378, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8690147399902344, "num_tokens": 113071221.0, "step": 2975 }, { "epoch": 0.3785777890853581, "ewc_loss": 0.046142578125, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6492276191711426, "learning_rate": 1e-05, "loss": 0.5556, "mean_token_accuracy": 0.8348567485809326, "num_tokens": 113105804.0, "step": 2976 }, { "epoch": 0.3787049993639486, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3588685989379883, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.86069256067276, "num_tokens": 113146051.0, "step": 2977 }, { "epoch": 0.3788322096425391, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.645752429962158, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8672712445259094, "num_tokens": 113186696.0, "step": 2978 }, { "epoch": 0.37895941992112964, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3719003200531006, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8454710245132446, "num_tokens": 113228402.0, "step": 2979 }, { "epoch": 0.37908663019972016, "ewc_loss": 0.046630859375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.639890432357788, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8427519202232361, "num_tokens": 113270407.0, "step": 2980 }, { "epoch": 0.37921384047831064, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.396052837371826, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8460304737091064, "num_tokens": 113304523.0, "step": 2981 }, { "epoch": 0.37934105075690117, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6122586727142334, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8555685877799988, "num_tokens": 113345224.0, "step": 2982 }, { "epoch": 0.3794682610354917, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.395770311355591, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8635495901107788, "num_tokens": 113382287.0, "step": 2983 }, { "epoch": 0.37959547131408217, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6105294227600098, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8415663242340088, "num_tokens": 113427308.0, "step": 2984 }, { "epoch": 0.3797226815926727, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.249282836914062e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4022653102874756, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8512987494468689, "num_tokens": 113461272.0, "step": 2985 }, { "epoch": 0.3798498918712632, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5652055740356445, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8721979260444641, "num_tokens": 113499942.0, "step": 2986 }, { "epoch": 0.3799771021498537, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.39859676361084, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8706796169281006, "num_tokens": 113531801.0, "step": 2987 }, { "epoch": 0.3801043124284442, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.54887056350708, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8695964813232422, "num_tokens": 113567475.0, "step": 2988 }, { "epoch": 0.38023152270703475, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4045042991638184, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8412982821464539, "num_tokens": 113603767.0, "step": 2989 }, { "epoch": 0.3803587329856252, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.5448715686798096, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8486596345901489, "num_tokens": 113641213.0, "step": 2990 }, { "epoch": 0.38048594326421575, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3900346755981445, "learning_rate": 1e-05, "loss": 0.5618, "mean_token_accuracy": 0.8287246823310852, "num_tokens": 113683364.0, "step": 2991 }, { "epoch": 0.3806131535428063, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.542848825454712, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8504989147186279, "num_tokens": 113718488.0, "step": 2992 }, { "epoch": 0.38074036382139675, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.375027656555176, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8667393922805786, "num_tokens": 113756664.0, "step": 2993 }, { "epoch": 0.3808675740999873, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.53961443901062, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8593149781227112, "num_tokens": 113790711.0, "step": 2994 }, { "epoch": 0.3809947843785778, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3642258644104004, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8635043501853943, "num_tokens": 113827091.0, "step": 2995 }, { "epoch": 0.3811219946571683, "ewc_loss": 0.044189453125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.572722911834717, "learning_rate": 1e-05, "loss": 0.5251, "mean_token_accuracy": 0.8402616381645203, "num_tokens": 113862971.0, "step": 2996 }, { "epoch": 0.3812492049357588, "ewc_loss": 0.042236328125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.3855438232421875e-05, "grad_norm": 2.3663883209228516, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8481655716896057, "num_tokens": 113897430.0, "step": 2997 }, { "epoch": 0.38137641521434934, "ewc_loss": 0.044189453125, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.583165168762207, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8513840436935425, "num_tokens": 113926626.0, "step": 2998 }, { "epoch": 0.3815036254929398, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.340986728668213, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8617271184921265, "num_tokens": 113969656.0, "step": 2999 }, { "epoch": 0.38163083577153034, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5877623558044434, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.867787778377533, "num_tokens": 114006140.0, "step": 3000 }, { "epoch": 0.38175804605012087, "ewc_loss": 0.04248046875, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3412764072418213, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8587583899497986, "num_tokens": 114047300.0, "step": 3001 }, { "epoch": 0.38188525632871134, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6151540279388428, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.848584771156311, "num_tokens": 114086129.0, "step": 3002 }, { "epoch": 0.38201246660730187, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.296966552734375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.344017744064331, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8535327911376953, "num_tokens": 114123345.0, "step": 3003 }, { "epoch": 0.3821396768858924, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6332128047943115, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8513064384460449, "num_tokens": 114158983.0, "step": 3004 }, { "epoch": 0.38226688716448287, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3721160888671875, "learning_rate": 1e-05, "loss": 0.5517, "mean_token_accuracy": 0.83014976978302, "num_tokens": 114199055.0, "step": 3005 }, { "epoch": 0.3823940974430734, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.344650268554688e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6224822998046875, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8589318990707397, "num_tokens": 114236355.0, "step": 3006 }, { "epoch": 0.38252130772166393, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.350973129272461, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8471273183822632, "num_tokens": 114285111.0, "step": 3007 }, { "epoch": 0.3826485180002544, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6499032974243164, "learning_rate": 1e-05, "loss": 0.5461, "mean_token_accuracy": 0.8306602835655212, "num_tokens": 114319306.0, "step": 3008 }, { "epoch": 0.38277572827884493, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3609237670898438, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8658204078674316, "num_tokens": 114354117.0, "step": 3009 }, { "epoch": 0.38290293855743546, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.639186382293701, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8429906368255615, "num_tokens": 114395865.0, "step": 3010 }, { "epoch": 0.38303014883602593, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.383704423904419, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8595567345619202, "num_tokens": 114427862.0, "step": 3011 }, { "epoch": 0.38315735911461646, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6288681030273438, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.863304615020752, "num_tokens": 114462892.0, "step": 3012 }, { "epoch": 0.383284569393207, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.375523328781128, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8648616075515747, "num_tokens": 114500477.0, "step": 3013 }, { "epoch": 0.38341177967179746, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6250674724578857, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8514479398727417, "num_tokens": 114534695.0, "step": 3014 }, { "epoch": 0.383538989950388, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.380650281906128, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8373353481292725, "num_tokens": 114579969.0, "step": 3015 }, { "epoch": 0.3836662002289785, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6350862979888916, "learning_rate": 1e-05, "loss": 0.5474, "mean_token_accuracy": 0.8360649347305298, "num_tokens": 114614904.0, "step": 3016 }, { "epoch": 0.383793410507569, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4507193565368652, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.863716721534729, "num_tokens": 114647126.0, "step": 3017 }, { "epoch": 0.3839206207861595, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5712549686431885, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8522779941558838, "num_tokens": 114683529.0, "step": 3018 }, { "epoch": 0.38404783106475004, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.393421173095703, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8607318997383118, "num_tokens": 114722471.0, "step": 3019 }, { "epoch": 0.3841750413433405, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5510125160217285, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8477637767791748, "num_tokens": 114765243.0, "step": 3020 }, { "epoch": 0.38430225162193105, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3853061199188232, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8556431531906128, "num_tokens": 114806341.0, "step": 3021 }, { "epoch": 0.3844294619005216, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5554182529449463, "learning_rate": 1e-05, "loss": 0.5814, "mean_token_accuracy": 0.8286047577857971, "num_tokens": 114850014.0, "step": 3022 }, { "epoch": 0.38455667217911205, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.4019861221313477, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8479997515678406, "num_tokens": 114884771.0, "step": 3023 }, { "epoch": 0.3846838824577026, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5232417583465576, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8598134517669678, "num_tokens": 114926319.0, "step": 3024 }, { "epoch": 0.3848110927362931, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.386348247528076, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8452650308609009, "num_tokens": 114971008.0, "step": 3025 }, { "epoch": 0.3849383030148836, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.545555830001831, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8497552871704102, "num_tokens": 115007294.0, "step": 3026 }, { "epoch": 0.3850655132934741, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.392333984375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3734662532806396, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8570626378059387, "num_tokens": 115043066.0, "step": 3027 }, { "epoch": 0.38519272357206463, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.544431686401367, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8579992055892944, "num_tokens": 115077404.0, "step": 3028 }, { "epoch": 0.3853199338506551, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.440017700195312e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3707921504974365, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.864352822303772, "num_tokens": 115114092.0, "step": 3029 }, { "epoch": 0.38544714412924563, "ewc_loss": 0.044677734375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5314154624938965, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.865773618221283, "num_tokens": 115157095.0, "step": 3030 }, { "epoch": 0.38557435440783616, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.377443313598633, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8676055669784546, "num_tokens": 115189752.0, "step": 3031 }, { "epoch": 0.3857015646864267, "ewc_loss": 0.04443359375, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5385947227478027, "learning_rate": 1e-05, "loss": 0.5582, "mean_token_accuracy": 0.8327720165252686, "num_tokens": 115232650.0, "step": 3032 }, { "epoch": 0.38582877496501716, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3713982105255127, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8673508167266846, "num_tokens": 115267812.0, "step": 3033 }, { "epoch": 0.3859559852436077, "ewc_loss": 0.044677734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5507094860076904, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8601184487342834, "num_tokens": 115300905.0, "step": 3034 }, { "epoch": 0.3860831955221982, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3536295890808105, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8637958765029907, "num_tokens": 115341869.0, "step": 3035 }, { "epoch": 0.3862104058007887, "ewc_loss": 0.044677734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.559372901916504, "learning_rate": 1e-05, "loss": 0.5571, "mean_token_accuracy": 0.8324781656265259, "num_tokens": 115380499.0, "step": 3036 }, { "epoch": 0.3863376160793792, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.356457233428955, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8626202344894409, "num_tokens": 115413573.0, "step": 3037 }, { "epoch": 0.38646482635796975, "ewc_loss": 0.044677734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.562323808670044, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8629036545753479, "num_tokens": 115452634.0, "step": 3038 }, { "epoch": 0.3865920366365602, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3688223361968994, "learning_rate": 1e-05, "loss": 0.5721, "mean_token_accuracy": 0.8300373554229736, "num_tokens": 115492509.0, "step": 3039 }, { "epoch": 0.38671924691515075, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.487701416015625e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.561541795730591, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8651782274246216, "num_tokens": 115531256.0, "step": 3040 }, { "epoch": 0.3868464571937413, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.366741180419922, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8457769155502319, "num_tokens": 115564173.0, "step": 3041 }, { "epoch": 0.38697366747233175, "ewc_loss": 0.044677734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5675339698791504, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8589473366737366, "num_tokens": 115598359.0, "step": 3042 }, { "epoch": 0.3871008777509223, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.355862855911255, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8510595560073853, "num_tokens": 115637782.0, "step": 3043 }, { "epoch": 0.3872280880295128, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5728514194488525, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8398349285125732, "num_tokens": 115680809.0, "step": 3044 }, { "epoch": 0.3873552983081033, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.383486032485962, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8508493304252625, "num_tokens": 115712092.0, "step": 3045 }, { "epoch": 0.3874825085866938, "ewc_loss": 0.044677734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5619630813598633, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8515917062759399, "num_tokens": 115749400.0, "step": 3046 }, { "epoch": 0.38760971886528434, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.374612331390381, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8483268022537231, "num_tokens": 115783932.0, "step": 3047 }, { "epoch": 0.3877369291438748, "ewc_loss": 0.044677734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5727169513702393, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8535531759262085, "num_tokens": 115819358.0, "step": 3048 }, { "epoch": 0.38786413942246534, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3518354892730713, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8649535775184631, "num_tokens": 115855370.0, "step": 3049 }, { "epoch": 0.38799134970105587, "ewc_loss": 0.044677734375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.584847927093506, "learning_rate": 1e-05, "loss": 0.5388, "mean_token_accuracy": 0.8364684581756592, "num_tokens": 115894722.0, "step": 3050 }, { "epoch": 0.38811855997964634, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.350348711013794, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8594306707382202, "num_tokens": 115935276.0, "step": 3051 }, { "epoch": 0.38824577025823687, "ewc_loss": 0.045654296875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6017305850982666, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8522869944572449, "num_tokens": 115971255.0, "step": 3052 }, { "epoch": 0.3883729805368274, "ewc_loss": 0.042724609375, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3556013107299805, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8646366000175476, "num_tokens": 116003818.0, "step": 3053 }, { "epoch": 0.38850019081541787, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6110341548919678, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8579246997833252, "num_tokens": 116040720.0, "step": 3054 }, { "epoch": 0.3886274010940084, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.535385131835938e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3430607318878174, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8669792413711548, "num_tokens": 116076485.0, "step": 3055 }, { "epoch": 0.3887546113725989, "ewc_loss": 0.046142578125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6249570846557617, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8668325543403625, "num_tokens": 116109802.0, "step": 3056 }, { "epoch": 0.3888818216511894, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3538475036621094, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8537024259567261, "num_tokens": 116149865.0, "step": 3057 }, { "epoch": 0.3890090319297799, "ewc_loss": 0.046630859375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.64795184135437, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8560954332351685, "num_tokens": 116185887.0, "step": 3058 }, { "epoch": 0.38913624220837045, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.350762367248535, "learning_rate": 1e-05, "loss": 0.5236, "mean_token_accuracy": 0.8464639186859131, "num_tokens": 116233090.0, "step": 3059 }, { "epoch": 0.3892634524869609, "ewc_loss": 0.04736328125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.653691053390503, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8510363101959229, "num_tokens": 116274433.0, "step": 3060 }, { "epoch": 0.38939066276555145, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.387791633605957, "learning_rate": 1e-05, "loss": 0.5566, "mean_token_accuracy": 0.8355410099029541, "num_tokens": 116309104.0, "step": 3061 }, { "epoch": 0.389517873044142, "ewc_loss": 0.046630859375, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.638461112976074, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8601801991462708, "num_tokens": 116350248.0, "step": 3062 }, { "epoch": 0.38964508332273246, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.384467124938965, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8421264290809631, "num_tokens": 116391078.0, "step": 3063 }, { "epoch": 0.389772293601323, "ewc_loss": 0.045654296875, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.625398635864258, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8502511978149414, "num_tokens": 116430652.0, "step": 3064 }, { "epoch": 0.3898995038799135, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3911678791046143, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8654416799545288, "num_tokens": 116467024.0, "step": 3065 }, { "epoch": 0.390026714158504, "ewc_loss": 0.045654296875, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6125683784484863, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8620348572731018, "num_tokens": 116504304.0, "step": 3066 }, { "epoch": 0.3901539244370945, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3962390422821045, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8487085700035095, "num_tokens": 116541079.0, "step": 3067 }, { "epoch": 0.39028113471568504, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6077115535736084, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8457618951797485, "num_tokens": 116578690.0, "step": 3068 }, { "epoch": 0.3904083449942755, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3898379802703857, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8526498079299927, "num_tokens": 116614945.0, "step": 3069 }, { "epoch": 0.39053555527286604, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6012649536132812, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8680382370948792, "num_tokens": 116646795.0, "step": 3070 }, { "epoch": 0.39066276555145657, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3757970333099365, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8469295501708984, "num_tokens": 116685731.0, "step": 3071 }, { "epoch": 0.39078997583004704, "ewc_loss": 0.045654296875, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.593658924102783, "learning_rate": 1e-05, "loss": 0.4044, "mean_token_accuracy": 0.8742903470993042, "num_tokens": 116722559.0, "step": 3072 }, { "epoch": 0.39091718610863757, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.385756015777588, "learning_rate": 1e-05, "loss": 0.5398, "mean_token_accuracy": 0.8379786014556885, "num_tokens": 116762954.0, "step": 3073 }, { "epoch": 0.3910443963872281, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.587679862976074, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.864309549331665, "num_tokens": 116801745.0, "step": 3074 }, { "epoch": 0.39117160666581857, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.372652769088745, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8673746585845947, "num_tokens": 116837650.0, "step": 3075 }, { "epoch": 0.3912988169444091, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5979297161102295, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8437781929969788, "num_tokens": 116872722.0, "step": 3076 }, { "epoch": 0.39142602722299963, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.374418258666992, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8572587370872498, "num_tokens": 116911175.0, "step": 3077 }, { "epoch": 0.3915532375015901, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.595773458480835, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.852194607257843, "num_tokens": 116946317.0, "step": 3078 }, { "epoch": 0.39168044778018063, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3713459968566895, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.8367413282394409, "num_tokens": 116988661.0, "step": 3079 }, { "epoch": 0.39180765805877116, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.578084945678711, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.860378086566925, "num_tokens": 117027309.0, "step": 3080 }, { "epoch": 0.3919348683373617, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3644115924835205, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8575697541236877, "num_tokens": 117065608.0, "step": 3081 }, { "epoch": 0.39206207861595216, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6279408931732178, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8501298427581787, "num_tokens": 117101824.0, "step": 3082 }, { "epoch": 0.3921892888945427, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.58306884765625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3554482460021973, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8488162755966187, "num_tokens": 117141985.0, "step": 3083 }, { "epoch": 0.3923164991731332, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6234405040740967, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8607996702194214, "num_tokens": 117177606.0, "step": 3084 }, { "epoch": 0.3924437094517237, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.352631092071533, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8536860942840576, "num_tokens": 117220214.0, "step": 3085 }, { "epoch": 0.3925709197303142, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6361067295074463, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.849600076675415, "num_tokens": 117259188.0, "step": 3086 }, { "epoch": 0.39269813000890474, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.630752563476562e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.363600492477417, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8518198132514954, "num_tokens": 117299146.0, "step": 3087 }, { "epoch": 0.3928253402874952, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.635362148284912, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8516260385513306, "num_tokens": 117337560.0, "step": 3088 }, { "epoch": 0.39295255056608575, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.375261068344116, "learning_rate": 1e-05, "loss": 0.5321, "mean_token_accuracy": 0.8406379818916321, "num_tokens": 117376935.0, "step": 3089 }, { "epoch": 0.3930797608446763, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6204895973205566, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8551282286643982, "num_tokens": 117420273.0, "step": 3090 }, { "epoch": 0.39320697112326675, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3803908824920654, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8674361109733582, "num_tokens": 117457338.0, "step": 3091 }, { "epoch": 0.3933341814018573, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6120896339416504, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8645027279853821, "num_tokens": 117491082.0, "step": 3092 }, { "epoch": 0.3934613916804478, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3861966133117676, "learning_rate": 1e-05, "loss": 0.3825, "mean_token_accuracy": 0.8834800720214844, "num_tokens": 117520425.0, "step": 3093 }, { "epoch": 0.3935886019590383, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.592316150665283, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8652228116989136, "num_tokens": 117559021.0, "step": 3094 }, { "epoch": 0.3937158122376288, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3878467082977295, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8636912703514099, "num_tokens": 117598705.0, "step": 3095 }, { "epoch": 0.39384302251621933, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.585663080215454, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8659764528274536, "num_tokens": 117636185.0, "step": 3096 }, { "epoch": 0.3939702327948098, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.382664442062378, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8661750555038452, "num_tokens": 117674290.0, "step": 3097 }, { "epoch": 0.39409744307340033, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.584176778793335, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8588972091674805, "num_tokens": 117708708.0, "step": 3098 }, { "epoch": 0.39422465335199086, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.402224540710449, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8543674349784851, "num_tokens": 117742528.0, "step": 3099 }, { "epoch": 0.39435186363058133, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5732710361480713, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.843232274055481, "num_tokens": 117779050.0, "step": 3100 }, { "epoch": 0.39447907390917186, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.378476858139038, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8432838916778564, "num_tokens": 117824004.0, "step": 3101 }, { "epoch": 0.3946062841877624, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5542750358581543, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8660439252853394, "num_tokens": 117864837.0, "step": 3102 }, { "epoch": 0.39473349446635286, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.378291368484497, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8591541051864624, "num_tokens": 117903675.0, "step": 3103 }, { "epoch": 0.3948607047449434, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.566056251525879, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8504019975662231, "num_tokens": 117942561.0, "step": 3104 }, { "epoch": 0.3949879150235339, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3751111030578613, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8495451807975769, "num_tokens": 117982439.0, "step": 3105 }, { "epoch": 0.3951151253021244, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5580356121063232, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8511310815811157, "num_tokens": 118025153.0, "step": 3106 }, { "epoch": 0.3952423355807149, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.366265058517456, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8495123386383057, "num_tokens": 118064508.0, "step": 3107 }, { "epoch": 0.39536954585930545, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.590360641479492, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8603305220603943, "num_tokens": 118093368.0, "step": 3108 }, { "epoch": 0.3954967561378959, "ewc_loss": 0.04296875, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3618390560150146, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8675574064254761, "num_tokens": 118129164.0, "step": 3109 }, { "epoch": 0.39562396641648645, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.678436279296875e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.582689046859741, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8621851205825806, "num_tokens": 118169092.0, "step": 3110 }, { "epoch": 0.395751176695077, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.351633310317993, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8570401072502136, "num_tokens": 118211616.0, "step": 3111 }, { "epoch": 0.39587838697366745, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.608211040496826, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8489207029342651, "num_tokens": 118248483.0, "step": 3112 }, { "epoch": 0.396005597252258, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3518946170806885, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8557226657867432, "num_tokens": 118288540.0, "step": 3113 }, { "epoch": 0.3961328075308485, "ewc_loss": 0.045654296875, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6160647869110107, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8559439182281494, "num_tokens": 118326988.0, "step": 3114 }, { "epoch": 0.396260017809439, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.726119995117188e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3509278297424316, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8558965921401978, "num_tokens": 118366100.0, "step": 3115 }, { "epoch": 0.3963872280880295, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.641024112701416, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8533560633659363, "num_tokens": 118397243.0, "step": 3116 }, { "epoch": 0.39651443836662004, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3709137439727783, "learning_rate": 1e-05, "loss": 0.5404, "mean_token_accuracy": 0.8387634754180908, "num_tokens": 118438198.0, "step": 3117 }, { "epoch": 0.3966416486452105, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6416501998901367, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8521707653999329, "num_tokens": 118473496.0, "step": 3118 }, { "epoch": 0.39676885892380104, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.391618013381958, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8649196624755859, "num_tokens": 118511988.0, "step": 3119 }, { "epoch": 0.39689606920239157, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.610509157180786, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8640064597129822, "num_tokens": 118548287.0, "step": 3120 }, { "epoch": 0.39702327948098204, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3891444206237793, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8633952140808105, "num_tokens": 118580637.0, "step": 3121 }, { "epoch": 0.39715048975957257, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5999550819396973, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8464062213897705, "num_tokens": 118622316.0, "step": 3122 }, { "epoch": 0.3972777000381631, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.389251947402954, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8430098295211792, "num_tokens": 118660497.0, "step": 3123 }, { "epoch": 0.39740491031675357, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5977728366851807, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8494514226913452, "num_tokens": 118691733.0, "step": 3124 }, { "epoch": 0.3975321205953441, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.390758752822876, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8432998657226562, "num_tokens": 118728576.0, "step": 3125 }, { "epoch": 0.3976593308739346, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.583390951156616, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8439363241195679, "num_tokens": 118765394.0, "step": 3126 }, { "epoch": 0.3977865411525251, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.7738037109375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3863813877105713, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8425518274307251, "num_tokens": 118804856.0, "step": 3127 }, { "epoch": 0.3979137514311156, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5848190784454346, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8544629812240601, "num_tokens": 118839274.0, "step": 3128 }, { "epoch": 0.39804096170970615, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.37756609916687, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8443086743354797, "num_tokens": 118877866.0, "step": 3129 }, { "epoch": 0.3981681719882967, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.584486484527588, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.868977427482605, "num_tokens": 118908942.0, "step": 3130 }, { "epoch": 0.39829538226688715, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3791160583496094, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8398314714431763, "num_tokens": 118948957.0, "step": 3131 }, { "epoch": 0.3984225925454777, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.569925308227539, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8610415458679199, "num_tokens": 118991246.0, "step": 3132 }, { "epoch": 0.3985498028240682, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3990955352783203, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8395471572875977, "num_tokens": 119025661.0, "step": 3133 }, { "epoch": 0.3986770131026587, "ewc_loss": 0.044677734375, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.569035768508911, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8511130809783936, "num_tokens": 119059524.0, "step": 3134 }, { "epoch": 0.3988042233812492, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.37441086769104, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8448680639266968, "num_tokens": 119101636.0, "step": 3135 }, { "epoch": 0.39893143365983974, "ewc_loss": 0.044921875, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.567103385925293, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8480707406997681, "num_tokens": 119138448.0, "step": 3136 }, { "epoch": 0.3990586439384302, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3660924434661865, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.842849612236023, "num_tokens": 119183337.0, "step": 3137 }, { "epoch": 0.39918585421702074, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5731894969940186, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8795543313026428, "num_tokens": 119217503.0, "step": 3138 }, { "epoch": 0.39931306449561127, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.378336191177368, "learning_rate": 1e-05, "loss": 0.5354, "mean_token_accuracy": 0.8406272530555725, "num_tokens": 119255824.0, "step": 3139 }, { "epoch": 0.39944027477420174, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5807876586914062, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.850261926651001, "num_tokens": 119292677.0, "step": 3140 }, { "epoch": 0.39956748505279227, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3652877807617188, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8570576906204224, "num_tokens": 119330961.0, "step": 3141 }, { "epoch": 0.3996946953313828, "ewc_loss": 0.045166015625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.6021065711975098, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.843766450881958, "num_tokens": 119362122.0, "step": 3142 }, { "epoch": 0.39982190560997327, "ewc_loss": 0.043212890625, "ewc_loss_diag": 8.869171142578125e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3532886505126953, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8574249744415283, "num_tokens": 119398018.0, "step": 3143 }, { "epoch": 0.3999491158885638, "ewc_loss": 0.04541015625, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6050913333892822, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8461679816246033, "num_tokens": 119438552.0, "step": 3144 }, { "epoch": 0.40007632616715433, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.916854858398438e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.370866537094116, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8519878387451172, "num_tokens": 119470777.0, "step": 3145 }, { "epoch": 0.4002035364457448, "ewc_loss": 0.045654296875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.607511281967163, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8680086731910706, "num_tokens": 119505062.0, "step": 3146 }, { "epoch": 0.40033074672433533, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3767173290252686, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8490535020828247, "num_tokens": 119540394.0, "step": 3147 }, { "epoch": 0.40045795700292586, "ewc_loss": 0.046142578125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.592374801635742, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8750482201576233, "num_tokens": 119581657.0, "step": 3148 }, { "epoch": 0.40058516728151633, "ewc_loss": 0.044189453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3804116249084473, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8621255159378052, "num_tokens": 119617140.0, "step": 3149 }, { "epoch": 0.40071237756010686, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5902881622314453, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8390154242515564, "num_tokens": 119660155.0, "step": 3150 }, { "epoch": 0.4008395878386974, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3714277744293213, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.841484785079956, "num_tokens": 119706258.0, "step": 3151 }, { "epoch": 0.40096679811728786, "ewc_loss": 0.046142578125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.590097188949585, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8572102785110474, "num_tokens": 119742564.0, "step": 3152 }, { "epoch": 0.4010940083958784, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.366222381591797, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8547346591949463, "num_tokens": 119781301.0, "step": 3153 }, { "epoch": 0.4012212186744689, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6018502712249756, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8482733964920044, "num_tokens": 119816513.0, "step": 3154 }, { "epoch": 0.4013484289530594, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3570590019226074, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8584327101707458, "num_tokens": 119858510.0, "step": 3155 }, { "epoch": 0.4014756392316499, "ewc_loss": 0.046630859375, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6039860248565674, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8579118847846985, "num_tokens": 119902243.0, "step": 3156 }, { "epoch": 0.40160284951024044, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.366331100463867, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.862500786781311, "num_tokens": 119938956.0, "step": 3157 }, { "epoch": 0.4017300597888309, "ewc_loss": 0.046630859375, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.611201524734497, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8651567697525024, "num_tokens": 119976070.0, "step": 3158 }, { "epoch": 0.40185727006742145, "ewc_loss": 0.043701171875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.369650363922119, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8602652549743652, "num_tokens": 120011811.0, "step": 3159 }, { "epoch": 0.401984480346012, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6382360458374023, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8491697311401367, "num_tokens": 120042278.0, "step": 3160 }, { "epoch": 0.40211169062460245, "ewc_loss": 0.04345703125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3697400093078613, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.854039192199707, "num_tokens": 120078379.0, "step": 3161 }, { "epoch": 0.402238900903193, "ewc_loss": 0.046630859375, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6175575256347656, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.865636944770813, "num_tokens": 120118852.0, "step": 3162 }, { "epoch": 0.4023661111817835, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3785502910614014, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8656294345855713, "num_tokens": 120153825.0, "step": 3163 }, { "epoch": 0.402493321460374, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.625136137008667, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8517318964004517, "num_tokens": 120190945.0, "step": 3164 }, { "epoch": 0.4026205317389645, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3925957679748535, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8623310327529907, "num_tokens": 120221435.0, "step": 3165 }, { "epoch": 0.40274774201755503, "ewc_loss": 0.04638671875, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6186575889587402, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8618395924568176, "num_tokens": 120253553.0, "step": 3166 }, { "epoch": 0.4028749522961455, "ewc_loss": 0.0439453125, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3862950801849365, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8502954244613647, "num_tokens": 120291438.0, "step": 3167 }, { "epoch": 0.40300216257473603, "ewc_loss": 0.0458984375, "ewc_loss_diag": 8.96453857421875e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.607879877090454, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8505273461341858, "num_tokens": 120332005.0, "step": 3168 }, { "epoch": 0.40312937285332656, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.386604070663452, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8373017311096191, "num_tokens": 120369280.0, "step": 3169 }, { "epoch": 0.40325658313191703, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6102707386016846, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8564709424972534, "num_tokens": 120402350.0, "step": 3170 }, { "epoch": 0.40338379341050756, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3784539699554443, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8611035943031311, "num_tokens": 120441791.0, "step": 3171 }, { "epoch": 0.4035110036890981, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5953311920166016, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8691733479499817, "num_tokens": 120476895.0, "step": 3172 }, { "epoch": 0.40363821396768856, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3776791095733643, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.848810613155365, "num_tokens": 120518016.0, "step": 3173 }, { "epoch": 0.4037654242462791, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5965726375579834, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8679471015930176, "num_tokens": 120551500.0, "step": 3174 }, { "epoch": 0.4038926345248696, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3796916007995605, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8641332387924194, "num_tokens": 120588543.0, "step": 3175 }, { "epoch": 0.4040198448034601, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5928094387054443, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8527741432189941, "num_tokens": 120634394.0, "step": 3176 }, { "epoch": 0.4041470550820506, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3784117698669434, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8510924577713013, "num_tokens": 120672982.0, "step": 3177 }, { "epoch": 0.40427426536064115, "ewc_loss": 0.045654296875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.594125270843506, "learning_rate": 1e-05, "loss": 0.5486, "mean_token_accuracy": 0.8324121236801147, "num_tokens": 120712091.0, "step": 3178 }, { "epoch": 0.4044014756392316, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3944814205169678, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8665761947631836, "num_tokens": 120743530.0, "step": 3179 }, { "epoch": 0.40452868591782215, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.571744203567505, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8673147559165955, "num_tokens": 120778884.0, "step": 3180 }, { "epoch": 0.4046558961964127, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3800034523010254, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8507133722305298, "num_tokens": 120815951.0, "step": 3181 }, { "epoch": 0.4047831064750032, "ewc_loss": 0.045654296875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5698530673980713, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8550107479095459, "num_tokens": 120853111.0, "step": 3182 }, { "epoch": 0.4049103167535937, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3768718242645264, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8501812815666199, "num_tokens": 120894041.0, "step": 3183 }, { "epoch": 0.4050375270321842, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.567376136779785, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8606624007225037, "num_tokens": 120930730.0, "step": 3184 }, { "epoch": 0.40516473731077474, "ewc_loss": 0.04345703125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3959403038024902, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8512240052223206, "num_tokens": 120965044.0, "step": 3185 }, { "epoch": 0.4052919475893652, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5550403594970703, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8527424335479736, "num_tokens": 120999777.0, "step": 3186 }, { "epoch": 0.40541915786795574, "ewc_loss": 0.04345703125, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.4006693363189697, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8390454053878784, "num_tokens": 121031910.0, "step": 3187 }, { "epoch": 0.40554636814654627, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.540968894958496, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8598665595054626, "num_tokens": 121069600.0, "step": 3188 }, { "epoch": 0.40567357842513674, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.4116532802581787, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8417139053344727, "num_tokens": 121103790.0, "step": 3189 }, { "epoch": 0.40580078870372727, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5341038703918457, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8479009866714478, "num_tokens": 121135199.0, "step": 3190 }, { "epoch": 0.4059279989823178, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3814597129821777, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8541792631149292, "num_tokens": 121176040.0, "step": 3191 }, { "epoch": 0.40605520926090827, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.533594846725464, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8440757989883423, "num_tokens": 121214675.0, "step": 3192 }, { "epoch": 0.4061824195394988, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.382089614868164, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8401229977607727, "num_tokens": 121252834.0, "step": 3193 }, { "epoch": 0.4063096298180893, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5335237979888916, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8512399196624756, "num_tokens": 121292837.0, "step": 3194 }, { "epoch": 0.4064368400966798, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3680648803710938, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8596148490905762, "num_tokens": 121332353.0, "step": 3195 }, { "epoch": 0.4065640503752703, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.54123854637146, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8633826971054077, "num_tokens": 121371095.0, "step": 3196 }, { "epoch": 0.40669126065386085, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.107589721679688e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3685853481292725, "learning_rate": 1e-05, "loss": 0.5368, "mean_token_accuracy": 0.8376134037971497, "num_tokens": 121409653.0, "step": 3197 }, { "epoch": 0.4068184709324513, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5585339069366455, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8657797574996948, "num_tokens": 121449309.0, "step": 3198 }, { "epoch": 0.40694568121104185, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3490219116210938, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8607475757598877, "num_tokens": 121495004.0, "step": 3199 }, { "epoch": 0.4070728914896324, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.059906005859375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.568265199661255, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8631303310394287, "num_tokens": 121537679.0, "step": 3200 }, { "epoch": 0.40720010176822286, "ewc_loss": 0.04345703125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.344972610473633, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8528826236724854, "num_tokens": 121579156.0, "step": 3201 }, { "epoch": 0.4073273120468134, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6273598670959473, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8520604968070984, "num_tokens": 121611056.0, "step": 3202 }, { "epoch": 0.4074545223254039, "ewc_loss": 0.04345703125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3443191051483154, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8648227453231812, "num_tokens": 121642821.0, "step": 3203 }, { "epoch": 0.4075817326039944, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6514840126037598, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8502000570297241, "num_tokens": 121674021.0, "step": 3204 }, { "epoch": 0.4077089428825849, "ewc_loss": 0.04345703125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3380253314971924, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.858353853225708, "num_tokens": 121711434.0, "step": 3205 }, { "epoch": 0.40783615316117544, "ewc_loss": 0.048095703125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6680915355682373, "learning_rate": 1e-05, "loss": 0.5253, "mean_token_accuracy": 0.8404142260551453, "num_tokens": 121755357.0, "step": 3206 }, { "epoch": 0.4079633634397659, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3571112155914307, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8623801469802856, "num_tokens": 121791703.0, "step": 3207 }, { "epoch": 0.40809057371835644, "ewc_loss": 0.048583984375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.684473752975464, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8676499128341675, "num_tokens": 121824940.0, "step": 3208 }, { "epoch": 0.40821778399694697, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3895583152770996, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8588760495185852, "num_tokens": 121860434.0, "step": 3209 }, { "epoch": 0.40834499427553744, "ewc_loss": 0.0478515625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6854681968688965, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8421186208724976, "num_tokens": 121897536.0, "step": 3210 }, { "epoch": 0.40847220455412797, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3862814903259277, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.839211106300354, "num_tokens": 121936721.0, "step": 3211 }, { "epoch": 0.4085994148327185, "ewc_loss": 0.048095703125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6667613983154297, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8485846519470215, "num_tokens": 121977407.0, "step": 3212 }, { "epoch": 0.40872662511130897, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4150149822235107, "learning_rate": 1e-05, "loss": 0.5379, "mean_token_accuracy": 0.8375126719474792, "num_tokens": 122015068.0, "step": 3213 }, { "epoch": 0.4088538353898995, "ewc_loss": 0.04736328125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.629213809967041, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8525213599205017, "num_tokens": 122058899.0, "step": 3214 }, { "epoch": 0.40898104566849003, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.408764362335205, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.857901394367218, "num_tokens": 122095407.0, "step": 3215 }, { "epoch": 0.4091082559470805, "ewc_loss": 0.047119140625, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.607330322265625, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8598989844322205, "num_tokens": 122134836.0, "step": 3216 }, { "epoch": 0.40923546622567103, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.431288242340088, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8554098010063171, "num_tokens": 122165959.0, "step": 3217 }, { "epoch": 0.40936267650426156, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5704097747802734, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8489298820495605, "num_tokens": 122206223.0, "step": 3218 }, { "epoch": 0.40948988678285203, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.404141902923584, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8615097999572754, "num_tokens": 122245416.0, "step": 3219 }, { "epoch": 0.40961709706144256, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.56333065032959, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8393539190292358, "num_tokens": 122285264.0, "step": 3220 }, { "epoch": 0.4097443073400331, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.390472888946533, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8522305488586426, "num_tokens": 122326956.0, "step": 3221 }, { "epoch": 0.40987151761862356, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5649349689483643, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8628171682357788, "num_tokens": 122362409.0, "step": 3222 }, { "epoch": 0.4099987278972141, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.386719226837158, "learning_rate": 1e-05, "loss": 0.5421, "mean_token_accuracy": 0.8325170874595642, "num_tokens": 122405011.0, "step": 3223 }, { "epoch": 0.4101259381758046, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.558997392654419, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8596185445785522, "num_tokens": 122445482.0, "step": 3224 }, { "epoch": 0.4102531484543951, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3718619346618652, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8595881462097168, "num_tokens": 122485024.0, "step": 3225 }, { "epoch": 0.4103803587329856, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.561116933822632, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8563171625137329, "num_tokens": 122524766.0, "step": 3226 }, { "epoch": 0.41050756901157615, "ewc_loss": 0.04345703125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3644118309020996, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8565322160720825, "num_tokens": 122566664.0, "step": 3227 }, { "epoch": 0.4106347792901666, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.63670015335083, "learning_rate": 1e-05, "loss": 0.5787, "mean_token_accuracy": 0.8215270042419434, "num_tokens": 122606724.0, "step": 3228 }, { "epoch": 0.41076198956875715, "ewc_loss": 0.04345703125, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.370584011077881, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8484679460525513, "num_tokens": 122641664.0, "step": 3229 }, { "epoch": 0.4108891998473477, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.1552734375e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6125142574310303, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8730829954147339, "num_tokens": 122673370.0, "step": 3230 }, { "epoch": 0.4110164101259382, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.358330726623535, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8532685041427612, "num_tokens": 122711545.0, "step": 3231 }, { "epoch": 0.4111436204045287, "ewc_loss": 0.046875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.636824607849121, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8575811386108398, "num_tokens": 122743656.0, "step": 3232 }, { "epoch": 0.4112708306831192, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3734352588653564, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8489699959754944, "num_tokens": 122778233.0, "step": 3233 }, { "epoch": 0.41139804096170973, "ewc_loss": 0.047119140625, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6207101345062256, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8636173605918884, "num_tokens": 122821063.0, "step": 3234 }, { "epoch": 0.4115252512403002, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3870387077331543, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8523616194725037, "num_tokens": 122855871.0, "step": 3235 }, { "epoch": 0.41165246151889073, "ewc_loss": 0.046875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6200122833251953, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.856788158416748, "num_tokens": 122893990.0, "step": 3236 }, { "epoch": 0.41177967179748126, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.380592107772827, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8585902452468872, "num_tokens": 122932017.0, "step": 3237 }, { "epoch": 0.41190688207607173, "ewc_loss": 0.046875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6055479049682617, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8608957529067993, "num_tokens": 122975581.0, "step": 3238 }, { "epoch": 0.41203409235466226, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4080076217651367, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8502339124679565, "num_tokens": 123006051.0, "step": 3239 }, { "epoch": 0.4121613026332528, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5853021144866943, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8504990339279175, "num_tokens": 123049756.0, "step": 3240 }, { "epoch": 0.41228851291184326, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3934314250946045, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8464946746826172, "num_tokens": 123091059.0, "step": 3241 }, { "epoch": 0.4124157231904338, "ewc_loss": 0.045654296875, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.570437431335449, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8629682064056396, "num_tokens": 123130369.0, "step": 3242 }, { "epoch": 0.4125429334690243, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3819286823272705, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8560678958892822, "num_tokens": 123173507.0, "step": 3243 }, { "epoch": 0.4126701437476148, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.202957153320312e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.569995641708374, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.864851713180542, "num_tokens": 123211428.0, "step": 3244 }, { "epoch": 0.4127973540262053, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.373509407043457, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8490121364593506, "num_tokens": 123256115.0, "step": 3245 }, { "epoch": 0.41292456430479585, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.571805000305176, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8562291860580444, "num_tokens": 123297319.0, "step": 3246 }, { "epoch": 0.4130517745833863, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.377497911453247, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8401584625244141, "num_tokens": 123341241.0, "step": 3247 }, { "epoch": 0.41317898486197685, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5788726806640625, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8450193405151367, "num_tokens": 123381832.0, "step": 3248 }, { "epoch": 0.4133061951405674, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.37749981880188, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8576767444610596, "num_tokens": 123418393.0, "step": 3249 }, { "epoch": 0.41343340541915785, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5868563652038574, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8658915162086487, "num_tokens": 123449916.0, "step": 3250 }, { "epoch": 0.4135606156977484, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3836514949798584, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8774977922439575, "num_tokens": 123484025.0, "step": 3251 }, { "epoch": 0.4136878259763389, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6251752376556396, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8585892915725708, "num_tokens": 123522345.0, "step": 3252 }, { "epoch": 0.4138150362549294, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.362597942352295, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.854408323764801, "num_tokens": 123562123.0, "step": 3253 }, { "epoch": 0.4139422465335199, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.606825351715088, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8480530977249146, "num_tokens": 123606014.0, "step": 3254 }, { "epoch": 0.41406945681211044, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.358213424682617, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8659589290618896, "num_tokens": 123643483.0, "step": 3255 }, { "epoch": 0.4141966670907009, "ewc_loss": 0.046630859375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6038527488708496, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8577135801315308, "num_tokens": 123689303.0, "step": 3256 }, { "epoch": 0.41432387736929144, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.390418767929077, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8612315654754639, "num_tokens": 123722388.0, "step": 3257 }, { "epoch": 0.41445108764788197, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.603224992752075, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8548444509506226, "num_tokens": 123761761.0, "step": 3258 }, { "epoch": 0.41457829792647244, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.365502119064331, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8773910999298096, "num_tokens": 123799138.0, "step": 3259 }, { "epoch": 0.41470550820506297, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.595576763153076, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8614758253097534, "num_tokens": 123840803.0, "step": 3260 }, { "epoch": 0.4148327184836535, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.386270761489868, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8520007133483887, "num_tokens": 123875705.0, "step": 3261 }, { "epoch": 0.41495992876224397, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5990042686462402, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8418675661087036, "num_tokens": 123915830.0, "step": 3262 }, { "epoch": 0.4150871390408345, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3854055404663086, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8444100618362427, "num_tokens": 123953855.0, "step": 3263 }, { "epoch": 0.415214349319425, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.594348669052124, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8558738827705383, "num_tokens": 123985697.0, "step": 3264 }, { "epoch": 0.4153415595980155, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3791980743408203, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8529999852180481, "num_tokens": 124021682.0, "step": 3265 }, { "epoch": 0.415468769876606, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.6074814796447754, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8419622182846069, "num_tokens": 124057472.0, "step": 3266 }, { "epoch": 0.41559598015519655, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.361372232437134, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8573598861694336, "num_tokens": 124100160.0, "step": 3267 }, { "epoch": 0.415723190433787, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.611111879348755, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8542124032974243, "num_tokens": 124138503.0, "step": 3268 }, { "epoch": 0.41585040071237755, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3804216384887695, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8451884984970093, "num_tokens": 124176082.0, "step": 3269 }, { "epoch": 0.4159776109909681, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6081974506378174, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8496649861335754, "num_tokens": 124213853.0, "step": 3270 }, { "epoch": 0.41610482126955856, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3663222789764404, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8722847700119019, "num_tokens": 124250562.0, "step": 3271 }, { "epoch": 0.4162320315481491, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6382534503936768, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.8389155268669128, "num_tokens": 124282321.0, "step": 3272 }, { "epoch": 0.4163592418267396, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.372549057006836, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8603081703186035, "num_tokens": 124323172.0, "step": 3273 }, { "epoch": 0.4164864521053301, "ewc_loss": 0.046875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.630280017852783, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8530037999153137, "num_tokens": 124361585.0, "step": 3274 }, { "epoch": 0.4166136623839206, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.381833076477051, "learning_rate": 1e-05, "loss": 0.562, "mean_token_accuracy": 0.8331493139266968, "num_tokens": 124398926.0, "step": 3275 }, { "epoch": 0.41674087266251114, "ewc_loss": 0.046630859375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6209142208099365, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8694422841072083, "num_tokens": 124437914.0, "step": 3276 }, { "epoch": 0.4168680829411016, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3795931339263916, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8569545745849609, "num_tokens": 124478850.0, "step": 3277 }, { "epoch": 0.41699529321969214, "ewc_loss": 0.046630859375, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6114816665649414, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8698239922523499, "num_tokens": 124519361.0, "step": 3278 }, { "epoch": 0.41712250349828267, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3935892581939697, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.852638304233551, "num_tokens": 124555616.0, "step": 3279 }, { "epoch": 0.4172497137768732, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6127336025238037, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.8375377655029297, "num_tokens": 124593045.0, "step": 3280 }, { "epoch": 0.41737692405546367, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.395080089569092, "learning_rate": 1e-05, "loss": 0.5445, "mean_token_accuracy": 0.8326922655105591, "num_tokens": 124631842.0, "step": 3281 }, { "epoch": 0.4175041343340542, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5799787044525146, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8507453203201294, "num_tokens": 124674466.0, "step": 3282 }, { "epoch": 0.41763134461264473, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.39339280128479, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8593370914459229, "num_tokens": 124707893.0, "step": 3283 }, { "epoch": 0.4177585548912352, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5726218223571777, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8538622260093689, "num_tokens": 124746230.0, "step": 3284 }, { "epoch": 0.41788576516982573, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.391465425491333, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8541254997253418, "num_tokens": 124786988.0, "step": 3285 }, { "epoch": 0.41801297544841626, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5616986751556396, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8474042415618896, "num_tokens": 124831061.0, "step": 3286 }, { "epoch": 0.41814018572700673, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.387390613555908, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8511466979980469, "num_tokens": 124870486.0, "step": 3287 }, { "epoch": 0.41826739600559726, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5510692596435547, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8674333095550537, "num_tokens": 124909675.0, "step": 3288 }, { "epoch": 0.4183946062841878, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.383157730102539, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.843241274356842, "num_tokens": 124953854.0, "step": 3289 }, { "epoch": 0.41852181656277826, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5512337684631348, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8530805110931396, "num_tokens": 124991117.0, "step": 3290 }, { "epoch": 0.4186490268413688, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3844316005706787, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8531994819641113, "num_tokens": 125027160.0, "step": 3291 }, { "epoch": 0.4187762371199593, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5529286861419678, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.860175609588623, "num_tokens": 125061324.0, "step": 3292 }, { "epoch": 0.4189034473985498, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.383540391921997, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8522350788116455, "num_tokens": 125099616.0, "step": 3293 }, { "epoch": 0.4190306576771403, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.250640869140625e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5629942417144775, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8505693674087524, "num_tokens": 125132855.0, "step": 3294 }, { "epoch": 0.41915786795573085, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.380715847015381, "learning_rate": 1e-05, "loss": 0.5215, "mean_token_accuracy": 0.8470197916030884, "num_tokens": 125168040.0, "step": 3295 }, { "epoch": 0.4192850782343213, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.566879987716675, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8600673675537109, "num_tokens": 125198788.0, "step": 3296 }, { "epoch": 0.41941228851291185, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3621444702148438, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8616109490394592, "num_tokens": 125238387.0, "step": 3297 }, { "epoch": 0.4195394987915024, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.579692840576172, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.865591824054718, "num_tokens": 125273666.0, "step": 3298 }, { "epoch": 0.41966670907009285, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3652801513671875, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8451615571975708, "num_tokens": 125315961.0, "step": 3299 }, { "epoch": 0.4197939193486834, "ewc_loss": 0.045654296875, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.583057165145874, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.84342360496521, "num_tokens": 125358830.0, "step": 3300 }, { "epoch": 0.4199211296272739, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.360027551651001, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8479887247085571, "num_tokens": 125397497.0, "step": 3301 }, { "epoch": 0.4200483399058644, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5988993644714355, "learning_rate": 1e-05, "loss": 0.5424, "mean_token_accuracy": 0.8341964483261108, "num_tokens": 125437337.0, "step": 3302 }, { "epoch": 0.4201755501844549, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3589277267456055, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8634554743766785, "num_tokens": 125470535.0, "step": 3303 }, { "epoch": 0.42030276046304543, "ewc_loss": 0.046630859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.60986065864563, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8412457704544067, "num_tokens": 125510692.0, "step": 3304 }, { "epoch": 0.4204299707416359, "ewc_loss": 0.043701171875, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3447985649108887, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8392329216003418, "num_tokens": 125557328.0, "step": 3305 }, { "epoch": 0.42055718102022643, "ewc_loss": 0.046630859375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6384100914001465, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8501003384590149, "num_tokens": 125594006.0, "step": 3306 }, { "epoch": 0.42068439129881696, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3724958896636963, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8496994376182556, "num_tokens": 125627692.0, "step": 3307 }, { "epoch": 0.42081160157740743, "ewc_loss": 0.04736328125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6428096294403076, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.857229471206665, "num_tokens": 125660218.0, "step": 3308 }, { "epoch": 0.42093881185599796, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.370608329772949, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8647462725639343, "num_tokens": 125695793.0, "step": 3309 }, { "epoch": 0.4210660221345885, "ewc_loss": 0.047119140625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6485607624053955, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8565177917480469, "num_tokens": 125732438.0, "step": 3310 }, { "epoch": 0.42119323241317896, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3863277435302734, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8616752624511719, "num_tokens": 125764919.0, "step": 3311 }, { "epoch": 0.4213204426917695, "ewc_loss": 0.04736328125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6554017066955566, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8528986573219299, "num_tokens": 125797172.0, "step": 3312 }, { "epoch": 0.42144765297036, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.380755662918091, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8574193120002747, "num_tokens": 125838144.0, "step": 3313 }, { "epoch": 0.4215748632489505, "ewc_loss": 0.047119140625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.640902519226074, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.865309476852417, "num_tokens": 125873897.0, "step": 3314 }, { "epoch": 0.421702073527541, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.404618501663208, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8403415679931641, "num_tokens": 125909773.0, "step": 3315 }, { "epoch": 0.42182928380613155, "ewc_loss": 0.047119140625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.617102861404419, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8585762977600098, "num_tokens": 125947661.0, "step": 3316 }, { "epoch": 0.421956494084722, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4139182567596436, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8586896657943726, "num_tokens": 125979483.0, "step": 3317 }, { "epoch": 0.42208370436331255, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5885939598083496, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8656558394432068, "num_tokens": 126016907.0, "step": 3318 }, { "epoch": 0.4222109146419031, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.392115592956543, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8499754667282104, "num_tokens": 126059475.0, "step": 3319 }, { "epoch": 0.42233812492049355, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.586364984512329, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8714174032211304, "num_tokens": 126097373.0, "step": 3320 }, { "epoch": 0.4224653351990841, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3862836360931396, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8670107126235962, "num_tokens": 126138884.0, "step": 3321 }, { "epoch": 0.4225925454776746, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5836362838745117, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8612127304077148, "num_tokens": 126178453.0, "step": 3322 }, { "epoch": 0.4227197557562651, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.409789800643921, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8499143123626709, "num_tokens": 126208979.0, "step": 3323 }, { "epoch": 0.4228469660348556, "ewc_loss": 0.045654296875, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5558907985687256, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8587510585784912, "num_tokens": 126252195.0, "step": 3324 }, { "epoch": 0.42297417631344614, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.393311023712158, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8607999682426453, "num_tokens": 126287526.0, "step": 3325 }, { "epoch": 0.4231013865920366, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5493061542510986, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8662161827087402, "num_tokens": 126325868.0, "step": 3326 }, { "epoch": 0.42322859687062714, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.298324584960938e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.404772996902466, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8584674596786499, "num_tokens": 126364613.0, "step": 3327 }, { "epoch": 0.42335580714921767, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5393171310424805, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.867283284664154, "num_tokens": 126404373.0, "step": 3328 }, { "epoch": 0.42348301742780814, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4036753177642822, "learning_rate": 1e-05, "loss": 0.5419, "mean_token_accuracy": 0.8349983096122742, "num_tokens": 126444557.0, "step": 3329 }, { "epoch": 0.42361022770639867, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.531656503677368, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.850824236869812, "num_tokens": 126479657.0, "step": 3330 }, { "epoch": 0.4237374379849892, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.4016435146331787, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8469980955123901, "num_tokens": 126518004.0, "step": 3331 }, { "epoch": 0.4238646482635797, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.525747537612915, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8487280607223511, "num_tokens": 126557854.0, "step": 3332 }, { "epoch": 0.4239918585421702, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3868556022644043, "learning_rate": 1e-05, "loss": 0.5167, "mean_token_accuracy": 0.8511785268783569, "num_tokens": 126593657.0, "step": 3333 }, { "epoch": 0.4241190688207607, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5161023139953613, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8780274391174316, "num_tokens": 126635913.0, "step": 3334 }, { "epoch": 0.42424627909935125, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3999645709991455, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8425620794296265, "num_tokens": 126668703.0, "step": 3335 }, { "epoch": 0.4243734893779417, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.506361961364746, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8697273135185242, "num_tokens": 126707952.0, "step": 3336 }, { "epoch": 0.42450069965653225, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.981560230255127, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8552253246307373, "num_tokens": 126748756.0, "step": 3337 }, { "epoch": 0.4246279099351228, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3789546489715576, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8657935857772827, "num_tokens": 126784780.0, "step": 3338 }, { "epoch": 0.42475512021371326, "ewc_loss": 0.046875, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6547815799713135, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8598805069923401, "num_tokens": 126827371.0, "step": 3339 }, { "epoch": 0.4248823304923038, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3236114978790283, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8506285548210144, "num_tokens": 126862357.0, "step": 3340 }, { "epoch": 0.4250095407708943, "ewc_loss": 0.05126953125, "ewc_loss_diag": 9.393692016601562e-05, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.740175247192383, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8750658631324768, "num_tokens": 126902433.0, "step": 3341 }, { "epoch": 0.4251367510494848, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.35649037361145, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8492597341537476, "num_tokens": 126946958.0, "step": 3342 }, { "epoch": 0.4252639613280753, "ewc_loss": 0.0517578125, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.773928165435791, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.865250825881958, "num_tokens": 126980267.0, "step": 3343 }, { "epoch": 0.42539117160666584, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.377035140991211, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8742086291313171, "num_tokens": 127018215.0, "step": 3344 }, { "epoch": 0.4255183818852563, "ewc_loss": 0.05322265625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.802386999130249, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8473743200302124, "num_tokens": 127055844.0, "step": 3345 }, { "epoch": 0.42564559216384684, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.401888370513916, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8601312637329102, "num_tokens": 127099373.0, "step": 3346 }, { "epoch": 0.42577280244243737, "ewc_loss": 0.053955078125, "ewc_loss_diag": 9.441375732421875e-05, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.797001600265503, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8578875660896301, "num_tokens": 127143755.0, "step": 3347 }, { "epoch": 0.42590001272102784, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4527804851531982, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8643943667411804, "num_tokens": 127180543.0, "step": 3348 }, { "epoch": 0.42602722299961837, "ewc_loss": 0.05322265625, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7733094692230225, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.850066065788269, "num_tokens": 127225203.0, "step": 3349 }, { "epoch": 0.4261544332782089, "ewc_loss": 0.046875, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4896254539489746, "learning_rate": 1e-05, "loss": 0.5458, "mean_token_accuracy": 0.8367466926574707, "num_tokens": 127260949.0, "step": 3350 }, { "epoch": 0.4262816435567994, "ewc_loss": 0.0517578125, "ewc_loss_diag": 9.489059448242188e-05, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.723127841949463, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8718012571334839, "num_tokens": 127297487.0, "step": 3351 }, { "epoch": 0.4264088538353899, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4710538387298584, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.8401591777801514, "num_tokens": 127342612.0, "step": 3352 }, { "epoch": 0.42653606411398043, "ewc_loss": 0.049072265625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.67383074760437, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8652278184890747, "num_tokens": 127383915.0, "step": 3353 }, { "epoch": 0.4266632743925709, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4666976928710938, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.849815845489502, "num_tokens": 127421564.0, "step": 3354 }, { "epoch": 0.42679048467116143, "ewc_loss": 0.048095703125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6463723182678223, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8391432166099548, "num_tokens": 127463750.0, "step": 3355 }, { "epoch": 0.42691769494975196, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.459397315979004, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8490282893180847, "num_tokens": 127502793.0, "step": 3356 }, { "epoch": 0.42704490522834243, "ewc_loss": 0.047607421875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6066160202026367, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8543727397918701, "num_tokens": 127542650.0, "step": 3357 }, { "epoch": 0.42717211550693296, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.418956756591797, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8656017780303955, "num_tokens": 127584753.0, "step": 3358 }, { "epoch": 0.4272993257855235, "ewc_loss": 0.046875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.599106788635254, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8476214408874512, "num_tokens": 127624427.0, "step": 3359 }, { "epoch": 0.42742653606411396, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4097487926483154, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8646920919418335, "num_tokens": 127665636.0, "step": 3360 }, { "epoch": 0.4275537463427045, "ewc_loss": 0.046875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.590852975845337, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8505664467811584, "num_tokens": 127706066.0, "step": 3361 }, { "epoch": 0.427680956621295, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.396021842956543, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8660100698471069, "num_tokens": 127741680.0, "step": 3362 }, { "epoch": 0.4278081668998855, "ewc_loss": 0.046630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.595496892929077, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8505662679672241, "num_tokens": 127781147.0, "step": 3363 }, { "epoch": 0.427935377178476, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3942337036132812, "learning_rate": 1e-05, "loss": 0.5437, "mean_token_accuracy": 0.8356181383132935, "num_tokens": 127823787.0, "step": 3364 }, { "epoch": 0.42806258745706655, "ewc_loss": 0.046630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.595371723175049, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8339571952819824, "num_tokens": 127862239.0, "step": 3365 }, { "epoch": 0.428189797735657, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3815622329711914, "learning_rate": 1e-05, "loss": 0.5305, "mean_token_accuracy": 0.8427090048789978, "num_tokens": 127904376.0, "step": 3366 }, { "epoch": 0.42831700801424755, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6016533374786377, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8428683280944824, "num_tokens": 127944132.0, "step": 3367 }, { "epoch": 0.4284442182928381, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3883135318756104, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8450776934623718, "num_tokens": 127983512.0, "step": 3368 }, { "epoch": 0.42857142857142855, "ewc_loss": 0.046630859375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6036300659179688, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8622646331787109, "num_tokens": 128018864.0, "step": 3369 }, { "epoch": 0.4286986388500191, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3666930198669434, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8707324266433716, "num_tokens": 128056586.0, "step": 3370 }, { "epoch": 0.4288258491286096, "ewc_loss": 0.046875, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6270956993103027, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8503815531730652, "num_tokens": 128091131.0, "step": 3371 }, { "epoch": 0.4289530594072001, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.5367431640625e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3893728256225586, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8484907150268555, "num_tokens": 128125455.0, "step": 3372 }, { "epoch": 0.4290802696857906, "ewc_loss": 0.046875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.61218523979187, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.851188063621521, "num_tokens": 128164357.0, "step": 3373 }, { "epoch": 0.42920747996438113, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3879554271698, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.849909245967865, "num_tokens": 128203590.0, "step": 3374 }, { "epoch": 0.4293346902429716, "ewc_loss": 0.046630859375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6091973781585693, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8510440587997437, "num_tokens": 128238819.0, "step": 3375 }, { "epoch": 0.42946190052156213, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3831048011779785, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8711830973625183, "num_tokens": 128273879.0, "step": 3376 }, { "epoch": 0.42958911080015266, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.612687826156616, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8572260141372681, "num_tokens": 128309688.0, "step": 3377 }, { "epoch": 0.42971632107874314, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.584426879882812e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4024343490600586, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8567181825637817, "num_tokens": 128341515.0, "step": 3378 }, { "epoch": 0.42984353135733366, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.590703010559082, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8610355257987976, "num_tokens": 128382676.0, "step": 3379 }, { "epoch": 0.4299707416359242, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.391375780105591, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8732864856719971, "num_tokens": 128417512.0, "step": 3380 }, { "epoch": 0.4300979519145147, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.572173595428467, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8601487874984741, "num_tokens": 128457695.0, "step": 3381 }, { "epoch": 0.4302251621931052, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.415250062942505, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8548562526702881, "num_tokens": 128487459.0, "step": 3382 }, { "epoch": 0.4303523724716957, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5626347064971924, "learning_rate": 1e-05, "loss": 0.5405, "mean_token_accuracy": 0.8320211172103882, "num_tokens": 128526823.0, "step": 3383 }, { "epoch": 0.43047958275028625, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3946619033813477, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8510268330574036, "num_tokens": 128566841.0, "step": 3384 }, { "epoch": 0.4306067930288767, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5471386909484863, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8457227945327759, "num_tokens": 128608147.0, "step": 3385 }, { "epoch": 0.43073400330746725, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3833515644073486, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8589736223220825, "num_tokens": 128648277.0, "step": 3386 }, { "epoch": 0.4308612135860578, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.562589406967163, "learning_rate": 1e-05, "loss": 0.5366, "mean_token_accuracy": 0.8389893770217896, "num_tokens": 128682547.0, "step": 3387 }, { "epoch": 0.43098842386464825, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3913633823394775, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8595341444015503, "num_tokens": 128715103.0, "step": 3388 }, { "epoch": 0.4311156341432388, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.570197582244873, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8509536385536194, "num_tokens": 128748924.0, "step": 3389 }, { "epoch": 0.4312428444218293, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3715415000915527, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8504902124404907, "num_tokens": 128786879.0, "step": 3390 }, { "epoch": 0.4313700547004198, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5729756355285645, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8528107404708862, "num_tokens": 128822121.0, "step": 3391 }, { "epoch": 0.4314972649790103, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3900537490844727, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8353861570358276, "num_tokens": 128858594.0, "step": 3392 }, { "epoch": 0.43162447525760084, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5675106048583984, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8536137938499451, "num_tokens": 128896147.0, "step": 3393 }, { "epoch": 0.4317516855361913, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.429168462753296, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8422854542732239, "num_tokens": 128930055.0, "step": 3394 }, { "epoch": 0.43187889581478184, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.553034782409668, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8532009124755859, "num_tokens": 128968577.0, "step": 3395 }, { "epoch": 0.43200610609337237, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.379833698272705, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8491071462631226, "num_tokens": 129007074.0, "step": 3396 }, { "epoch": 0.43213331637196284, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5498080253601074, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8459200859069824, "num_tokens": 129044722.0, "step": 3397 }, { "epoch": 0.43226052665055337, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3812766075134277, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8550786375999451, "num_tokens": 129081567.0, "step": 3398 }, { "epoch": 0.4323877369291439, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5531277656555176, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8505754470825195, "num_tokens": 129119354.0, "step": 3399 }, { "epoch": 0.43251494720773437, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3726389408111572, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8667713403701782, "num_tokens": 129157871.0, "step": 3400 }, { "epoch": 0.4326421574863249, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.555107593536377, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8635364174842834, "num_tokens": 129200164.0, "step": 3401 }, { "epoch": 0.4327693677649154, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.376555919647217, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8516402840614319, "num_tokens": 129239382.0, "step": 3402 }, { "epoch": 0.4328965780435059, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.632110595703125e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.577202558517456, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8471498489379883, "num_tokens": 129273201.0, "step": 3403 }, { "epoch": 0.4330237883220964, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3615806102752686, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8637552857398987, "num_tokens": 129309636.0, "step": 3404 }, { "epoch": 0.43315099860068695, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5815141201019287, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.86335289478302, "num_tokens": 129345803.0, "step": 3405 }, { "epoch": 0.4332782088792774, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.357337713241577, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8605896830558777, "num_tokens": 129384955.0, "step": 3406 }, { "epoch": 0.43340541915786795, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5956056118011475, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8712623119354248, "num_tokens": 129425129.0, "step": 3407 }, { "epoch": 0.4335326294364585, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3520114421844482, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8514200448989868, "num_tokens": 129465931.0, "step": 3408 }, { "epoch": 0.43365983971504896, "ewc_loss": 0.047607421875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6281886100769043, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8645824193954468, "num_tokens": 129497296.0, "step": 3409 }, { "epoch": 0.4337870499936395, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3448448181152344, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8600414395332336, "num_tokens": 129539037.0, "step": 3410 }, { "epoch": 0.43391426027223, "ewc_loss": 0.04833984375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.645660638809204, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8495734930038452, "num_tokens": 129581036.0, "step": 3411 }, { "epoch": 0.4340414705508205, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3685061931610107, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8491413593292236, "num_tokens": 129617515.0, "step": 3412 }, { "epoch": 0.434168680829411, "ewc_loss": 0.04833984375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.642982244491577, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8753623962402344, "num_tokens": 129659348.0, "step": 3413 }, { "epoch": 0.43429589110800154, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.403564691543579, "learning_rate": 1e-05, "loss": 0.539, "mean_token_accuracy": 0.834166407585144, "num_tokens": 129695000.0, "step": 3414 }, { "epoch": 0.434423101386592, "ewc_loss": 0.047607421875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6344096660614014, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8491069078445435, "num_tokens": 129734125.0, "step": 3415 }, { "epoch": 0.43455031166518254, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.383007287979126, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8541418313980103, "num_tokens": 129774799.0, "step": 3416 }, { "epoch": 0.43467752194377307, "ewc_loss": 0.047607421875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6172597408294678, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8615289330482483, "num_tokens": 129814390.0, "step": 3417 }, { "epoch": 0.43480473222236354, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3913872241973877, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8625103235244751, "num_tokens": 129848172.0, "step": 3418 }, { "epoch": 0.43493194250095407, "ewc_loss": 0.047607421875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.62341046333313, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8690405488014221, "num_tokens": 129878728.0, "step": 3419 }, { "epoch": 0.4350591527795446, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3864824771881104, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8527908325195312, "num_tokens": 129916843.0, "step": 3420 }, { "epoch": 0.4351863630581351, "ewc_loss": 0.047607421875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6198575496673584, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8443409204483032, "num_tokens": 129954614.0, "step": 3421 }, { "epoch": 0.4353135733367256, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.399095296859741, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.836093544960022, "num_tokens": 129993084.0, "step": 3422 }, { "epoch": 0.43544078361531613, "ewc_loss": 0.047607421875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6091747283935547, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8620120286941528, "num_tokens": 130028313.0, "step": 3423 }, { "epoch": 0.4355679938939066, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.391845703125, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8524876832962036, "num_tokens": 130066722.0, "step": 3424 }, { "epoch": 0.43569520417249713, "ewc_loss": 0.047119140625, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.593346357345581, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8515864610671997, "num_tokens": 130108628.0, "step": 3425 }, { "epoch": 0.43582241445108766, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.381756544113159, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8678135871887207, "num_tokens": 130145770.0, "step": 3426 }, { "epoch": 0.43594962472967813, "ewc_loss": 0.046875, "ewc_loss_diag": 9.72747802734375e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5874805450439453, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8546930551528931, "num_tokens": 130191539.0, "step": 3427 }, { "epoch": 0.43607683500826866, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.387681722640991, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.866517186164856, "num_tokens": 130229358.0, "step": 3428 }, { "epoch": 0.4362040452868592, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5800256729125977, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8553590178489685, "num_tokens": 130272173.0, "step": 3429 }, { "epoch": 0.4363312555654497, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.386805772781372, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.852576732635498, "num_tokens": 130313135.0, "step": 3430 }, { "epoch": 0.4364584658440402, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5795767307281494, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8650738000869751, "num_tokens": 130350595.0, "step": 3431 }, { "epoch": 0.4365856761226307, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3857693672180176, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8618051409721375, "num_tokens": 130390194.0, "step": 3432 }, { "epoch": 0.43671288640122125, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.585172176361084, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8571454286575317, "num_tokens": 130424768.0, "step": 3433 }, { "epoch": 0.4368400966798117, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3920602798461914, "learning_rate": 1e-05, "loss": 0.5298, "mean_token_accuracy": 0.8410266637802124, "num_tokens": 130469457.0, "step": 3434 }, { "epoch": 0.43696730695840225, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.575650691986084, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8704727292060852, "num_tokens": 130506250.0, "step": 3435 }, { "epoch": 0.4370945172369928, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.400681257247925, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8558024764060974, "num_tokens": 130538605.0, "step": 3436 }, { "epoch": 0.43722172751558325, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5554251670837402, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8698924779891968, "num_tokens": 130577172.0, "step": 3437 }, { "epoch": 0.4373489377941738, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3962504863739014, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8535189628601074, "num_tokens": 130613067.0, "step": 3438 }, { "epoch": 0.4374761480727643, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5434629917144775, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8729690909385681, "num_tokens": 130649176.0, "step": 3439 }, { "epoch": 0.4376033583513548, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.438687562942505, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8581580519676208, "num_tokens": 130685904.0, "step": 3440 }, { "epoch": 0.4377305686299453, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.515357494354248, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8691328763961792, "num_tokens": 130727502.0, "step": 3441 }, { "epoch": 0.43785777890853583, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.421231985092163, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8466306924819946, "num_tokens": 130765038.0, "step": 3442 }, { "epoch": 0.4379849891871263, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.484037399291992, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8618294596672058, "num_tokens": 130810349.0, "step": 3443 }, { "epoch": 0.43811219946571683, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4385223388671875, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8488262295722961, "num_tokens": 130843383.0, "step": 3444 }, { "epoch": 0.43823940974430736, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4751269817352295, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8510518074035645, "num_tokens": 130880399.0, "step": 3445 }, { "epoch": 0.43836662002289783, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.431548595428467, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8397956490516663, "num_tokens": 130915537.0, "step": 3446 }, { "epoch": 0.43849383030148836, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4779350757598877, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.864963710308075, "num_tokens": 130948560.0, "step": 3447 }, { "epoch": 0.4386210405800789, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.41532039642334, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.863591194152832, "num_tokens": 130982405.0, "step": 3448 }, { "epoch": 0.43874825085866936, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4590108394622803, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.86122727394104, "num_tokens": 131022673.0, "step": 3449 }, { "epoch": 0.4388754611372599, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.426018714904785, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8534532785415649, "num_tokens": 131054647.0, "step": 3450 }, { "epoch": 0.4390026714158504, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4542417526245117, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8617657423019409, "num_tokens": 131091616.0, "step": 3451 }, { "epoch": 0.4391298816944409, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.432373046875, "learning_rate": 1e-05, "loss": 0.5533, "mean_token_accuracy": 0.8306373357772827, "num_tokens": 131129663.0, "step": 3452 }, { "epoch": 0.4392570919730314, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4439878463745117, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8597808480262756, "num_tokens": 131170858.0, "step": 3453 }, { "epoch": 0.43938430225162195, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4266774654388428, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8544104099273682, "num_tokens": 131210892.0, "step": 3454 }, { "epoch": 0.4395115125302124, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.461921215057373, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8542073369026184, "num_tokens": 131246616.0, "step": 3455 }, { "epoch": 0.43963872280880295, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4194886684417725, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8587844371795654, "num_tokens": 131286515.0, "step": 3456 }, { "epoch": 0.4397659330873935, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.465311050415039, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8544729351997375, "num_tokens": 131321715.0, "step": 3457 }, { "epoch": 0.43989314336598395, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.4188942909240723, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8470489382743835, "num_tokens": 131359235.0, "step": 3458 }, { "epoch": 0.4400203536445745, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4762325286865234, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8750079870223999, "num_tokens": 131390551.0, "step": 3459 }, { "epoch": 0.440147563923165, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4131274223327637, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8595271706581116, "num_tokens": 131424107.0, "step": 3460 }, { "epoch": 0.4402747742017555, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4680538177490234, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8631772994995117, "num_tokens": 131459797.0, "step": 3461 }, { "epoch": 0.440401984480346, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.4091968536376953, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8556587100028992, "num_tokens": 131496877.0, "step": 3462 }, { "epoch": 0.44052919475893654, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.480114698410034, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8648593425750732, "num_tokens": 131528958.0, "step": 3463 }, { "epoch": 0.440656405037527, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3953778743743896, "learning_rate": 1e-05, "loss": 0.5371, "mean_token_accuracy": 0.8364192843437195, "num_tokens": 131572157.0, "step": 3464 }, { "epoch": 0.44078361531611754, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4917452335357666, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8493673801422119, "num_tokens": 131613855.0, "step": 3465 }, { "epoch": 0.44091082559470807, "ewc_loss": 0.044189453125, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.378394365310669, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8541837930679321, "num_tokens": 131654056.0, "step": 3466 }, { "epoch": 0.44103803587329854, "ewc_loss": 0.045654296875, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.509305477142334, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8351095914840698, "num_tokens": 131694865.0, "step": 3467 }, { "epoch": 0.44116524615188907, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3767385482788086, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8566429615020752, "num_tokens": 131732883.0, "step": 3468 }, { "epoch": 0.4412924564304796, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.540235996246338, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8496991395950317, "num_tokens": 131770384.0, "step": 3469 }, { "epoch": 0.44141966670907007, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.822845458984375e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.373148202896118, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8513068556785583, "num_tokens": 131801611.0, "step": 3470 }, { "epoch": 0.4415468769876606, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5703718662261963, "learning_rate": 1e-05, "loss": 0.5833, "mean_token_accuracy": 0.8312239646911621, "num_tokens": 131842522.0, "step": 3471 }, { "epoch": 0.4416740872662511, "ewc_loss": 0.0439453125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.409385681152344e-05, "grad_norm": 2.3510210514068604, "learning_rate": 1e-05, "loss": 0.542, "mean_token_accuracy": 0.8411144018173218, "num_tokens": 131880041.0, "step": 3472 }, { "epoch": 0.4418012975448416, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.590237617492676, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8586398363113403, "num_tokens": 131917682.0, "step": 3473 }, { "epoch": 0.4419285078234321, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3263626098632812, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8530492782592773, "num_tokens": 131958149.0, "step": 3474 }, { "epoch": 0.44205571810202265, "ewc_loss": 0.04736328125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.646876573562622, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8423737287521362, "num_tokens": 131996648.0, "step": 3475 }, { "epoch": 0.4421829283806131, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3244423866271973, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8573290705680847, "num_tokens": 132036524.0, "step": 3476 }, { "epoch": 0.44231013865920366, "ewc_loss": 0.04931640625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6805567741394043, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8569420576095581, "num_tokens": 132077474.0, "step": 3477 }, { "epoch": 0.4424373489377942, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3379712104797363, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8400665521621704, "num_tokens": 132117153.0, "step": 3478 }, { "epoch": 0.44256455921638466, "ewc_loss": 0.05029296875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.7107038497924805, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8685956001281738, "num_tokens": 132156832.0, "step": 3479 }, { "epoch": 0.4426917694949752, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3686392307281494, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8571898937225342, "num_tokens": 132189245.0, "step": 3480 }, { "epoch": 0.4428189797735657, "ewc_loss": 0.05078125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7182257175445557, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8545706272125244, "num_tokens": 132230572.0, "step": 3481 }, { "epoch": 0.44294619005215624, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3817152976989746, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.854712188243866, "num_tokens": 132271850.0, "step": 3482 }, { "epoch": 0.4430734003307467, "ewc_loss": 0.05078125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7268924713134766, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8596866726875305, "num_tokens": 132312461.0, "step": 3483 }, { "epoch": 0.44320061060933724, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.398911476135254, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8617932200431824, "num_tokens": 132350628.0, "step": 3484 }, { "epoch": 0.44332782088792777, "ewc_loss": 0.05126953125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7188079357147217, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8534590005874634, "num_tokens": 132394040.0, "step": 3485 }, { "epoch": 0.44345503116651824, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4214088916778564, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8557447791099548, "num_tokens": 132435727.0, "step": 3486 }, { "epoch": 0.44358224144510877, "ewc_loss": 0.05029296875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.7267653942108154, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8524593710899353, "num_tokens": 132475769.0, "step": 3487 }, { "epoch": 0.4437094517236993, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.426239013671875, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8526477813720703, "num_tokens": 132520944.0, "step": 3488 }, { "epoch": 0.4438366620022898, "ewc_loss": 0.04931640625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6966044902801514, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8537156581878662, "num_tokens": 132558138.0, "step": 3489 }, { "epoch": 0.4439638722808803, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.449664831161499, "learning_rate": 1e-05, "loss": 0.5496, "mean_token_accuracy": 0.8304173946380615, "num_tokens": 132594937.0, "step": 3490 }, { "epoch": 0.44409108255947083, "ewc_loss": 0.048828125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6751952171325684, "learning_rate": 1e-05, "loss": 0.5399, "mean_token_accuracy": 0.8374966979026794, "num_tokens": 132629168.0, "step": 3491 }, { "epoch": 0.4442182928380613, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4991235733032227, "learning_rate": 1e-05, "loss": 0.5696, "mean_token_accuracy": 0.8278905153274536, "num_tokens": 132673239.0, "step": 3492 }, { "epoch": 0.44434550311665183, "ewc_loss": 0.0478515625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6007087230682373, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8624340295791626, "num_tokens": 132707429.0, "step": 3493 }, { "epoch": 0.44447271339524236, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4343085289001465, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8572548031806946, "num_tokens": 132745592.0, "step": 3494 }, { "epoch": 0.44459992367383283, "ewc_loss": 0.046875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.582069158554077, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8632619976997375, "num_tokens": 132778260.0, "step": 3495 }, { "epoch": 0.44472713395242336, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4189155101776123, "learning_rate": 1e-05, "loss": 0.3792, "mean_token_accuracy": 0.8843614459037781, "num_tokens": 132813156.0, "step": 3496 }, { "epoch": 0.4448543442310139, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.583773612976074, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8527320623397827, "num_tokens": 132847654.0, "step": 3497 }, { "epoch": 0.44498155450960436, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.40226411819458, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8632285594940186, "num_tokens": 132889258.0, "step": 3498 }, { "epoch": 0.4451087647881949, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.6185545921325684, "learning_rate": 1e-05, "loss": 0.5449, "mean_token_accuracy": 0.8341959714889526, "num_tokens": 132919962.0, "step": 3499 }, { "epoch": 0.4452359750667854, "ewc_loss": 0.04443359375, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3924734592437744, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8483083844184875, "num_tokens": 132957333.0, "step": 3500 }, { "epoch": 0.4453631853453759, "ewc_loss": 0.046875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6171257495880127, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8545763492584229, "num_tokens": 132990645.0, "step": 3501 }, { "epoch": 0.4454903956239664, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3807361125946045, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8663849830627441, "num_tokens": 133026234.0, "step": 3502 }, { "epoch": 0.44561760590255695, "ewc_loss": 0.04736328125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6215322017669678, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8573606014251709, "num_tokens": 133062833.0, "step": 3503 }, { "epoch": 0.4457448161811474, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3866450786590576, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8684558868408203, "num_tokens": 133097868.0, "step": 3504 }, { "epoch": 0.44587202645973795, "ewc_loss": 0.04736328125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.608304738998413, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8620849847793579, "num_tokens": 133142062.0, "step": 3505 }, { "epoch": 0.4459992367383285, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3893649578094482, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8509821891784668, "num_tokens": 133181415.0, "step": 3506 }, { "epoch": 0.44612644701691895, "ewc_loss": 0.04736328125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.614788055419922, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.857523500919342, "num_tokens": 133217849.0, "step": 3507 }, { "epoch": 0.4462536572955095, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.384181499481201, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8540908098220825, "num_tokens": 133257639.0, "step": 3508 }, { "epoch": 0.4463808675741, "ewc_loss": 0.04736328125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.608975410461426, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8509752154350281, "num_tokens": 133299692.0, "step": 3509 }, { "epoch": 0.4465080778526905, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.410815477371216, "learning_rate": 1e-05, "loss": 0.5281, "mean_token_accuracy": 0.8394637703895569, "num_tokens": 133339315.0, "step": 3510 }, { "epoch": 0.446635288131281, "ewc_loss": 0.04736328125, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.687502145767212, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8599411845207214, "num_tokens": 133378477.0, "step": 3511 }, { "epoch": 0.44676249840987153, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3975603580474854, "learning_rate": 1e-05, "loss": 0.5585, "mean_token_accuracy": 0.8305084705352783, "num_tokens": 133416041.0, "step": 3512 }, { "epoch": 0.446889708688462, "ewc_loss": 0.0478515625, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6203129291534424, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.859969973564148, "num_tokens": 133458201.0, "step": 3513 }, { "epoch": 0.44701691896705253, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.870529174804688e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.389918088912964, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8575443625450134, "num_tokens": 133495645.0, "step": 3514 }, { "epoch": 0.44714412924564306, "ewc_loss": 0.04833984375, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6709673404693604, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8498685359954834, "num_tokens": 133535385.0, "step": 3515 }, { "epoch": 0.44727133952423354, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.918212890625e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.375101327896118, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8603491187095642, "num_tokens": 133575093.0, "step": 3516 }, { "epoch": 0.44739854980282406, "ewc_loss": 0.04833984375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6614460945129395, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.84687340259552, "num_tokens": 133610447.0, "step": 3517 }, { "epoch": 0.4475257600814146, "ewc_loss": 0.04541015625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.389176368713379, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8644999861717224, "num_tokens": 133643506.0, "step": 3518 }, { "epoch": 0.44765297036000506, "ewc_loss": 0.048583984375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.634979009628296, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8766070604324341, "num_tokens": 133683081.0, "step": 3519 }, { "epoch": 0.4477801806385956, "ewc_loss": 0.045654296875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.529193639755249, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8467615842819214, "num_tokens": 133721836.0, "step": 3520 }, { "epoch": 0.4479073909171861, "ewc_loss": 0.046630859375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.556023597717285, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8611727952957153, "num_tokens": 133756243.0, "step": 3521 }, { "epoch": 0.4480346011957766, "ewc_loss": 0.045654296875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4192614555358887, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8443418741226196, "num_tokens": 133807126.0, "step": 3522 }, { "epoch": 0.4481618114743671, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5339980125427246, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8527573347091675, "num_tokens": 133842036.0, "step": 3523 }, { "epoch": 0.44828902175295765, "ewc_loss": 0.045654296875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.433835744857788, "learning_rate": 1e-05, "loss": 0.5276, "mean_token_accuracy": 0.8408757448196411, "num_tokens": 133881948.0, "step": 3524 }, { "epoch": 0.4484162320315481, "ewc_loss": 0.045654296875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.518979072570801, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8638569712638855, "num_tokens": 133917765.0, "step": 3525 }, { "epoch": 0.44854344231013865, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4104533195495605, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8528587818145752, "num_tokens": 133959199.0, "step": 3526 }, { "epoch": 0.4486706525887292, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5389597415924072, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8430348038673401, "num_tokens": 133991484.0, "step": 3527 }, { "epoch": 0.44879786286731965, "ewc_loss": 0.045166015625, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4125735759735107, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8657926321029663, "num_tokens": 134025817.0, "step": 3528 }, { "epoch": 0.4489250731459102, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.520616054534912, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8552349805831909, "num_tokens": 134063956.0, "step": 3529 }, { "epoch": 0.4490522834245007, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3977818489074707, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8541052341461182, "num_tokens": 134101959.0, "step": 3530 }, { "epoch": 0.44917949370309124, "ewc_loss": 0.0458984375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5278854370117188, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8511700630187988, "num_tokens": 134139961.0, "step": 3531 }, { "epoch": 0.4493067039816817, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.382969856262207, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8487902283668518, "num_tokens": 134180368.0, "step": 3532 }, { "epoch": 0.44943391426027224, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.53131103515625, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8656818866729736, "num_tokens": 134216842.0, "step": 3533 }, { "epoch": 0.44956112453886277, "ewc_loss": 0.044921875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.377319574356079, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8594218492507935, "num_tokens": 134254089.0, "step": 3534 }, { "epoch": 0.44968833481745324, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5456128120422363, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8620900511741638, "num_tokens": 134289559.0, "step": 3535 }, { "epoch": 0.44981554509604377, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3849949836730957, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8568228483200073, "num_tokens": 134323029.0, "step": 3536 }, { "epoch": 0.4499427553746343, "ewc_loss": 0.046142578125, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5504045486450195, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8537169098854065, "num_tokens": 134362657.0, "step": 3537 }, { "epoch": 0.45006996565322477, "ewc_loss": 0.044677734375, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.359635591506958, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8596527576446533, "num_tokens": 134403555.0, "step": 3538 }, { "epoch": 0.4501971759318153, "ewc_loss": 0.04638671875, "ewc_loss_diag": 9.965896606445312e-05, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.574620246887207, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8604261875152588, "num_tokens": 134438780.0, "step": 3539 }, { "epoch": 0.4503243862104058, "ewc_loss": 0.044677734375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.358480930328369, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8505554795265198, "num_tokens": 134477331.0, "step": 3540 }, { "epoch": 0.4504515964889963, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5927035808563232, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8687866926193237, "num_tokens": 134513898.0, "step": 3541 }, { "epoch": 0.4505788067675868, "ewc_loss": 0.044677734375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.369652271270752, "learning_rate": 1e-05, "loss": 0.5396, "mean_token_accuracy": 0.8410885334014893, "num_tokens": 134548507.0, "step": 3542 }, { "epoch": 0.45070601704617735, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.603626251220703, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8500024676322937, "num_tokens": 134590243.0, "step": 3543 }, { "epoch": 0.4508332273247678, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.363813638687134, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8577585220336914, "num_tokens": 134624693.0, "step": 3544 }, { "epoch": 0.45096043760335836, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.603512763977051, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8773186206817627, "num_tokens": 134661365.0, "step": 3545 }, { "epoch": 0.4510876478819489, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3702943325042725, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8594093322753906, "num_tokens": 134698218.0, "step": 3546 }, { "epoch": 0.45121485816053936, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6148760318756104, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8499526381492615, "num_tokens": 134736011.0, "step": 3547 }, { "epoch": 0.4513420684391299, "ewc_loss": 0.045166015625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.366361379623413, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8592315912246704, "num_tokens": 134774457.0, "step": 3548 }, { "epoch": 0.4514692787177204, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6481621265411377, "learning_rate": 1e-05, "loss": 0.5411, "mean_token_accuracy": 0.8347442150115967, "num_tokens": 134813788.0, "step": 3549 }, { "epoch": 0.4515964889963109, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3724114894866943, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.844592809677124, "num_tokens": 134851916.0, "step": 3550 }, { "epoch": 0.4517236992749014, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6352217197418213, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8614454865455627, "num_tokens": 134892655.0, "step": 3551 }, { "epoch": 0.45185090955349194, "ewc_loss": 0.045166015625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.367586851119995, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8542025089263916, "num_tokens": 134931391.0, "step": 3552 }, { "epoch": 0.4519781198320824, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6622464656829834, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8509812951087952, "num_tokens": 134966285.0, "step": 3553 }, { "epoch": 0.45210533011067294, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3800277709960938, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8572620153427124, "num_tokens": 135003893.0, "step": 3554 }, { "epoch": 0.45223254038926347, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.653256416320801, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8571841716766357, "num_tokens": 135043245.0, "step": 3555 }, { "epoch": 0.45235975066785394, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.388239860534668, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8527406454086304, "num_tokens": 135082867.0, "step": 3556 }, { "epoch": 0.45248696094644447, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.646350145339966, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8778349161148071, "num_tokens": 135121153.0, "step": 3557 }, { "epoch": 0.452614171225035, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4137468338012695, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8523614406585693, "num_tokens": 135154221.0, "step": 3558 }, { "epoch": 0.4527413815036255, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6260805130004883, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8585480451583862, "num_tokens": 135196065.0, "step": 3559 }, { "epoch": 0.452868591782216, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4013664722442627, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8654009699821472, "num_tokens": 135233692.0, "step": 3560 }, { "epoch": 0.45299580206080653, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.622891664505005, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8720346689224243, "num_tokens": 135267538.0, "step": 3561 }, { "epoch": 0.453123012339397, "ewc_loss": 0.045166015625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.402890920639038, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.865469217300415, "num_tokens": 135303946.0, "step": 3562 }, { "epoch": 0.45325022261798753, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.619946002960205, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8465242385864258, "num_tokens": 135340940.0, "step": 3563 }, { "epoch": 0.45337743289657806, "ewc_loss": 0.045166015625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4070253372192383, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8466137647628784, "num_tokens": 135377472.0, "step": 3564 }, { "epoch": 0.45350464317516853, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.623351573944092, "learning_rate": 1e-05, "loss": 0.5312, "mean_token_accuracy": 0.8367824554443359, "num_tokens": 135411396.0, "step": 3565 }, { "epoch": 0.45363185345375906, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.403686761856079, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8467286229133606, "num_tokens": 135445323.0, "step": 3566 }, { "epoch": 0.4537590637323496, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5933828353881836, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8603520393371582, "num_tokens": 135487364.0, "step": 3567 }, { "epoch": 0.45388627401094006, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.00010061264038085938, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.386000871658325, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8563617467880249, "num_tokens": 135529879.0, "step": 3568 }, { "epoch": 0.4540134842895306, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5926265716552734, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8609768152236938, "num_tokens": 135568812.0, "step": 3569 }, { "epoch": 0.4541406945681211, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.389070510864258, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8517045378684998, "num_tokens": 135608096.0, "step": 3570 }, { "epoch": 0.4542679048467116, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6093013286590576, "learning_rate": 1e-05, "loss": 0.5645, "mean_token_accuracy": 0.8311062455177307, "num_tokens": 135645796.0, "step": 3571 }, { "epoch": 0.4543951151253021, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3811497688293457, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8539718389511108, "num_tokens": 135686535.0, "step": 3572 }, { "epoch": 0.45452232540389265, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6225874423980713, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8562202453613281, "num_tokens": 135719555.0, "step": 3573 }, { "epoch": 0.4546495356824831, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.375631809234619, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8677325248718262, "num_tokens": 135756498.0, "step": 3574 }, { "epoch": 0.45477674596107365, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.628650188446045, "learning_rate": 1e-05, "loss": 0.5424, "mean_token_accuracy": 0.8378129601478577, "num_tokens": 135797438.0, "step": 3575 }, { "epoch": 0.4549039562396642, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.377009630203247, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8512604236602783, "num_tokens": 135836085.0, "step": 3576 }, { "epoch": 0.45503116651825465, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6298887729644775, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8616636991500854, "num_tokens": 135872272.0, "step": 3577 }, { "epoch": 0.4551583767968452, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4054203033447266, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8373196125030518, "num_tokens": 135908667.0, "step": 3578 }, { "epoch": 0.4552855870754357, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.628481388092041, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8663464784622192, "num_tokens": 135939205.0, "step": 3579 }, { "epoch": 0.4554127973540262, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.396045684814453, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8593841791152954, "num_tokens": 135977084.0, "step": 3580 }, { "epoch": 0.4555400076326167, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.609595537185669, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8494336605072021, "num_tokens": 136012712.0, "step": 3581 }, { "epoch": 0.45566721791120723, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3899054527282715, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8605670928955078, "num_tokens": 136050253.0, "step": 3582 }, { "epoch": 0.45579442818979776, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6088850498199463, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8544787764549255, "num_tokens": 136088072.0, "step": 3583 }, { "epoch": 0.45592163846838824, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4053497314453125, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8510253429412842, "num_tokens": 136122434.0, "step": 3584 }, { "epoch": 0.45604884874697876, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5896811485290527, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8639187812805176, "num_tokens": 136164489.0, "step": 3585 }, { "epoch": 0.4561760590255693, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.406205177307129, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8527111411094666, "num_tokens": 136198534.0, "step": 3586 }, { "epoch": 0.45630326930415976, "ewc_loss": 0.046875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5659689903259277, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8573272824287415, "num_tokens": 136242261.0, "step": 3587 }, { "epoch": 0.4564304795827503, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4132602214813232, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.847006618976593, "num_tokens": 136277106.0, "step": 3588 }, { "epoch": 0.4565576898613408, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.55666446685791, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8587295413017273, "num_tokens": 136315872.0, "step": 3589 }, { "epoch": 0.4566849001399313, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.6288158893585205, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8563985824584961, "num_tokens": 136359513.0, "step": 3590 }, { "epoch": 0.4568121104185218, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4368913173675537, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.850354790687561, "num_tokens": 136398575.0, "step": 3591 }, { "epoch": 0.45693932069711235, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.0001010894775390625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.493454694747925, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8670822381973267, "num_tokens": 136439062.0, "step": 3592 }, { "epoch": 0.4570665309757028, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.4098398685455322, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8543616533279419, "num_tokens": 136475044.0, "step": 3593 }, { "epoch": 0.45719374125429335, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.522270679473877, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8450928330421448, "num_tokens": 136513674.0, "step": 3594 }, { "epoch": 0.4573209515328839, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.390770196914673, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8601194620132446, "num_tokens": 136551218.0, "step": 3595 }, { "epoch": 0.45744816181147435, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5265042781829834, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8697519302368164, "num_tokens": 136589088.0, "step": 3596 }, { "epoch": 0.4575753720900649, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3856372833251953, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8580372929573059, "num_tokens": 136625269.0, "step": 3597 }, { "epoch": 0.4577025823686554, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5375678539276123, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8605650663375854, "num_tokens": 136660212.0, "step": 3598 }, { "epoch": 0.4578297926472459, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3644328117370605, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8495017290115356, "num_tokens": 136703372.0, "step": 3599 }, { "epoch": 0.4579570029258364, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5603556632995605, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8508022427558899, "num_tokens": 136741241.0, "step": 3600 }, { "epoch": 0.45808421320442694, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3515541553497314, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8564499616622925, "num_tokens": 136780329.0, "step": 3601 }, { "epoch": 0.4582114234830174, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6133997440338135, "learning_rate": 1e-05, "loss": 0.5278, "mean_token_accuracy": 0.8413301706314087, "num_tokens": 136812567.0, "step": 3602 }, { "epoch": 0.45833863376160794, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.341205358505249, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8472913503646851, "num_tokens": 136854370.0, "step": 3603 }, { "epoch": 0.45846584404019847, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.648200750350952, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8545143008232117, "num_tokens": 136890536.0, "step": 3604 }, { "epoch": 0.45859305431878894, "ewc_loss": 0.04443359375, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.3446996212005615, "learning_rate": 1e-05, "loss": 0.5639, "mean_token_accuracy": 0.8295553922653198, "num_tokens": 136929376.0, "step": 3605 }, { "epoch": 0.45872026459737947, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6697425842285156, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8623394966125488, "num_tokens": 136967857.0, "step": 3606 }, { "epoch": 0.45884747487597, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.00010156631469726562, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.346107006072998, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8555270433425903, "num_tokens": 137007853.0, "step": 3607 }, { "epoch": 0.45897468515456047, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.688743829727173, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.858166515827179, "num_tokens": 137048172.0, "step": 3608 }, { "epoch": 0.459101895433151, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.374305248260498, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8391110301017761, "num_tokens": 137083598.0, "step": 3609 }, { "epoch": 0.4592291057117415, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7068142890930176, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8464639186859131, "num_tokens": 137117626.0, "step": 3610 }, { "epoch": 0.459356315990332, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3823893070220947, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8464465737342834, "num_tokens": 137161536.0, "step": 3611 }, { "epoch": 0.4594835262689225, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.710817337036133, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8633555173873901, "num_tokens": 137195932.0, "step": 3612 }, { "epoch": 0.45961073654751305, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4101061820983887, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8651362657546997, "num_tokens": 137229263.0, "step": 3613 }, { "epoch": 0.4597379468261035, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.702127456665039, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8560861349105835, "num_tokens": 137265250.0, "step": 3614 }, { "epoch": 0.45986515710469406, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.407212018966675, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8677119612693787, "num_tokens": 137307612.0, "step": 3615 }, { "epoch": 0.4599923673832846, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.709074020385742, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8687742352485657, "num_tokens": 137336075.0, "step": 3616 }, { "epoch": 0.46011957766187506, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4348702430725098, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8546895980834961, "num_tokens": 137374571.0, "step": 3617 }, { "epoch": 0.4602467879404656, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.691523551940918, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8485891819000244, "num_tokens": 137409275.0, "step": 3618 }, { "epoch": 0.4603739982190561, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.427107334136963, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8584922552108765, "num_tokens": 137447665.0, "step": 3619 }, { "epoch": 0.4605012084976466, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6822521686553955, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8649486303329468, "num_tokens": 137484256.0, "step": 3620 }, { "epoch": 0.4606284187762371, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.427351951599121, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8584977388381958, "num_tokens": 137523868.0, "step": 3621 }, { "epoch": 0.46075562905482764, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00010204315185546875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.647707223892212, "learning_rate": 1e-05, "loss": 0.5504, "mean_token_accuracy": 0.8315741419792175, "num_tokens": 137564707.0, "step": 3622 }, { "epoch": 0.4608828393334181, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.433607816696167, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8634982109069824, "num_tokens": 137600347.0, "step": 3623 }, { "epoch": 0.46101004961200864, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6202337741851807, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8498780727386475, "num_tokens": 137642874.0, "step": 3624 }, { "epoch": 0.46113725989059917, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4229040145874023, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.849433422088623, "num_tokens": 137685253.0, "step": 3625 }, { "epoch": 0.46126447016918964, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6515614986419678, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8732223510742188, "num_tokens": 137722180.0, "step": 3626 }, { "epoch": 0.4613916804477802, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4256467819213867, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8624268174171448, "num_tokens": 137756665.0, "step": 3627 }, { "epoch": 0.4615188907263707, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6045491695404053, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8630787134170532, "num_tokens": 137795928.0, "step": 3628 }, { "epoch": 0.4616461010049612, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4283270835876465, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8422991037368774, "num_tokens": 137833695.0, "step": 3629 }, { "epoch": 0.4617733112835517, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5804903507232666, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8600941300392151, "num_tokens": 137871894.0, "step": 3630 }, { "epoch": 0.46190052156214223, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4156618118286133, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8503444194793701, "num_tokens": 137907818.0, "step": 3631 }, { "epoch": 0.46202773184073276, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.56648850440979, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8690795302391052, "num_tokens": 137943506.0, "step": 3632 }, { "epoch": 0.46215494211932323, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010251998901367188, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.440452814102173, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8493815660476685, "num_tokens": 137975113.0, "step": 3633 }, { "epoch": 0.46228215239791376, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.566718339920044, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8628249168395996, "num_tokens": 138006503.0, "step": 3634 }, { "epoch": 0.4624093626765043, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.406170129776001, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8650014400482178, "num_tokens": 138046667.0, "step": 3635 }, { "epoch": 0.46253657295509476, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5532336235046387, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8443769216537476, "num_tokens": 138084821.0, "step": 3636 }, { "epoch": 0.4626637832336853, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3992626667022705, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8552085161209106, "num_tokens": 138123994.0, "step": 3637 }, { "epoch": 0.4627909935122758, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5588181018829346, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8527127504348755, "num_tokens": 138161225.0, "step": 3638 }, { "epoch": 0.4629182037908663, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3929240703582764, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8644282817840576, "num_tokens": 138197441.0, "step": 3639 }, { "epoch": 0.4630454140694568, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5666186809539795, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8589648008346558, "num_tokens": 138229663.0, "step": 3640 }, { "epoch": 0.46317262434804735, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.388734817504883, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8540422916412354, "num_tokens": 138267955.0, "step": 3641 }, { "epoch": 0.4632998346266378, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00010347366333007812, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5649044513702393, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8679815530776978, "num_tokens": 138303968.0, "step": 3642 }, { "epoch": 0.46342704490522835, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.377709150314331, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8569539785385132, "num_tokens": 138344667.0, "step": 3643 }, { "epoch": 0.4635542551838189, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5600035190582275, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8677929043769836, "num_tokens": 138387513.0, "step": 3644 }, { "epoch": 0.46368146546240935, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3667659759521484, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8738031387329102, "num_tokens": 138428070.0, "step": 3645 }, { "epoch": 0.4638086757409999, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6221325397491455, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8597930669784546, "num_tokens": 138462303.0, "step": 3646 }, { "epoch": 0.4639358860195904, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.366588830947876, "learning_rate": 1e-05, "loss": 0.546, "mean_token_accuracy": 0.8368515968322754, "num_tokens": 138502099.0, "step": 3647 }, { "epoch": 0.4640630962981809, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6231234073638916, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8492701053619385, "num_tokens": 138539400.0, "step": 3648 }, { "epoch": 0.4641903065767714, "ewc_loss": 0.045166015625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3684959411621094, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8575763702392578, "num_tokens": 138573082.0, "step": 3649 }, { "epoch": 0.46431751685536193, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.624002456665039, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8620089292526245, "num_tokens": 138613693.0, "step": 3650 }, { "epoch": 0.4644447271339524, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010395050048828125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3585402965545654, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8713055849075317, "num_tokens": 138655138.0, "step": 3651 }, { "epoch": 0.46457193741254293, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6553823947906494, "learning_rate": 1e-05, "loss": 0.5746, "mean_token_accuracy": 0.8274880647659302, "num_tokens": 138697117.0, "step": 3652 }, { "epoch": 0.46469914769113346, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.377030849456787, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8534305691719055, "num_tokens": 138732723.0, "step": 3653 }, { "epoch": 0.46482635796972394, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6519277095794678, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.85927414894104, "num_tokens": 138767680.0, "step": 3654 }, { "epoch": 0.46495356824831446, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.383686065673828, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8643897175788879, "num_tokens": 138803119.0, "step": 3655 }, { "epoch": 0.465080778526905, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6505720615386963, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8633650541305542, "num_tokens": 138845032.0, "step": 3656 }, { "epoch": 0.46520798880549546, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3851966857910156, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8631387948989868, "num_tokens": 138886521.0, "step": 3657 }, { "epoch": 0.465335199084086, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6423001289367676, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8578214645385742, "num_tokens": 138927586.0, "step": 3658 }, { "epoch": 0.4654624093626765, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.41131591796875, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.853218674659729, "num_tokens": 138967382.0, "step": 3659 }, { "epoch": 0.465589619641267, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.637760639190674, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8408850431442261, "num_tokens": 139003498.0, "step": 3660 }, { "epoch": 0.4657168299198575, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4006526470184326, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8464361429214478, "num_tokens": 139046724.0, "step": 3661 }, { "epoch": 0.46584404019844805, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.612025737762451, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8540680408477783, "num_tokens": 139091513.0, "step": 3662 }, { "epoch": 0.4659712504770385, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.423853874206543, "learning_rate": 1e-05, "loss": 0.548, "mean_token_accuracy": 0.8341150283813477, "num_tokens": 139130082.0, "step": 3663 }, { "epoch": 0.46609846075562905, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5855884552001953, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.858555018901825, "num_tokens": 139172974.0, "step": 3664 }, { "epoch": 0.4662256710342196, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.410922050476074, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8648463487625122, "num_tokens": 139212164.0, "step": 3665 }, { "epoch": 0.46635288131281005, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.597419023513794, "learning_rate": 1e-05, "loss": 0.5631, "mean_token_accuracy": 0.8280839920043945, "num_tokens": 139252960.0, "step": 3666 }, { "epoch": 0.4664800915914006, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.425477981567383, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8486748933792114, "num_tokens": 139287078.0, "step": 3667 }, { "epoch": 0.4666073018699911, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00010442733764648438, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.623136520385742, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.84613037109375, "num_tokens": 139321545.0, "step": 3668 }, { "epoch": 0.4667345121485816, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.392153739929199, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8393920063972473, "num_tokens": 139360829.0, "step": 3669 }, { "epoch": 0.4668617224271721, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.594330310821533, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.858301043510437, "num_tokens": 139396347.0, "step": 3670 }, { "epoch": 0.46698893270576264, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.38533353805542, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8600300550460815, "num_tokens": 139430367.0, "step": 3671 }, { "epoch": 0.4671161429843531, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.604768753051758, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8479986190795898, "num_tokens": 139468233.0, "step": 3672 }, { "epoch": 0.46724335326294364, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.391382932662964, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.853151798248291, "num_tokens": 139502377.0, "step": 3673 }, { "epoch": 0.46737056354153417, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5994160175323486, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8481730222702026, "num_tokens": 139540503.0, "step": 3674 }, { "epoch": 0.46749777382012464, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3891615867614746, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8434441089630127, "num_tokens": 139578063.0, "step": 3675 }, { "epoch": 0.46762498409871517, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6059134006500244, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8708693981170654, "num_tokens": 139615451.0, "step": 3676 }, { "epoch": 0.4677521943773057, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3849899768829346, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8462690114974976, "num_tokens": 139657779.0, "step": 3677 }, { "epoch": 0.46787940465589617, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.606194496154785, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8589804172515869, "num_tokens": 139698166.0, "step": 3678 }, { "epoch": 0.4680066149344867, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.391150712966919, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8407852649688721, "num_tokens": 139736970.0, "step": 3679 }, { "epoch": 0.4681338252130772, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.61118221282959, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8557750582695007, "num_tokens": 139771570.0, "step": 3680 }, { "epoch": 0.46826103549166775, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3910205364227295, "learning_rate": 1e-05, "loss": 0.5519, "mean_token_accuracy": 0.8347718715667725, "num_tokens": 139811927.0, "step": 3681 }, { "epoch": 0.4683882457702582, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.603970766067505, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8441075682640076, "num_tokens": 139852903.0, "step": 3682 }, { "epoch": 0.46851545604884876, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3803505897521973, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8453273773193359, "num_tokens": 139895375.0, "step": 3683 }, { "epoch": 0.4686426663274393, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.638045310974121, "learning_rate": 1e-05, "loss": 0.5777, "mean_token_accuracy": 0.8335506319999695, "num_tokens": 139927878.0, "step": 3684 }, { "epoch": 0.46876987660602976, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.0001049041748046875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3938493728637695, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8538378477096558, "num_tokens": 139959864.0, "step": 3685 }, { "epoch": 0.4688970868846203, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6086716651916504, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.84342360496521, "num_tokens": 140003815.0, "step": 3686 }, { "epoch": 0.4690242971632108, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3920180797576904, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8399620652198792, "num_tokens": 140043126.0, "step": 3687 }, { "epoch": 0.4691515074418013, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6180617809295654, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8518215417861938, "num_tokens": 140078466.0, "step": 3688 }, { "epoch": 0.4692787177203918, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.394425392150879, "learning_rate": 1e-05, "loss": 0.5031, "mean_token_accuracy": 0.8449043035507202, "num_tokens": 140114733.0, "step": 3689 }, { "epoch": 0.46940592799898234, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.610233783721924, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8520578742027283, "num_tokens": 140157017.0, "step": 3690 }, { "epoch": 0.4695331382775728, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3858561515808105, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8502055406570435, "num_tokens": 140197408.0, "step": 3691 }, { "epoch": 0.46966034855616334, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.630599021911621, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.8474565744400024, "num_tokens": 140232458.0, "step": 3692 }, { "epoch": 0.46978755883475387, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.38846492767334, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8532159328460693, "num_tokens": 140273790.0, "step": 3693 }, { "epoch": 0.46991476911334434, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.609865427017212, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8712635040283203, "num_tokens": 140313339.0, "step": 3694 }, { "epoch": 0.47004197939193487, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3964428901672363, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8530811667442322, "num_tokens": 140352572.0, "step": 3695 }, { "epoch": 0.4701691896705254, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6313295364379883, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8468766808509827, "num_tokens": 140385221.0, "step": 3696 }, { "epoch": 0.4702963999491159, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3819069862365723, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.870240330696106, "num_tokens": 140427041.0, "step": 3697 }, { "epoch": 0.4704236102277064, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6288599967956543, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8706419467926025, "num_tokens": 140460956.0, "step": 3698 }, { "epoch": 0.47055082050629693, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.381028890609741, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.870450496673584, "num_tokens": 140497644.0, "step": 3699 }, { "epoch": 0.4706780307848874, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.636763334274292, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8563323020935059, "num_tokens": 140532764.0, "step": 3700 }, { "epoch": 0.47080524106347793, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4153754711151123, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8551212549209595, "num_tokens": 140564705.0, "step": 3701 }, { "epoch": 0.47093245134206846, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6136233806610107, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.85280442237854, "num_tokens": 140606959.0, "step": 3702 }, { "epoch": 0.47105966162065893, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.39258074760437, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8622596859931946, "num_tokens": 140646034.0, "step": 3703 }, { "epoch": 0.47118687189924946, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6215102672576904, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.854841947555542, "num_tokens": 140679948.0, "step": 3704 }, { "epoch": 0.47131408217784, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.412069320678711, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8469600081443787, "num_tokens": 140713071.0, "step": 3705 }, { "epoch": 0.47144129245643046, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6087090969085693, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8628900051116943, "num_tokens": 140744243.0, "step": 3706 }, { "epoch": 0.471568502735021, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.400834083557129, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8732946515083313, "num_tokens": 140776716.0, "step": 3707 }, { "epoch": 0.4716957130136115, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6132357120513916, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8754774332046509, "num_tokens": 140808874.0, "step": 3708 }, { "epoch": 0.471822923292202, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3872129917144775, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8661496639251709, "num_tokens": 140846198.0, "step": 3709 }, { "epoch": 0.4719501335707925, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5969767570495605, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8592255115509033, "num_tokens": 140889228.0, "step": 3710 }, { "epoch": 0.47207734384938305, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.399754047393799, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8579993844032288, "num_tokens": 140924549.0, "step": 3711 }, { "epoch": 0.4722045541279735, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.604440450668335, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8679728507995605, "num_tokens": 140960995.0, "step": 3712 }, { "epoch": 0.47233176440656405, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 3.1760823726654053, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8581901788711548, "num_tokens": 141000899.0, "step": 3713 }, { "epoch": 0.4724589746851546, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.440246820449829, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8646713495254517, "num_tokens": 141037736.0, "step": 3714 }, { "epoch": 0.47258618496374505, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00010538101196289062, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6826984882354736, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8394098281860352, "num_tokens": 141068916.0, "step": 3715 }, { "epoch": 0.4727133952423356, "ewc_loss": 0.044921875, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 3.4332275390625e-05, "grad_norm": 2.347349166870117, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8393576741218567, "num_tokens": 141102353.0, "step": 3716 }, { "epoch": 0.4728406055209261, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.738743782043457, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8654897809028625, "num_tokens": 141142653.0, "step": 3717 }, { "epoch": 0.4729678157995166, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3527140617370605, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8592227101325989, "num_tokens": 141180091.0, "step": 3718 }, { "epoch": 0.4730950260781071, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7677409648895264, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8729540705680847, "num_tokens": 141218674.0, "step": 3719 }, { "epoch": 0.47322223635669763, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010633468627929688, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.38481068611145, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8678497076034546, "num_tokens": 141258680.0, "step": 3720 }, { "epoch": 0.4733494466352881, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7963433265686035, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8503341674804688, "num_tokens": 141297883.0, "step": 3721 }, { "epoch": 0.47347665691387864, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4411187171936035, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8577980995178223, "num_tokens": 141336549.0, "step": 3722 }, { "epoch": 0.47360386719246916, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7907094955444336, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.857384204864502, "num_tokens": 141368697.0, "step": 3723 }, { "epoch": 0.47373107747105964, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.464233160018921, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8620036244392395, "num_tokens": 141399349.0, "step": 3724 }, { "epoch": 0.47385828774965016, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7484207153320312, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8714846968650818, "num_tokens": 141435198.0, "step": 3725 }, { "epoch": 0.4739854980282407, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00010585784912109375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4792096614837646, "learning_rate": 1e-05, "loss": 0.541, "mean_token_accuracy": 0.8425052165985107, "num_tokens": 141470659.0, "step": 3726 }, { "epoch": 0.47411270830683117, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.709993839263916, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8550723791122437, "num_tokens": 141514774.0, "step": 3727 }, { "epoch": 0.4742399185854217, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.0001068115234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.467604637145996, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8619844317436218, "num_tokens": 141551071.0, "step": 3728 }, { "epoch": 0.4743671288640122, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6712284088134766, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8575572371482849, "num_tokens": 141590201.0, "step": 3729 }, { "epoch": 0.4744943391426027, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4428489208221436, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8645528554916382, "num_tokens": 141637580.0, "step": 3730 }, { "epoch": 0.4746215494211932, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.67427134513855, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8447071313858032, "num_tokens": 141674092.0, "step": 3731 }, { "epoch": 0.47474875969978375, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.466644287109375, "learning_rate": 1e-05, "loss": 0.5333, "mean_token_accuracy": 0.8382343053817749, "num_tokens": 141709416.0, "step": 3732 }, { "epoch": 0.4748759699783743, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.644993543624878, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8560791611671448, "num_tokens": 141742445.0, "step": 3733 }, { "epoch": 0.47500318025696475, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.430128812789917, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8664008378982544, "num_tokens": 141778224.0, "step": 3734 }, { "epoch": 0.4751303905355553, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.647819995880127, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8467638492584229, "num_tokens": 141818239.0, "step": 3735 }, { "epoch": 0.4752576008141458, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4247090816497803, "learning_rate": 1e-05, "loss": 0.5259, "mean_token_accuracy": 0.8424840569496155, "num_tokens": 141853624.0, "step": 3736 }, { "epoch": 0.4753848110927363, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.621328592300415, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8563317060470581, "num_tokens": 141893620.0, "step": 3737 }, { "epoch": 0.4755120213713268, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.411958694458008, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8645293116569519, "num_tokens": 141930504.0, "step": 3738 }, { "epoch": 0.47563923164991734, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6212189197540283, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8484320044517517, "num_tokens": 141968195.0, "step": 3739 }, { "epoch": 0.4757664419285078, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.5863780975341797, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8592478036880493, "num_tokens": 142010143.0, "step": 3740 }, { "epoch": 0.47589365220709834, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.506378412246704, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.845883846282959, "num_tokens": 142051170.0, "step": 3741 }, { "epoch": 0.47602086248568887, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.480159282684326, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.8374345302581787, "num_tokens": 142084327.0, "step": 3742 }, { "epoch": 0.47614807276427934, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.466892957687378, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8470567464828491, "num_tokens": 142125512.0, "step": 3743 }, { "epoch": 0.47627528304286987, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4756007194519043, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8577284812927246, "num_tokens": 142159117.0, "step": 3744 }, { "epoch": 0.4764024933214604, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.462904691696167, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.856927752494812, "num_tokens": 142194580.0, "step": 3745 }, { "epoch": 0.47652970360005087, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4643502235412598, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8715631365776062, "num_tokens": 142230945.0, "step": 3746 }, { "epoch": 0.4766569138786414, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.447960376739502, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8660115599632263, "num_tokens": 142264792.0, "step": 3747 }, { "epoch": 0.4767841241572319, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.45900821685791, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8668088912963867, "num_tokens": 142302210.0, "step": 3748 }, { "epoch": 0.4769113344358224, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.44242262840271, "learning_rate": 1e-05, "loss": 0.5041, "mean_token_accuracy": 0.8490091562271118, "num_tokens": 142345646.0, "step": 3749 }, { "epoch": 0.4770385447144129, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.461726427078247, "learning_rate": 1e-05, "loss": 0.5526, "mean_token_accuracy": 0.8327036499977112, "num_tokens": 142387097.0, "step": 3750 }, { "epoch": 0.47716575499300345, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4521431922912598, "learning_rate": 1e-05, "loss": 0.5398, "mean_token_accuracy": 0.8345566391944885, "num_tokens": 142424087.0, "step": 3751 }, { "epoch": 0.4772929652715939, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4582390785217285, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8660475015640259, "num_tokens": 142458837.0, "step": 3752 }, { "epoch": 0.47742017555018446, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.440807342529297, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8645268678665161, "num_tokens": 142495661.0, "step": 3753 }, { "epoch": 0.477547385828775, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.467374563217163, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8461371660232544, "num_tokens": 142532439.0, "step": 3754 }, { "epoch": 0.47767459610736546, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4301323890686035, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8447682857513428, "num_tokens": 142574032.0, "step": 3755 }, { "epoch": 0.477801806385956, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4657936096191406, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8611970543861389, "num_tokens": 142607438.0, "step": 3756 }, { "epoch": 0.4779290166645465, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.454103708267212, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8449714779853821, "num_tokens": 142642527.0, "step": 3757 }, { "epoch": 0.478056226943137, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.463463306427002, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8567835092544556, "num_tokens": 142677447.0, "step": 3758 }, { "epoch": 0.4781834372217275, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.446470260620117, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8481847643852234, "num_tokens": 142714429.0, "step": 3759 }, { "epoch": 0.47831064750031804, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4482157230377197, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8525568246841431, "num_tokens": 142752742.0, "step": 3760 }, { "epoch": 0.4784378577789085, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4435815811157227, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8570677638053894, "num_tokens": 142788925.0, "step": 3761 }, { "epoch": 0.47856506805749904, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4458391666412354, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8659956455230713, "num_tokens": 142823215.0, "step": 3762 }, { "epoch": 0.47869227833608957, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4437978267669678, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8453162312507629, "num_tokens": 142860346.0, "step": 3763 }, { "epoch": 0.47881948861468004, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4510114192962646, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8547281622886658, "num_tokens": 142895814.0, "step": 3764 }, { "epoch": 0.4789466988932706, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4553604125976562, "learning_rate": 1e-05, "loss": 0.5463, "mean_token_accuracy": 0.833066999912262, "num_tokens": 142934379.0, "step": 3765 }, { "epoch": 0.4790739091718611, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4573850631713867, "learning_rate": 1e-05, "loss": 0.5334, "mean_token_accuracy": 0.8361064791679382, "num_tokens": 142970512.0, "step": 3766 }, { "epoch": 0.4792011194504516, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.448140859603882, "learning_rate": 1e-05, "loss": 0.518, "mean_token_accuracy": 0.8416245579719543, "num_tokens": 143008635.0, "step": 3767 }, { "epoch": 0.4793283297290421, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.441234588623047, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8623121976852417, "num_tokens": 143047638.0, "step": 3768 }, { "epoch": 0.47945554000763263, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.456479787826538, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8480867147445679, "num_tokens": 143082470.0, "step": 3769 }, { "epoch": 0.4795827502862231, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.423835277557373, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8628771901130676, "num_tokens": 143126869.0, "step": 3770 }, { "epoch": 0.47970996056481363, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4647092819213867, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8751336336135864, "num_tokens": 143161371.0, "step": 3771 }, { "epoch": 0.47983717084340416, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4114584922790527, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8707728981971741, "num_tokens": 143199896.0, "step": 3772 }, { "epoch": 0.47996438112199463, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.492908477783203, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8506870269775391, "num_tokens": 143236577.0, "step": 3773 }, { "epoch": 0.48009159140058516, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.403343677520752, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8559614419937134, "num_tokens": 143279780.0, "step": 3774 }, { "epoch": 0.4802188016791757, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4939606189727783, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8615244030952454, "num_tokens": 143321946.0, "step": 3775 }, { "epoch": 0.48034601195776616, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.388461112976074, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8667593002319336, "num_tokens": 143359278.0, "step": 3776 }, { "epoch": 0.4804732222363567, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5167901515960693, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8498321771621704, "num_tokens": 143398968.0, "step": 3777 }, { "epoch": 0.4806004325149472, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.391185998916626, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.850841760635376, "num_tokens": 143438020.0, "step": 3778 }, { "epoch": 0.4807276427935377, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.525256633758545, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.86125248670578, "num_tokens": 143478813.0, "step": 3779 }, { "epoch": 0.4808548530721282, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3743014335632324, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.849372386932373, "num_tokens": 143519766.0, "step": 3780 }, { "epoch": 0.48098206335071875, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00010728836059570312, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5367417335510254, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8627511262893677, "num_tokens": 143562570.0, "step": 3781 }, { "epoch": 0.4811092736293093, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3900210857391357, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.834078311920166, "num_tokens": 143597203.0, "step": 3782 }, { "epoch": 0.48123648390789975, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.565934181213379, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8548128008842468, "num_tokens": 143635979.0, "step": 3783 }, { "epoch": 0.4813636941864903, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3610007762908936, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8454516530036926, "num_tokens": 143679915.0, "step": 3784 }, { "epoch": 0.4814909044650808, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00010776519775390625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.593441963195801, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8403457403182983, "num_tokens": 143716275.0, "step": 3785 }, { "epoch": 0.4816181147436713, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3538286685943604, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8467457294464111, "num_tokens": 143752029.0, "step": 3786 }, { "epoch": 0.4817453250222618, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6155612468719482, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8622814416885376, "num_tokens": 143793139.0, "step": 3787 }, { "epoch": 0.48187253530085233, "ewc_loss": 0.04541015625, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3339898586273193, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8689525127410889, "num_tokens": 143832790.0, "step": 3788 }, { "epoch": 0.4819997455794428, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 3.0076708793640137, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8524789810180664, "num_tokens": 143876046.0, "step": 3789 }, { "epoch": 0.48212695585803333, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.375849723815918, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8483872413635254, "num_tokens": 143917019.0, "step": 3790 }, { "epoch": 0.48225416613662386, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.8435239791870117, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8622281551361084, "num_tokens": 143957789.0, "step": 3791 }, { "epoch": 0.48238137641521434, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.414170980453491, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8618097305297852, "num_tokens": 143993797.0, "step": 3792 }, { "epoch": 0.48250858669380486, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.884995460510254, "learning_rate": 1e-05, "loss": 0.5526, "mean_token_accuracy": 0.8392500281333923, "num_tokens": 144033583.0, "step": 3793 }, { "epoch": 0.4826357969723954, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.480196237564087, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8572824597358704, "num_tokens": 144074358.0, "step": 3794 }, { "epoch": 0.48276300725098586, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.849297523498535, "learning_rate": 1e-05, "loss": 0.5245, "mean_token_accuracy": 0.8488819599151611, "num_tokens": 144114480.0, "step": 3795 }, { "epoch": 0.4828902175295764, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5203919410705566, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8619938492774963, "num_tokens": 144152306.0, "step": 3796 }, { "epoch": 0.4830174278081669, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.799309730529785, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8576008081436157, "num_tokens": 144191134.0, "step": 3797 }, { "epoch": 0.4831446380867574, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5305979251861572, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8493214845657349, "num_tokens": 144230165.0, "step": 3798 }, { "epoch": 0.4832718483653479, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.733225107192993, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8599315881729126, "num_tokens": 144265035.0, "step": 3799 }, { "epoch": 0.48339905864393845, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.502553701400757, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.874343991279602, "num_tokens": 144304692.0, "step": 3800 }, { "epoch": 0.4835262689225289, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.692213296890259, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8503389358520508, "num_tokens": 144345288.0, "step": 3801 }, { "epoch": 0.48365347920111945, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.512784719467163, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8481941223144531, "num_tokens": 144377427.0, "step": 3802 }, { "epoch": 0.48378068947971, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6449310779571533, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8654226064682007, "num_tokens": 144412405.0, "step": 3803 }, { "epoch": 0.48390789975830045, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4735095500946045, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8684353828430176, "num_tokens": 144448093.0, "step": 3804 }, { "epoch": 0.484035110036891, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6115498542785645, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8528138995170593, "num_tokens": 144483280.0, "step": 3805 }, { "epoch": 0.4841623203154815, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010824203491210938, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4424939155578613, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8759662508964539, "num_tokens": 144520827.0, "step": 3806 }, { "epoch": 0.484289530594072, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.594827175140381, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8643102645874023, "num_tokens": 144562314.0, "step": 3807 }, { "epoch": 0.4844167408726625, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4286253452301025, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8616282343864441, "num_tokens": 144599101.0, "step": 3808 }, { "epoch": 0.48454395115125304, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.585319995880127, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8600897192955017, "num_tokens": 144641308.0, "step": 3809 }, { "epoch": 0.4846711614298435, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4378838539123535, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8445846438407898, "num_tokens": 144677267.0, "step": 3810 }, { "epoch": 0.48479837170843404, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5821235179901123, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8448304533958435, "num_tokens": 144713904.0, "step": 3811 }, { "epoch": 0.48492558198702457, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4109272956848145, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8546996712684631, "num_tokens": 144750523.0, "step": 3812 }, { "epoch": 0.48505279226561504, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.574957847595215, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8560026288032532, "num_tokens": 144791569.0, "step": 3813 }, { "epoch": 0.48518000254420557, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.390632390975952, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8644540309906006, "num_tokens": 144827919.0, "step": 3814 }, { "epoch": 0.4853072128227961, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.574592113494873, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8735071420669556, "num_tokens": 144866859.0, "step": 3815 }, { "epoch": 0.48543442310138657, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.38655424118042, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8708226680755615, "num_tokens": 144903088.0, "step": 3816 }, { "epoch": 0.4855616333799771, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5892274379730225, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8694874048233032, "num_tokens": 144938485.0, "step": 3817 }, { "epoch": 0.4856888436585676, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4163320064544678, "learning_rate": 1e-05, "loss": 0.557, "mean_token_accuracy": 0.8410231471061707, "num_tokens": 144974296.0, "step": 3818 }, { "epoch": 0.4858160539371581, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5766725540161133, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8587884306907654, "num_tokens": 145012391.0, "step": 3819 }, { "epoch": 0.4859432642157486, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.415919542312622, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.841094970703125, "num_tokens": 145045409.0, "step": 3820 }, { "epoch": 0.48607047449433916, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5613441467285156, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8643893003463745, "num_tokens": 145081162.0, "step": 3821 }, { "epoch": 0.48619768477292963, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3938090801239014, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8624008893966675, "num_tokens": 145117045.0, "step": 3822 }, { "epoch": 0.48632489505152016, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5677645206451416, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8651883006095886, "num_tokens": 145153931.0, "step": 3823 }, { "epoch": 0.4864521053301107, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3831405639648438, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.862800121307373, "num_tokens": 145193812.0, "step": 3824 }, { "epoch": 0.48657931560870116, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5781617164611816, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8502262830734253, "num_tokens": 145231814.0, "step": 3825 }, { "epoch": 0.4867065258872917, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.375065803527832, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8516932129859924, "num_tokens": 145279821.0, "step": 3826 }, { "epoch": 0.4868337361658822, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.597754716873169, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.84995436668396, "num_tokens": 145318972.0, "step": 3827 }, { "epoch": 0.4869609464444727, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.367563247680664, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8593085408210754, "num_tokens": 145359479.0, "step": 3828 }, { "epoch": 0.4870881567230632, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.607701063156128, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8472985029220581, "num_tokens": 145403119.0, "step": 3829 }, { "epoch": 0.48721536700165374, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.377774477005005, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8482980728149414, "num_tokens": 145439715.0, "step": 3830 }, { "epoch": 0.48734257728024427, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6106679439544678, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8651142120361328, "num_tokens": 145480715.0, "step": 3831 }, { "epoch": 0.48746978755883474, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3732738494873047, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8550201058387756, "num_tokens": 145523501.0, "step": 3832 }, { "epoch": 0.48759699783742527, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6292967796325684, "learning_rate": 1e-05, "loss": 0.5449, "mean_token_accuracy": 0.8333369493484497, "num_tokens": 145563387.0, "step": 3833 }, { "epoch": 0.4877242081160158, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.380152463912964, "learning_rate": 1e-05, "loss": 0.5221, "mean_token_accuracy": 0.8424777984619141, "num_tokens": 145607818.0, "step": 3834 }, { "epoch": 0.4878514183946063, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6375629901885986, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8437751531600952, "num_tokens": 145646115.0, "step": 3835 }, { "epoch": 0.4879786286731968, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.381911277770996, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8620869517326355, "num_tokens": 145681555.0, "step": 3836 }, { "epoch": 0.48810583895178733, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6405293941497803, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8661161661148071, "num_tokens": 145717459.0, "step": 3837 }, { "epoch": 0.4882330492303778, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.393411636352539, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8482524752616882, "num_tokens": 145753143.0, "step": 3838 }, { "epoch": 0.48836025950896833, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6429381370544434, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8477494716644287, "num_tokens": 145790345.0, "step": 3839 }, { "epoch": 0.48848746978755886, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.386690378189087, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.850361704826355, "num_tokens": 145832900.0, "step": 3840 }, { "epoch": 0.48861468006614933, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6398072242736816, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8611590266227722, "num_tokens": 145870309.0, "step": 3841 }, { "epoch": 0.48874189034473986, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.393080234527588, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.86272132396698, "num_tokens": 145910276.0, "step": 3842 }, { "epoch": 0.4888691006233304, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.629549980163574, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8659167289733887, "num_tokens": 145952945.0, "step": 3843 }, { "epoch": 0.48899631090192086, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4054148197174072, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8521807193756104, "num_tokens": 145987339.0, "step": 3844 }, { "epoch": 0.4891235211805114, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6228249073028564, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8573693037033081, "num_tokens": 146027447.0, "step": 3845 }, { "epoch": 0.4892507314591019, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4090752601623535, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8605265617370605, "num_tokens": 146064192.0, "step": 3846 }, { "epoch": 0.4893779417376924, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010919570922851562, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.621504545211792, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8590429425239563, "num_tokens": 146103468.0, "step": 3847 }, { "epoch": 0.4895051520162829, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3943490982055664, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8647592067718506, "num_tokens": 146142499.0, "step": 3848 }, { "epoch": 0.48963236229487345, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.630530595779419, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.852259635925293, "num_tokens": 146177538.0, "step": 3849 }, { "epoch": 0.4897595725734639, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3895604610443115, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8652544021606445, "num_tokens": 146218259.0, "step": 3850 }, { "epoch": 0.48988678285205445, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6345136165618896, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.857467532157898, "num_tokens": 146256630.0, "step": 3851 }, { "epoch": 0.490013993130645, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4138526916503906, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8385850191116333, "num_tokens": 146290891.0, "step": 3852 }, { "epoch": 0.49014120340923545, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6340038776397705, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8523032665252686, "num_tokens": 146327148.0, "step": 3853 }, { "epoch": 0.490268413687826, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.393990993499756, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.858871579170227, "num_tokens": 146367932.0, "step": 3854 }, { "epoch": 0.4903956239664165, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.614025115966797, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8627122640609741, "num_tokens": 146407639.0, "step": 3855 }, { "epoch": 0.490522834245007, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.404768943786621, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8485205769538879, "num_tokens": 146442974.0, "step": 3856 }, { "epoch": 0.4906500445235975, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.604966402053833, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.864356279373169, "num_tokens": 146482510.0, "step": 3857 }, { "epoch": 0.49077725480218803, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3964807987213135, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8701151609420776, "num_tokens": 146522616.0, "step": 3858 }, { "epoch": 0.4909044650807785, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6141207218170166, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8648316860198975, "num_tokens": 146558226.0, "step": 3859 }, { "epoch": 0.49103167535936904, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.388124704360962, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8516605496406555, "num_tokens": 146601873.0, "step": 3860 }, { "epoch": 0.49115888563795956, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.611109495162964, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8590199947357178, "num_tokens": 146645284.0, "step": 3861 }, { "epoch": 0.49128609591655004, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4085395336151123, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8599288463592529, "num_tokens": 146680374.0, "step": 3862 }, { "epoch": 0.49141330619514056, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6082353591918945, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8514470458030701, "num_tokens": 146721915.0, "step": 3863 }, { "epoch": 0.4915405164737311, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.391622543334961, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.867472767829895, "num_tokens": 146758981.0, "step": 3864 }, { "epoch": 0.49166772675232157, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6169490814208984, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.851240873336792, "num_tokens": 146795606.0, "step": 3865 }, { "epoch": 0.4917949370309121, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00010967254638671875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3954553604125977, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8676786422729492, "num_tokens": 146829681.0, "step": 3866 }, { "epoch": 0.4919221473095026, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.594592571258545, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8723709583282471, "num_tokens": 146874142.0, "step": 3867 }, { "epoch": 0.4920493575880931, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4163596630096436, "learning_rate": 1e-05, "loss": 0.5819, "mean_token_accuracy": 0.82586669921875, "num_tokens": 146911414.0, "step": 3868 }, { "epoch": 0.4921765678666836, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5938637256622314, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8390981554985046, "num_tokens": 146956376.0, "step": 3869 }, { "epoch": 0.49230377814527415, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4011876583099365, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8458805084228516, "num_tokens": 147001539.0, "step": 3870 }, { "epoch": 0.4924309884238646, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.58805775642395, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8457037210464478, "num_tokens": 147043782.0, "step": 3871 }, { "epoch": 0.49255819870245515, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.416718006134033, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8370267748832703, "num_tokens": 147077431.0, "step": 3872 }, { "epoch": 0.4926854089810457, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5774636268615723, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8669384717941284, "num_tokens": 147113325.0, "step": 3873 }, { "epoch": 0.49281261925963615, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4024710655212402, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8376525640487671, "num_tokens": 147158363.0, "step": 3874 }, { "epoch": 0.4929398295382267, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5812666416168213, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.8467875123023987, "num_tokens": 147195868.0, "step": 3875 }, { "epoch": 0.4930670398168172, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.391925573348999, "learning_rate": 1e-05, "loss": 0.547, "mean_token_accuracy": 0.8345391750335693, "num_tokens": 147237782.0, "step": 3876 }, { "epoch": 0.4931942500954077, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00011014938354492188, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5950498580932617, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8480238318443298, "num_tokens": 147274709.0, "step": 3877 }, { "epoch": 0.4933214603739982, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.383730888366699, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8622684478759766, "num_tokens": 147312261.0, "step": 3878 }, { "epoch": 0.49344867065258874, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.599379301071167, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8574481010437012, "num_tokens": 147352290.0, "step": 3879 }, { "epoch": 0.4935758809311792, "ewc_loss": 0.045654296875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3810300827026367, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.85953688621521, "num_tokens": 147390558.0, "step": 3880 }, { "epoch": 0.49370309120976974, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6153552532196045, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.850536584854126, "num_tokens": 147429597.0, "step": 3881 }, { "epoch": 0.49383030148836027, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.374446153640747, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8491891622543335, "num_tokens": 147472302.0, "step": 3882 }, { "epoch": 0.4939575117669508, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.634755849838257, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8700661063194275, "num_tokens": 147502846.0, "step": 3883 }, { "epoch": 0.49408472204554127, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3657386302948, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8642380237579346, "num_tokens": 147542351.0, "step": 3884 }, { "epoch": 0.4942119323241318, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6551566123962402, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8617421984672546, "num_tokens": 147576912.0, "step": 3885 }, { "epoch": 0.4943391426027223, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.000110626220703125, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.382702350616455, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8535947799682617, "num_tokens": 147614687.0, "step": 3886 }, { "epoch": 0.4944663528813128, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6639633178710938, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8652506470680237, "num_tokens": 147649048.0, "step": 3887 }, { "epoch": 0.4945935631599033, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3857288360595703, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8466607332229614, "num_tokens": 147689337.0, "step": 3888 }, { "epoch": 0.49472077343849385, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6708121299743652, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8581331968307495, "num_tokens": 147725040.0, "step": 3889 }, { "epoch": 0.4948479837170843, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4146461486816406, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8559120893478394, "num_tokens": 147761442.0, "step": 3890 }, { "epoch": 0.49497519399567486, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.661357879638672, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8509366512298584, "num_tokens": 147795421.0, "step": 3891 }, { "epoch": 0.4951024042742654, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3969855308532715, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.85623699426651, "num_tokens": 147833591.0, "step": 3892 }, { "epoch": 0.49522961455285586, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6581530570983887, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.8451898694038391, "num_tokens": 147870707.0, "step": 3893 }, { "epoch": 0.4953568248314464, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.413740396499634, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8563238978385925, "num_tokens": 147905231.0, "step": 3894 }, { "epoch": 0.4954840351100369, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6410484313964844, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8590136170387268, "num_tokens": 147945890.0, "step": 3895 }, { "epoch": 0.4956112453886274, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.418827533721924, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8583697080612183, "num_tokens": 147984572.0, "step": 3896 }, { "epoch": 0.4957384556672179, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.633051633834839, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8664168119430542, "num_tokens": 148020059.0, "step": 3897 }, { "epoch": 0.49586566594580844, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4093613624572754, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.871306836605072, "num_tokens": 148058442.0, "step": 3898 }, { "epoch": 0.4959928762243989, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6339046955108643, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8654603362083435, "num_tokens": 148093710.0, "step": 3899 }, { "epoch": 0.49612008650298944, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4042673110961914, "learning_rate": 1e-05, "loss": 0.398, "mean_token_accuracy": 0.8783605694770813, "num_tokens": 148135763.0, "step": 3900 }, { "epoch": 0.49624729678157997, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00011157989501953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.630295991897583, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8655587434768677, "num_tokens": 148169065.0, "step": 3901 }, { "epoch": 0.49637450706017044, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4184093475341797, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8451019525527954, "num_tokens": 148205779.0, "step": 3902 }, { "epoch": 0.496501717338761, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.7717864513397217, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8415021896362305, "num_tokens": 148241893.0, "step": 3903 }, { "epoch": 0.4966289276173515, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.413983106613159, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8608865737915039, "num_tokens": 148275384.0, "step": 3904 }, { "epoch": 0.496756137895942, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.675259590148926, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8579108119010925, "num_tokens": 148311248.0, "step": 3905 }, { "epoch": 0.4968833481745325, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011205673217773438, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.416100263595581, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8565285801887512, "num_tokens": 148347717.0, "step": 3906 }, { "epoch": 0.49701055845312303, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6641173362731934, "learning_rate": 1e-05, "loss": 0.5401, "mean_token_accuracy": 0.8390156626701355, "num_tokens": 148393395.0, "step": 3907 }, { "epoch": 0.4971377687317135, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.423337697982788, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8669555187225342, "num_tokens": 148427843.0, "step": 3908 }, { "epoch": 0.49726497901030403, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6474225521087646, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8453028202056885, "num_tokens": 148470813.0, "step": 3909 }, { "epoch": 0.49739218928889456, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4466514587402344, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8553862571716309, "num_tokens": 148504491.0, "step": 3910 }, { "epoch": 0.49751939956748503, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6158952713012695, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8510419130325317, "num_tokens": 148545934.0, "step": 3911 }, { "epoch": 0.49764660984607556, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.444746255874634, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8629918098449707, "num_tokens": 148583521.0, "step": 3912 }, { "epoch": 0.4977738201246661, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5873637199401855, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8524644374847412, "num_tokens": 148621484.0, "step": 3913 }, { "epoch": 0.49790103040325656, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.434791088104248, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8455339670181274, "num_tokens": 148661198.0, "step": 3914 }, { "epoch": 0.4980282406818471, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.558255672454834, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.856376051902771, "num_tokens": 148702620.0, "step": 3915 }, { "epoch": 0.4981554509604376, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.430424928665161, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8528647422790527, "num_tokens": 148739991.0, "step": 3916 }, { "epoch": 0.4982826612390281, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.540074348449707, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8733997344970703, "num_tokens": 148780520.0, "step": 3917 }, { "epoch": 0.4984098715176186, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4301300048828125, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8496357202529907, "num_tokens": 148817604.0, "step": 3918 }, { "epoch": 0.49853708179620915, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.539285659790039, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8534090518951416, "num_tokens": 148858028.0, "step": 3919 }, { "epoch": 0.4986642920747996, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4182629585266113, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8604202270507812, "num_tokens": 148895458.0, "step": 3920 }, { "epoch": 0.49879150235339015, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.541980504989624, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8517899513244629, "num_tokens": 148933305.0, "step": 3921 }, { "epoch": 0.4989187126319807, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.407076835632324, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8685626983642578, "num_tokens": 148970587.0, "step": 3922 }, { "epoch": 0.49904592291057115, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.553877830505371, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8506451845169067, "num_tokens": 149004874.0, "step": 3923 }, { "epoch": 0.4991731331891617, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.397052526473999, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.86191725730896, "num_tokens": 149045169.0, "step": 3924 }, { "epoch": 0.4993003434677522, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.567105770111084, "learning_rate": 1e-05, "loss": 0.5702, "mean_token_accuracy": 0.8257589340209961, "num_tokens": 149081554.0, "step": 3925 }, { "epoch": 0.4994275537463427, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.379279851913452, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8484607934951782, "num_tokens": 149124500.0, "step": 3926 }, { "epoch": 0.4995547640249332, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5785388946533203, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8497644066810608, "num_tokens": 149162522.0, "step": 3927 }, { "epoch": 0.49968197430352373, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3717856407165527, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8595542907714844, "num_tokens": 149198083.0, "step": 3928 }, { "epoch": 0.4998091845821142, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6073648929595947, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8431335687637329, "num_tokens": 149234786.0, "step": 3929 }, { "epoch": 0.49993639486070474, "ewc_loss": 0.046142578125, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.366168260574341, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8740510940551758, "num_tokens": 149269859.0, "step": 3930 }, { "epoch": 0.5000636051392953, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6247143745422363, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8360961079597473, "num_tokens": 149311209.0, "step": 3931 }, { "epoch": 0.5001908154178858, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011301040649414062, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3657498359680176, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8515466451644897, "num_tokens": 149351283.0, "step": 3932 }, { "epoch": 0.5003180256964763, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6375555992126465, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8566004037857056, "num_tokens": 149391068.0, "step": 3933 }, { "epoch": 0.5004452359750667, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.354785203933716, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8487241268157959, "num_tokens": 149435078.0, "step": 3934 }, { "epoch": 0.5005724462536573, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6651952266693115, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8440049886703491, "num_tokens": 149481336.0, "step": 3935 }, { "epoch": 0.5006996565322478, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.386004686355591, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8571467995643616, "num_tokens": 149523278.0, "step": 3936 }, { "epoch": 0.5008268668108383, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6553611755371094, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8746213912963867, "num_tokens": 149563011.0, "step": 3937 }, { "epoch": 0.5009540770894289, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3864035606384277, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8544410467147827, "num_tokens": 149606609.0, "step": 3938 }, { "epoch": 0.5010812873680194, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6578712463378906, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.868289589881897, "num_tokens": 149647012.0, "step": 3939 }, { "epoch": 0.5012084976466098, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.400254726409912, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8607231974601746, "num_tokens": 149680153.0, "step": 3940 }, { "epoch": 0.5013357079252003, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6539218425750732, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8637430667877197, "num_tokens": 149717080.0, "step": 3941 }, { "epoch": 0.5014629182037909, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4020326137542725, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8596071004867554, "num_tokens": 149755934.0, "step": 3942 }, { "epoch": 0.5015901284823814, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6525535583496094, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8582462668418884, "num_tokens": 149792714.0, "step": 3943 }, { "epoch": 0.5017173387609719, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4212355613708496, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8481464385986328, "num_tokens": 149828647.0, "step": 3944 }, { "epoch": 0.5018445490395624, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.631484031677246, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.856560230255127, "num_tokens": 149864303.0, "step": 3945 }, { "epoch": 0.5019717593181529, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.421384811401367, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.8419487476348877, "num_tokens": 149902154.0, "step": 3946 }, { "epoch": 0.5020989695967434, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.630622148513794, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8626563549041748, "num_tokens": 149930902.0, "step": 3947 }, { "epoch": 0.5022261798753339, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.40586519241333, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8549767136573792, "num_tokens": 149972033.0, "step": 3948 }, { "epoch": 0.5023533901539244, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.618490219116211, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8666999340057373, "num_tokens": 150007278.0, "step": 3949 }, { "epoch": 0.502480600432515, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4018540382385254, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8538388609886169, "num_tokens": 150054512.0, "step": 3950 }, { "epoch": 0.5026078107111055, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.614560842514038, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8547753095626831, "num_tokens": 150094614.0, "step": 3951 }, { "epoch": 0.5027350209896959, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.40585994720459, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8577048182487488, "num_tokens": 150133011.0, "step": 3952 }, { "epoch": 0.5028622312682864, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.615623950958252, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8527882099151611, "num_tokens": 150170041.0, "step": 3953 }, { "epoch": 0.502989441546877, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.410163164138794, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8537389039993286, "num_tokens": 150206034.0, "step": 3954 }, { "epoch": 0.5031166518254675, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6069083213806152, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8563089966773987, "num_tokens": 150241601.0, "step": 3955 }, { "epoch": 0.503243862104058, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3964571952819824, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8673321008682251, "num_tokens": 150279219.0, "step": 3956 }, { "epoch": 0.5033710723826486, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6030826568603516, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8537255525588989, "num_tokens": 150323160.0, "step": 3957 }, { "epoch": 0.5034982826612391, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3867876529693604, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.86063152551651, "num_tokens": 150363788.0, "step": 3958 }, { "epoch": 0.5036254929398295, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6316146850585938, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8511949777603149, "num_tokens": 150395481.0, "step": 3959 }, { "epoch": 0.50375270321842, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011348724365234375, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4063851833343506, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8613264560699463, "num_tokens": 150425726.0, "step": 3960 }, { "epoch": 0.5038799134970106, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.624946355819702, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8505498766899109, "num_tokens": 150463073.0, "step": 3961 }, { "epoch": 0.5040071237756011, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3969435691833496, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8429598808288574, "num_tokens": 150502938.0, "step": 3962 }, { "epoch": 0.5041343340541916, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.653712749481201, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8542719483375549, "num_tokens": 150538601.0, "step": 3963 }, { "epoch": 0.5042615443327821, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4043123722076416, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8536775708198547, "num_tokens": 150573745.0, "step": 3964 }, { "epoch": 0.5043887546113726, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6341702938079834, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8437627553939819, "num_tokens": 150613029.0, "step": 3965 }, { "epoch": 0.5045159648899631, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3957881927490234, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8632904291152954, "num_tokens": 150649773.0, "step": 3966 }, { "epoch": 0.5046431751685536, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6177196502685547, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8594251871109009, "num_tokens": 150694764.0, "step": 3967 }, { "epoch": 0.5047703854471441, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.403128147125244, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8501175045967102, "num_tokens": 150735233.0, "step": 3968 }, { "epoch": 0.5048975957257347, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6143107414245605, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8769209384918213, "num_tokens": 150769596.0, "step": 3969 }, { "epoch": 0.5050248060043252, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4181997776031494, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.850578784942627, "num_tokens": 150806096.0, "step": 3970 }, { "epoch": 0.5051520162829156, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6001055240631104, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8554648160934448, "num_tokens": 150844721.0, "step": 3971 }, { "epoch": 0.5052792265615061, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.399233818054199, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8588719964027405, "num_tokens": 150885711.0, "step": 3972 }, { "epoch": 0.5054064368400967, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.606405019760132, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.854019045829773, "num_tokens": 150924993.0, "step": 3973 }, { "epoch": 0.5055336471186872, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.413193464279175, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8680132031440735, "num_tokens": 150956139.0, "step": 3974 }, { "epoch": 0.5056608573972777, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.603632688522339, "learning_rate": 1e-05, "loss": 0.5455, "mean_token_accuracy": 0.8381402492523193, "num_tokens": 150995816.0, "step": 3975 }, { "epoch": 0.5057880676758683, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4066693782806396, "learning_rate": 1e-05, "loss": 0.5185, "mean_token_accuracy": 0.8399012088775635, "num_tokens": 151035191.0, "step": 3976 }, { "epoch": 0.5059152779544587, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.603153944015503, "learning_rate": 1e-05, "loss": 0.5261, "mean_token_accuracy": 0.8368332386016846, "num_tokens": 151069625.0, "step": 3977 }, { "epoch": 0.5060424882330492, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4182207584381104, "learning_rate": 1e-05, "loss": 0.5135, "mean_token_accuracy": 0.8462650179862976, "num_tokens": 151105999.0, "step": 3978 }, { "epoch": 0.5061696985116397, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5770950317382812, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8571069240570068, "num_tokens": 151148712.0, "step": 3979 }, { "epoch": 0.5062969087902303, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011396408081054688, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4103734493255615, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8709232807159424, "num_tokens": 151182842.0, "step": 3980 }, { "epoch": 0.5064241190688208, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.583404064178467, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8554301261901855, "num_tokens": 151223731.0, "step": 3981 }, { "epoch": 0.5065513293474113, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4087042808532715, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8535155653953552, "num_tokens": 151259711.0, "step": 3982 }, { "epoch": 0.5066785396260017, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.570840358734131, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8485326766967773, "num_tokens": 151301486.0, "step": 3983 }, { "epoch": 0.5068057499045923, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.403366804122925, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8555527925491333, "num_tokens": 151339068.0, "step": 3984 }, { "epoch": 0.5069329601831828, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5714681148529053, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8461044430732727, "num_tokens": 151380694.0, "step": 3985 }, { "epoch": 0.5070601704617733, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.396033763885498, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8592602014541626, "num_tokens": 151416522.0, "step": 3986 }, { "epoch": 0.5071873807403638, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5836238861083984, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8399007320404053, "num_tokens": 151453786.0, "step": 3987 }, { "epoch": 0.5073145910189544, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.386465311050415, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8553451299667358, "num_tokens": 151492155.0, "step": 3988 }, { "epoch": 0.5074418012975448, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.579754114151001, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.863089919090271, "num_tokens": 151534979.0, "step": 3989 }, { "epoch": 0.5075690115761353, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3892500400543213, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8507518768310547, "num_tokens": 151573670.0, "step": 3990 }, { "epoch": 0.5076962218547258, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.595085620880127, "learning_rate": 1e-05, "loss": 0.3943, "mean_token_accuracy": 0.8825985193252563, "num_tokens": 151605232.0, "step": 3991 }, { "epoch": 0.5078234321333164, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.385349750518799, "learning_rate": 1e-05, "loss": 0.5296, "mean_token_accuracy": 0.8384263515472412, "num_tokens": 151648696.0, "step": 3992 }, { "epoch": 0.5079506424119069, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6118240356445312, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8651847243309021, "num_tokens": 151684203.0, "step": 3993 }, { "epoch": 0.5080778526904974, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3970274925231934, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8487628102302551, "num_tokens": 151716749.0, "step": 3994 }, { "epoch": 0.5082050629690879, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6179137229919434, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8446749448776245, "num_tokens": 151752567.0, "step": 3995 }, { "epoch": 0.5083322732476784, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.365875720977783, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8772096633911133, "num_tokens": 151794201.0, "step": 3996 }, { "epoch": 0.5084594835262689, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.63335919380188, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8606278896331787, "num_tokens": 151833731.0, "step": 3997 }, { "epoch": 0.5085866938048594, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3807880878448486, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8578356504440308, "num_tokens": 151872738.0, "step": 3998 }, { "epoch": 0.50871390408345, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00011444091796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6858949661254883, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8643982410430908, "num_tokens": 151916801.0, "step": 3999 }, { "epoch": 0.5088411143620405, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3910610675811768, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8538076281547546, "num_tokens": 151949661.0, "step": 4000 }, { "epoch": 0.5089683246406309, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6788830757141113, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8490411043167114, "num_tokens": 151986719.0, "step": 4001 }, { "epoch": 0.5090955349192214, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3913967609405518, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8520601987838745, "num_tokens": 152023921.0, "step": 4002 }, { "epoch": 0.509222745197812, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.672581195831299, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8643094301223755, "num_tokens": 152061246.0, "step": 4003 }, { "epoch": 0.5093499554764025, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.385411500930786, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8458216190338135, "num_tokens": 152103299.0, "step": 4004 }, { "epoch": 0.509477165754993, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00011491775512695312, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6715903282165527, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8465029001235962, "num_tokens": 152148576.0, "step": 4005 }, { "epoch": 0.5096043760335836, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4082155227661133, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8549912571907043, "num_tokens": 152189221.0, "step": 4006 }, { "epoch": 0.5097315863121741, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6593031883239746, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.851374626159668, "num_tokens": 152232732.0, "step": 4007 }, { "epoch": 0.5098587965907645, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.419786214828491, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8616809844970703, "num_tokens": 152268715.0, "step": 4008 }, { "epoch": 0.509986006869355, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.638256311416626, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.849769115447998, "num_tokens": 152311141.0, "step": 4009 }, { "epoch": 0.5101132171479456, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.416987419128418, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8708449602127075, "num_tokens": 152348892.0, "step": 4010 }, { "epoch": 0.5102404274265361, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6351165771484375, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8596041202545166, "num_tokens": 152383300.0, "step": 4011 }, { "epoch": 0.5103676377051266, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.413944959640503, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8526577949523926, "num_tokens": 152424404.0, "step": 4012 }, { "epoch": 0.5104948479837171, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6119329929351807, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8732797503471375, "num_tokens": 152465226.0, "step": 4013 }, { "epoch": 0.5106220582623076, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.410519599914551, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8493210077285767, "num_tokens": 152507819.0, "step": 4014 }, { "epoch": 0.5107492685408981, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.598310947418213, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.878495991230011, "num_tokens": 152548194.0, "step": 4015 }, { "epoch": 0.5108764788194886, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.427473783493042, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8582846522331238, "num_tokens": 152580255.0, "step": 4016 }, { "epoch": 0.5110036890980791, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5906245708465576, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8554866313934326, "num_tokens": 152619813.0, "step": 4017 }, { "epoch": 0.5111308993766697, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.433459997177124, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8497610092163086, "num_tokens": 152655552.0, "step": 4018 }, { "epoch": 0.5112581096552602, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.577216625213623, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.844009280204773, "num_tokens": 152693695.0, "step": 4019 }, { "epoch": 0.5113853199338506, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4118897914886475, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8540817499160767, "num_tokens": 152732917.0, "step": 4020 }, { "epoch": 0.5115125302124411, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.577073335647583, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8636995553970337, "num_tokens": 152768052.0, "step": 4021 }, { "epoch": 0.5116397404910317, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.416196584701538, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.8427414298057556, "num_tokens": 152808106.0, "step": 4022 }, { "epoch": 0.5117669507696222, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5563535690307617, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8794050216674805, "num_tokens": 152849989.0, "step": 4023 }, { "epoch": 0.5118941610482127, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4044101238250732, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8619803190231323, "num_tokens": 152886599.0, "step": 4024 }, { "epoch": 0.5120213713268033, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.568293333053589, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8500655293464661, "num_tokens": 152926383.0, "step": 4025 }, { "epoch": 0.5121485816053937, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.397343873977661, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.851195216178894, "num_tokens": 152966141.0, "step": 4026 }, { "epoch": 0.5122757918839842, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5887420177459717, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8621336221694946, "num_tokens": 152998217.0, "step": 4027 }, { "epoch": 0.5124030021625747, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4063479900360107, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8414508700370789, "num_tokens": 153032258.0, "step": 4028 }, { "epoch": 0.5125302124411653, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5692121982574463, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.860259473323822, "num_tokens": 153075762.0, "step": 4029 }, { "epoch": 0.5126574227197558, "ewc_loss": 0.04638671875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3901071548461914, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8527641892433167, "num_tokens": 153112571.0, "step": 4030 }, { "epoch": 0.5127846329983463, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5733368396759033, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8654137849807739, "num_tokens": 153153348.0, "step": 4031 }, { "epoch": 0.5129118432769367, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3891899585723877, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8663792014122009, "num_tokens": 153189616.0, "step": 4032 }, { "epoch": 0.5130390535555273, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.597165584564209, "learning_rate": 1e-05, "loss": 0.5358, "mean_token_accuracy": 0.841728925704956, "num_tokens": 153230879.0, "step": 4033 }, { "epoch": 0.5131662638341178, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3778867721557617, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8519232273101807, "num_tokens": 153275234.0, "step": 4034 }, { "epoch": 0.5132934741127083, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011587142944335938, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.598506450653076, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8687127232551575, "num_tokens": 153312606.0, "step": 4035 }, { "epoch": 0.5134206843912988, "ewc_loss": 0.046875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.378074884414673, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8464673757553101, "num_tokens": 153352370.0, "step": 4036 }, { "epoch": 0.5135478946698894, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.612870454788208, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8589118123054504, "num_tokens": 153393030.0, "step": 4037 }, { "epoch": 0.5136751049484798, "ewc_loss": 0.046875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4038002490997314, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8539798855781555, "num_tokens": 153423831.0, "step": 4038 }, { "epoch": 0.5138023152270703, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6107990741729736, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8676687479019165, "num_tokens": 153458388.0, "step": 4039 }, { "epoch": 0.5139295255056608, "ewc_loss": 0.046875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3833022117614746, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.848278820514679, "num_tokens": 153503486.0, "step": 4040 }, { "epoch": 0.5140567357842514, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6225507259368896, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8604604601860046, "num_tokens": 153541474.0, "step": 4041 }, { "epoch": 0.5141839460628419, "ewc_loss": 0.046875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3822526931762695, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8465169668197632, "num_tokens": 153583518.0, "step": 4042 }, { "epoch": 0.5143111563414324, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.645320177078247, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8545795679092407, "num_tokens": 153617552.0, "step": 4043 }, { "epoch": 0.5144383666200228, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3868746757507324, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8445973992347717, "num_tokens": 153655616.0, "step": 4044 }, { "epoch": 0.5145655768986134, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.647188663482666, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8631337285041809, "num_tokens": 153692835.0, "step": 4045 }, { "epoch": 0.5146927871772039, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001163482666015625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.422175407409668, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8510564565658569, "num_tokens": 153720867.0, "step": 4046 }, { "epoch": 0.5148199974557944, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6334707736968994, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8611929416656494, "num_tokens": 153758798.0, "step": 4047 }, { "epoch": 0.514947207734385, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4174916744232178, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.842587947845459, "num_tokens": 153798257.0, "step": 4048 }, { "epoch": 0.5150744180129755, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6151654720306396, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8681553602218628, "num_tokens": 153839241.0, "step": 4049 }, { "epoch": 0.5152016282915659, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3942553997039795, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8562597036361694, "num_tokens": 153882522.0, "step": 4050 }, { "epoch": 0.5153288385701564, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6325490474700928, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8698164224624634, "num_tokens": 153912173.0, "step": 4051 }, { "epoch": 0.515456048848747, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.416304349899292, "learning_rate": 1e-05, "loss": 0.585, "mean_token_accuracy": 0.821056604385376, "num_tokens": 153955070.0, "step": 4052 }, { "epoch": 0.5155832591273375, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.605379819869995, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8617772459983826, "num_tokens": 153995526.0, "step": 4053 }, { "epoch": 0.515710469405928, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4046056270599365, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8538239002227783, "num_tokens": 154035476.0, "step": 4054 }, { "epoch": 0.5158376796845185, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6050732135772705, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8490813970565796, "num_tokens": 154079108.0, "step": 4055 }, { "epoch": 0.5159648899631091, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.411259651184082, "learning_rate": 1e-05, "loss": 0.5659, "mean_token_accuracy": 0.830193281173706, "num_tokens": 154118484.0, "step": 4056 }, { "epoch": 0.5160921002416995, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5833489894866943, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8574594259262085, "num_tokens": 154163981.0, "step": 4057 }, { "epoch": 0.51621931052029, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4095771312713623, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8441444039344788, "num_tokens": 154205035.0, "step": 4058 }, { "epoch": 0.5163465207988805, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.579425096511841, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8662182092666626, "num_tokens": 154241243.0, "step": 4059 }, { "epoch": 0.5164737310774711, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3936879634857178, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8422559499740601, "num_tokens": 154289182.0, "step": 4060 }, { "epoch": 0.5166009413560616, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5901973247528076, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8466466665267944, "num_tokens": 154326735.0, "step": 4061 }, { "epoch": 0.5167281516346521, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.452407121658325, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8569481372833252, "num_tokens": 154367795.0, "step": 4062 }, { "epoch": 0.5168553619132426, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5628957748413086, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8725876808166504, "num_tokens": 154401949.0, "step": 4063 }, { "epoch": 0.5169825721918331, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4020793437957764, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8672357201576233, "num_tokens": 154442693.0, "step": 4064 }, { "epoch": 0.5171097824704236, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5535778999328613, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.857250452041626, "num_tokens": 154489219.0, "step": 4065 }, { "epoch": 0.5172369927490141, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.402148485183716, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8718536496162415, "num_tokens": 154526639.0, "step": 4066 }, { "epoch": 0.5173642030276047, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5775022506713867, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.868580162525177, "num_tokens": 154557339.0, "step": 4067 }, { "epoch": 0.5174914133061952, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.410031795501709, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8511492013931274, "num_tokens": 154592738.0, "step": 4068 }, { "epoch": 0.5176186235847856, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.568483591079712, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8559920787811279, "num_tokens": 154631633.0, "step": 4069 }, { "epoch": 0.5177458338633761, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3953499794006348, "learning_rate": 1e-05, "loss": 0.5542, "mean_token_accuracy": 0.8334568738937378, "num_tokens": 154676274.0, "step": 4070 }, { "epoch": 0.5178730441419667, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.582209348678589, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8422448039054871, "num_tokens": 154716379.0, "step": 4071 }, { "epoch": 0.5180002544205572, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3903756141662598, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8516702651977539, "num_tokens": 154757038.0, "step": 4072 }, { "epoch": 0.5181274646991477, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5789430141448975, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8811880946159363, "num_tokens": 154795684.0, "step": 4073 }, { "epoch": 0.5182546749777382, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3844401836395264, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8590118288993835, "num_tokens": 154834858.0, "step": 4074 }, { "epoch": 0.5183818852563287, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00011682510375976562, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5924179553985596, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8610324859619141, "num_tokens": 154868919.0, "step": 4075 }, { "epoch": 0.5185090955349192, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3850343227386475, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8540396690368652, "num_tokens": 154910511.0, "step": 4076 }, { "epoch": 0.5186363058135097, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.600618600845337, "learning_rate": 1e-05, "loss": 0.5247, "mean_token_accuracy": 0.842284083366394, "num_tokens": 154955345.0, "step": 4077 }, { "epoch": 0.5187635160921003, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3831465244293213, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8619263768196106, "num_tokens": 154992541.0, "step": 4078 }, { "epoch": 0.5188907263706908, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6160120964050293, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8562774062156677, "num_tokens": 155033592.0, "step": 4079 }, { "epoch": 0.5190179366492813, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3640477657318115, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8584755063056946, "num_tokens": 155080034.0, "step": 4080 }, { "epoch": 0.5191451469278717, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6498818397521973, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8451438546180725, "num_tokens": 155116207.0, "step": 4081 }, { "epoch": 0.5192723572064623, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3774526119232178, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8534260392189026, "num_tokens": 155155787.0, "step": 4082 }, { "epoch": 0.5193995674850528, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6615471839904785, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8627274632453918, "num_tokens": 155194021.0, "step": 4083 }, { "epoch": 0.5195267777636433, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3886160850524902, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.861859917640686, "num_tokens": 155228685.0, "step": 4084 }, { "epoch": 0.5196539880422338, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6732265949249268, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8543850779533386, "num_tokens": 155269146.0, "step": 4085 }, { "epoch": 0.5197811983208244, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.384897470474243, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8569880127906799, "num_tokens": 155309080.0, "step": 4086 }, { "epoch": 0.5199084085994148, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00011730194091796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.678091526031494, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8578917980194092, "num_tokens": 155348661.0, "step": 4087 }, { "epoch": 0.5200356188780053, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.404360055923462, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.849790096282959, "num_tokens": 155384984.0, "step": 4088 }, { "epoch": 0.5201628291565958, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.684781789779663, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8496297597885132, "num_tokens": 155421914.0, "step": 4089 }, { "epoch": 0.5202900394351864, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.000118255615234375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.398432493209839, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8620874285697937, "num_tokens": 155466460.0, "step": 4090 }, { "epoch": 0.5204172497137769, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.687248706817627, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8606852889060974, "num_tokens": 155504034.0, "step": 4091 }, { "epoch": 0.5205444599923674, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4166412353515625, "learning_rate": 1e-05, "loss": 0.5215, "mean_token_accuracy": 0.8449146151542664, "num_tokens": 155542521.0, "step": 4092 }, { "epoch": 0.5206716702709578, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.688070058822632, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8617086410522461, "num_tokens": 155574170.0, "step": 4093 }, { "epoch": 0.5207988805495484, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.440155506134033, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8588273525238037, "num_tokens": 155604314.0, "step": 4094 }, { "epoch": 0.5209260908281389, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.652423143386841, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8598097562789917, "num_tokens": 155643259.0, "step": 4095 }, { "epoch": 0.5210533011067294, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4275600910186768, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8530428409576416, "num_tokens": 155682522.0, "step": 4096 }, { "epoch": 0.52118051138532, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.650561809539795, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8666424751281738, "num_tokens": 155714021.0, "step": 4097 }, { "epoch": 0.5213077216639105, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4392688274383545, "learning_rate": 1e-05, "loss": 0.5346, "mean_token_accuracy": 0.83750319480896, "num_tokens": 155754064.0, "step": 4098 }, { "epoch": 0.5214349319425009, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6201720237731934, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8584683537483215, "num_tokens": 155788428.0, "step": 4099 }, { "epoch": 0.5215621422210914, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.435098648071289, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8522757291793823, "num_tokens": 155826608.0, "step": 4100 }, { "epoch": 0.521689352499682, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5853450298309326, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8557455539703369, "num_tokens": 155870314.0, "step": 4101 }, { "epoch": 0.5218165627782725, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4295027256011963, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8446742296218872, "num_tokens": 155907742.0, "step": 4102 }, { "epoch": 0.521943773056863, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5701982975006104, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8715023994445801, "num_tokens": 155949837.0, "step": 4103 }, { "epoch": 0.5220709833354535, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4168195724487305, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8539437055587769, "num_tokens": 155990407.0, "step": 4104 }, { "epoch": 0.522198193614044, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.57351016998291, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8527677059173584, "num_tokens": 156027157.0, "step": 4105 }, { "epoch": 0.5223254038926345, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4192264080047607, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8548414707183838, "num_tokens": 156065074.0, "step": 4106 }, { "epoch": 0.522452614171225, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.557650566101074, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8656697273254395, "num_tokens": 156103399.0, "step": 4107 }, { "epoch": 0.5225798244498155, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3946986198425293, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.860909640789032, "num_tokens": 156143825.0, "step": 4108 }, { "epoch": 0.5227070347284061, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5701892375946045, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8611390590667725, "num_tokens": 156183841.0, "step": 4109 }, { "epoch": 0.5228342450069966, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.3908791542053223, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8615775108337402, "num_tokens": 156225288.0, "step": 4110 }, { "epoch": 0.5229614552855871, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.589582681655884, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8725202083587646, "num_tokens": 156257463.0, "step": 4111 }, { "epoch": 0.5230886655641775, "ewc_loss": 0.046630859375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3983781337738037, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8566655516624451, "num_tokens": 156293059.0, "step": 4112 }, { "epoch": 0.5232158758427681, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.600172281265259, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8344393968582153, "num_tokens": 156338605.0, "step": 4113 }, { "epoch": 0.5233430861213586, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.395843505859375, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8538296818733215, "num_tokens": 156375472.0, "step": 4114 }, { "epoch": 0.5234702963999491, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6070361137390137, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.858070969581604, "num_tokens": 156412290.0, "step": 4115 }, { "epoch": 0.5235975066785397, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.7098586559295654, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8550465703010559, "num_tokens": 156452268.0, "step": 4116 }, { "epoch": 0.5237247169571302, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4496631622314453, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8501633405685425, "num_tokens": 156500659.0, "step": 4117 }, { "epoch": 0.5238519272357206, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.534604787826538, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8612936735153198, "num_tokens": 156534529.0, "step": 4118 }, { "epoch": 0.5239791375143111, "ewc_loss": 0.046875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.385364294052124, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.86948561668396, "num_tokens": 156574459.0, "step": 4119 }, { "epoch": 0.5241063477929017, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.571643590927124, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8574118614196777, "num_tokens": 156611626.0, "step": 4120 }, { "epoch": 0.5242335580714922, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3778483867645264, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8614597320556641, "num_tokens": 156649454.0, "step": 4121 }, { "epoch": 0.5243607683500827, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5988242626190186, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8549450039863586, "num_tokens": 156687263.0, "step": 4122 }, { "epoch": 0.5244879786286732, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.383660078048706, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8533799648284912, "num_tokens": 156720384.0, "step": 4123 }, { "epoch": 0.5246151889072637, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.614607810974121, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.860429584980011, "num_tokens": 156758365.0, "step": 4124 }, { "epoch": 0.5247423991858542, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.380277633666992, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8565565347671509, "num_tokens": 156790411.0, "step": 4125 }, { "epoch": 0.5248696094644447, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6295011043548584, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8575313687324524, "num_tokens": 156829176.0, "step": 4126 }, { "epoch": 0.5249968197430352, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3709716796875, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8649919033050537, "num_tokens": 156871932.0, "step": 4127 }, { "epoch": 0.5251240300216258, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6524288654327393, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8689878582954407, "num_tokens": 156905810.0, "step": 4128 }, { "epoch": 0.5252512403002163, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.368474245071411, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8653486967086792, "num_tokens": 156947218.0, "step": 4129 }, { "epoch": 0.5253784505788067, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.700718402862549, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8471086025238037, "num_tokens": 156978943.0, "step": 4130 }, { "epoch": 0.5255056608573972, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.364614486694336, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8675961494445801, "num_tokens": 157015015.0, "step": 4131 }, { "epoch": 0.5256328711359878, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7339844703674316, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8560042381286621, "num_tokens": 157051789.0, "step": 4132 }, { "epoch": 0.5257600814145783, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3796801567077637, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8676117658615112, "num_tokens": 157090568.0, "step": 4133 }, { "epoch": 0.5258872916931688, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00011873245239257812, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.755998373031616, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8616558909416199, "num_tokens": 157125255.0, "step": 4134 }, { "epoch": 0.5260145019717594, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.409299612045288, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8557232022285461, "num_tokens": 157164917.0, "step": 4135 }, { "epoch": 0.5261417122503498, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.74949049949646, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8514095544815063, "num_tokens": 157205186.0, "step": 4136 }, { "epoch": 0.5262689225289403, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4558231830596924, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8467758893966675, "num_tokens": 157241511.0, "step": 4137 }, { "epoch": 0.5263961328075308, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.703810214996338, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8633013367652893, "num_tokens": 157283381.0, "step": 4138 }, { "epoch": 0.5265233430861214, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4519119262695312, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8549022674560547, "num_tokens": 157321315.0, "step": 4139 }, { "epoch": 0.5266505533647119, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.696526050567627, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8455108404159546, "num_tokens": 157359205.0, "step": 4140 }, { "epoch": 0.5267777636433024, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4470410346984863, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8650703430175781, "num_tokens": 157398889.0, "step": 4141 }, { "epoch": 0.5269049739218928, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6588363647460938, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8467658758163452, "num_tokens": 157440831.0, "step": 4142 }, { "epoch": 0.5270321842004834, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4548113346099854, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8590678572654724, "num_tokens": 157476275.0, "step": 4143 }, { "epoch": 0.5271593944790739, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00011920928955078125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.637566566467285, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8718358278274536, "num_tokens": 157508830.0, "step": 4144 }, { "epoch": 0.5272866047576644, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4623525142669678, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8634328246116638, "num_tokens": 157539077.0, "step": 4145 }, { "epoch": 0.527413815036255, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5904698371887207, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8686610460281372, "num_tokens": 157583501.0, "step": 4146 }, { "epoch": 0.5275410253148455, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.431626319885254, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8567166328430176, "num_tokens": 157629778.0, "step": 4147 }, { "epoch": 0.5276682355934359, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5897152423858643, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8617726564407349, "num_tokens": 157669375.0, "step": 4148 }, { "epoch": 0.5277954458720264, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4271676540374756, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8674826622009277, "num_tokens": 157707400.0, "step": 4149 }, { "epoch": 0.527922656150617, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.582996368408203, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8569772243499756, "num_tokens": 157747039.0, "step": 4150 }, { "epoch": 0.5280498664292075, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.42103910446167, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8576753735542297, "num_tokens": 157782790.0, "step": 4151 }, { "epoch": 0.528177076707798, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.576336622238159, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8616536259651184, "num_tokens": 157832002.0, "step": 4152 }, { "epoch": 0.5283042869863885, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4099068641662598, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8472734689712524, "num_tokens": 157869875.0, "step": 4153 }, { "epoch": 0.528431497264979, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5812623500823975, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8503484129905701, "num_tokens": 157908910.0, "step": 4154 }, { "epoch": 0.5285587075435695, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4123427867889404, "learning_rate": 1e-05, "loss": 0.4941, "mean_token_accuracy": 0.8433243036270142, "num_tokens": 157944366.0, "step": 4155 }, { "epoch": 0.52868591782216, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.58086895942688, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8598158359527588, "num_tokens": 157982053.0, "step": 4156 }, { "epoch": 0.5288131281007505, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 5.245651721954346, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.856645941734314, "num_tokens": 158022654.0, "step": 4157 }, { "epoch": 0.5289403383793411, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00011968612670898438, "ewc_loss_parallel": 5.817413330078125e-05, "grad_norm": 3.020439386367798, "learning_rate": 1e-05, "loss": 0.571, "mean_token_accuracy": 0.8347810506820679, "num_tokens": 158060508.0, "step": 4158 }, { "epoch": 0.5290675486579316, "ewc_loss": 0.0458984375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 3.361701965332031e-05, "grad_norm": 2.4067821502685547, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8454152345657349, "num_tokens": 158098956.0, "step": 4159 }, { "epoch": 0.5291947589365221, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6088309288024902, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8531848192214966, "num_tokens": 158134781.0, "step": 4160 }, { "epoch": 0.5293219692151125, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4910948276519775, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8648282885551453, "num_tokens": 158177495.0, "step": 4161 }, { "epoch": 0.5294491794937031, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.564448833465576, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8546297550201416, "num_tokens": 158216672.0, "step": 4162 }, { "epoch": 0.5295763897722936, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012063980102539062, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4637556076049805, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8488876223564148, "num_tokens": 158255397.0, "step": 4163 }, { "epoch": 0.5297036000508841, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5829262733459473, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8500438928604126, "num_tokens": 158292230.0, "step": 4164 }, { "epoch": 0.5298308103294747, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4447731971740723, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8588178157806396, "num_tokens": 158329884.0, "step": 4165 }, { "epoch": 0.5299580206080652, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.585442066192627, "learning_rate": 1e-05, "loss": 0.5734, "mean_token_accuracy": 0.8314160108566284, "num_tokens": 158367290.0, "step": 4166 }, { "epoch": 0.5300852308866556, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3973686695098877, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.877855122089386, "num_tokens": 158407999.0, "step": 4167 }, { "epoch": 0.5302124411652461, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5901644229888916, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8706778287887573, "num_tokens": 158444283.0, "step": 4168 }, { "epoch": 0.5303396514438367, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3877854347229004, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8550428152084351, "num_tokens": 158484807.0, "step": 4169 }, { "epoch": 0.5304668617224272, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6144304275512695, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8627703189849854, "num_tokens": 158518863.0, "step": 4170 }, { "epoch": 0.5305940720010177, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4370763301849365, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8641780614852905, "num_tokens": 158555057.0, "step": 4171 }, { "epoch": 0.5307212822796082, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.59340238571167, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8602467179298401, "num_tokens": 158596032.0, "step": 4172 }, { "epoch": 0.5308484925581987, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00012159347534179688, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.408750057220459, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8483123183250427, "num_tokens": 158631098.0, "step": 4173 }, { "epoch": 0.5309757028367892, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.607161283493042, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8490330576896667, "num_tokens": 158664579.0, "step": 4174 }, { "epoch": 0.5311029131153797, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3930704593658447, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8483877778053284, "num_tokens": 158701220.0, "step": 4175 }, { "epoch": 0.5312301233939702, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.59338641166687, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8647290468215942, "num_tokens": 158744235.0, "step": 4176 }, { "epoch": 0.5313573336725608, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.382718563079834, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.860572874546051, "num_tokens": 158782602.0, "step": 4177 }, { "epoch": 0.5314845439511513, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6257524490356445, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8626528382301331, "num_tokens": 158813345.0, "step": 4178 }, { "epoch": 0.5316117542297417, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4091217517852783, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8439292907714844, "num_tokens": 158847204.0, "step": 4179 }, { "epoch": 0.5317389645083322, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.625276565551758, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8538610935211182, "num_tokens": 158889481.0, "step": 4180 }, { "epoch": 0.5318661747869228, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3861889839172363, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8660480976104736, "num_tokens": 158929267.0, "step": 4181 }, { "epoch": 0.5319933850655133, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6247875690460205, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8456870317459106, "num_tokens": 158974526.0, "step": 4182 }, { "epoch": 0.5321205953441038, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3764495849609375, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8622136116027832, "num_tokens": 159015428.0, "step": 4183 }, { "epoch": 0.5322478056226944, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6598455905914307, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.845663845539093, "num_tokens": 159050492.0, "step": 4184 }, { "epoch": 0.5323750159012848, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3962459564208984, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8474010229110718, "num_tokens": 159085644.0, "step": 4185 }, { "epoch": 0.5325022261798753, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.669362783432007, "learning_rate": 1e-05, "loss": 0.5357, "mean_token_accuracy": 0.8442554473876953, "num_tokens": 159123093.0, "step": 4186 }, { "epoch": 0.5326294364584658, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.407982349395752, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8555414080619812, "num_tokens": 159157688.0, "step": 4187 }, { "epoch": 0.5327566467370564, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.706381320953369, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8510161638259888, "num_tokens": 159194473.0, "step": 4188 }, { "epoch": 0.5328838570156469, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.406687021255493, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8512811660766602, "num_tokens": 159230314.0, "step": 4189 }, { "epoch": 0.5330110672942374, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6723642349243164, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8637569546699524, "num_tokens": 159266579.0, "step": 4190 }, { "epoch": 0.5331382775728278, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3920793533325195, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8673141002655029, "num_tokens": 159306502.0, "step": 4191 }, { "epoch": 0.5332654878514184, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6934714317321777, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.840385913848877, "num_tokens": 159342512.0, "step": 4192 }, { "epoch": 0.5333926981300089, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4145970344543457, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8620749115943909, "num_tokens": 159378408.0, "step": 4193 }, { "epoch": 0.5335199084085994, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6946218013763428, "learning_rate": 1e-05, "loss": 0.5474, "mean_token_accuracy": 0.8369998335838318, "num_tokens": 159415843.0, "step": 4194 }, { "epoch": 0.53364711868719, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4447779655456543, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8566358089447021, "num_tokens": 159446880.0, "step": 4195 }, { "epoch": 0.5337743289657805, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6538965702056885, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8597609400749207, "num_tokens": 159491960.0, "step": 4196 }, { "epoch": 0.5339015392443709, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4241445064544678, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8607268333435059, "num_tokens": 159529157.0, "step": 4197 }, { "epoch": 0.5340287495229614, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6616404056549072, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8527398109436035, "num_tokens": 159563589.0, "step": 4198 }, { "epoch": 0.534155959801552, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4145264625549316, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8759483098983765, "num_tokens": 159601789.0, "step": 4199 }, { "epoch": 0.5342831700801425, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.668417453765869, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8338767290115356, "num_tokens": 159637357.0, "step": 4200 }, { "epoch": 0.534410380358733, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4122307300567627, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8705997467041016, "num_tokens": 159678919.0, "step": 4201 }, { "epoch": 0.5345375906373235, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6671884059906006, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.852073609828949, "num_tokens": 159714109.0, "step": 4202 }, { "epoch": 0.534664800915914, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4191956520080566, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8649603128433228, "num_tokens": 159752171.0, "step": 4203 }, { "epoch": 0.5347920111945045, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6562108993530273, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8475354909896851, "num_tokens": 159791251.0, "step": 4204 }, { "epoch": 0.534919221473095, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4383347034454346, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8668063879013062, "num_tokens": 159819925.0, "step": 4205 }, { "epoch": 0.5350464317516855, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6379597187042236, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8505960702896118, "num_tokens": 159859788.0, "step": 4206 }, { "epoch": 0.5351736420302761, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.437633514404297, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8492641448974609, "num_tokens": 159897202.0, "step": 4207 }, { "epoch": 0.5353008523088666, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6192331314086914, "learning_rate": 1e-05, "loss": 0.5089, "mean_token_accuracy": 0.8460750579833984, "num_tokens": 159935840.0, "step": 4208 }, { "epoch": 0.5354280625874571, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.423227071762085, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8669406175613403, "num_tokens": 159975120.0, "step": 4209 }, { "epoch": 0.5355552728660475, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5933830738067627, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8637557029724121, "num_tokens": 160020935.0, "step": 4210 }, { "epoch": 0.5356824831446381, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.420531749725342, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8707467317581177, "num_tokens": 160065029.0, "step": 4211 }, { "epoch": 0.5358096934232286, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6035091876983643, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8481212854385376, "num_tokens": 160104865.0, "step": 4212 }, { "epoch": 0.5359369037018191, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.413496732711792, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.866710901260376, "num_tokens": 160145477.0, "step": 4213 }, { "epoch": 0.5360641139804097, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.608306407928467, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8492295742034912, "num_tokens": 160184906.0, "step": 4214 }, { "epoch": 0.5361913242590002, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.412992238998413, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8531041145324707, "num_tokens": 160225011.0, "step": 4215 }, { "epoch": 0.5363185345375906, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.600351333618164, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8578636646270752, "num_tokens": 160263193.0, "step": 4216 }, { "epoch": 0.5364457448161811, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.400357961654663, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8819689154624939, "num_tokens": 160299552.0, "step": 4217 }, { "epoch": 0.5365729550947717, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6114768981933594, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8688910007476807, "num_tokens": 160334609.0, "step": 4218 }, { "epoch": 0.5367001653733622, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3914644718170166, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.862699031829834, "num_tokens": 160374295.0, "step": 4219 }, { "epoch": 0.5368273756519527, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6457457542419434, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8410319685935974, "num_tokens": 160404420.0, "step": 4220 }, { "epoch": 0.5369545859305432, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.391676425933838, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8548212051391602, "num_tokens": 160439278.0, "step": 4221 }, { "epoch": 0.5370817962091337, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.634006977081299, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8742312788963318, "num_tokens": 160475225.0, "step": 4222 }, { "epoch": 0.5372090064877242, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3973541259765625, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8605307340621948, "num_tokens": 160516073.0, "step": 4223 }, { "epoch": 0.5373362167663147, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6716456413269043, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8532098531723022, "num_tokens": 160550938.0, "step": 4224 }, { "epoch": 0.5374634270449052, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3914973735809326, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8652826547622681, "num_tokens": 160590382.0, "step": 4225 }, { "epoch": 0.5375906373234958, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.671051502227783, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8704344034194946, "num_tokens": 160624830.0, "step": 4226 }, { "epoch": 0.5377178476020863, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4055240154266357, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8544738292694092, "num_tokens": 160663042.0, "step": 4227 }, { "epoch": 0.5378450578806767, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6779675483703613, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8486915826797485, "num_tokens": 160700502.0, "step": 4228 }, { "epoch": 0.5379722681592672, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4264440536499023, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8418058753013611, "num_tokens": 160736746.0, "step": 4229 }, { "epoch": 0.5380994784378578, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6599221229553223, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8667833805084229, "num_tokens": 160771430.0, "step": 4230 }, { "epoch": 0.5382266887164483, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.428776264190674, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8578228950500488, "num_tokens": 160807047.0, "step": 4231 }, { "epoch": 0.5383538989950388, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6487154960632324, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8388057947158813, "num_tokens": 160844600.0, "step": 4232 }, { "epoch": 0.5384811092736294, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4177072048187256, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.861158549785614, "num_tokens": 160883132.0, "step": 4233 }, { "epoch": 0.5386083195522198, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6364059448242188, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8590258359909058, "num_tokens": 160922408.0, "step": 4234 }, { "epoch": 0.5387355298308103, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.448223114013672, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8619082570075989, "num_tokens": 160960563.0, "step": 4235 }, { "epoch": 0.5388627401094008, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6198949813842773, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8574680089950562, "num_tokens": 160995594.0, "step": 4236 }, { "epoch": 0.5389899503879914, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4452731609344482, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.854102611541748, "num_tokens": 161028341.0, "step": 4237 }, { "epoch": 0.5391171606665819, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.590813636779785, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8599028587341309, "num_tokens": 161064549.0, "step": 4238 }, { "epoch": 0.5392443709451724, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.433242082595825, "learning_rate": 1e-05, "loss": 0.5173, "mean_token_accuracy": 0.8430501818656921, "num_tokens": 161105170.0, "step": 4239 }, { "epoch": 0.5393715812237628, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.594961404800415, "learning_rate": 1e-05, "loss": 0.5341, "mean_token_accuracy": 0.8387340307235718, "num_tokens": 161141018.0, "step": 4240 }, { "epoch": 0.5394987915023534, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4425792694091797, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8419546484947205, "num_tokens": 161175363.0, "step": 4241 }, { "epoch": 0.5396260017809439, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.571186065673828, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8630630373954773, "num_tokens": 161210920.0, "step": 4242 }, { "epoch": 0.5397532120595344, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4132819175720215, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8702032566070557, "num_tokens": 161248743.0, "step": 4243 }, { "epoch": 0.5398804223381249, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5732204914093018, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8610022068023682, "num_tokens": 161282897.0, "step": 4244 }, { "epoch": 0.5400076326167155, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4125635623931885, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8583452105522156, "num_tokens": 161323103.0, "step": 4245 }, { "epoch": 0.5401348428953059, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5885281562805176, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8569965958595276, "num_tokens": 161359612.0, "step": 4246 }, { "epoch": 0.5402620531738964, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.401932716369629, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8701715469360352, "num_tokens": 161396961.0, "step": 4247 }, { "epoch": 0.540389263452487, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5818567276000977, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8523034453392029, "num_tokens": 161437380.0, "step": 4248 }, { "epoch": 0.5405164737310775, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3922312259674072, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8547992706298828, "num_tokens": 161480237.0, "step": 4249 }, { "epoch": 0.540643684009668, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5819945335388184, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8725773096084595, "num_tokens": 161519447.0, "step": 4250 }, { "epoch": 0.5407708942882585, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.40317702293396, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8492230176925659, "num_tokens": 161556588.0, "step": 4251 }, { "epoch": 0.540898104566849, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6081795692443848, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.861988365650177, "num_tokens": 161591542.0, "step": 4252 }, { "epoch": 0.5410253148454395, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.390784978866577, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.858305811882019, "num_tokens": 161630453.0, "step": 4253 }, { "epoch": 0.54115252512403, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6198112964630127, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8817107677459717, "num_tokens": 161660927.0, "step": 4254 }, { "epoch": 0.5412797354026205, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3786356449127197, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8591569066047668, "num_tokens": 161703284.0, "step": 4255 }, { "epoch": 0.5414069456812111, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.639011859893799, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.846791684627533, "num_tokens": 161739212.0, "step": 4256 }, { "epoch": 0.5415341559598016, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 6.990549087524414, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8562085628509521, "num_tokens": 161777635.0, "step": 4257 }, { "epoch": 0.5416613662383921, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 6.771087646484375e-05, "grad_norm": 3.3328487873077393, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8519263863563538, "num_tokens": 161816019.0, "step": 4258 }, { "epoch": 0.5417885765169825, "ewc_loss": 0.044189453125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.1948089599609375e-05, "grad_norm": 2.170168399810791, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8601881861686707, "num_tokens": 161849690.0, "step": 4259 }, { "epoch": 0.5419157867955731, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 4.9591064453125e-05, "grad_norm": 2.9023241996765137, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.872390866279602, "num_tokens": 161889596.0, "step": 4260 }, { "epoch": 0.5420429970741636, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.418316125869751, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8390566110610962, "num_tokens": 161929451.0, "step": 4261 }, { "epoch": 0.5421702073527541, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.841697931289673, "learning_rate": 1e-05, "loss": 0.5147, "mean_token_accuracy": 0.8473775386810303, "num_tokens": 161966422.0, "step": 4262 }, { "epoch": 0.5422974176313446, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.465701103210449, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8543780446052551, "num_tokens": 162005192.0, "step": 4263 }, { "epoch": 0.5424246279099352, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8041632175445557, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8547645807266235, "num_tokens": 162051039.0, "step": 4264 }, { "epoch": 0.5425518381885256, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.501467227935791, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8623538017272949, "num_tokens": 162086869.0, "step": 4265 }, { "epoch": 0.5426790484671161, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7669143676757812, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.858849048614502, "num_tokens": 162122442.0, "step": 4266 }, { "epoch": 0.5428062587457066, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.513030529022217, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8606282472610474, "num_tokens": 162153057.0, "step": 4267 }, { "epoch": 0.5429334690242972, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.706515073776245, "learning_rate": 1e-05, "loss": 0.5387, "mean_token_accuracy": 0.8423344492912292, "num_tokens": 162192086.0, "step": 4268 }, { "epoch": 0.5430606793028877, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.4839890003204346, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8684227466583252, "num_tokens": 162227206.0, "step": 4269 }, { "epoch": 0.5431878895814782, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.659900188446045, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8501795530319214, "num_tokens": 162263490.0, "step": 4270 }, { "epoch": 0.5433150998600687, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4773621559143066, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.869596004486084, "num_tokens": 162302750.0, "step": 4271 }, { "epoch": 0.5434423101386592, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.618795871734619, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.873767614364624, "num_tokens": 162339613.0, "step": 4272 }, { "epoch": 0.5435695204172497, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4615721702575684, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8527213335037231, "num_tokens": 162380163.0, "step": 4273 }, { "epoch": 0.5436967306958402, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.617406129837036, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8530805110931396, "num_tokens": 162413306.0, "step": 4274 }, { "epoch": 0.5438239409744308, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4489362239837646, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8651211857795715, "num_tokens": 162446861.0, "step": 4275 }, { "epoch": 0.5439511512530213, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5896177291870117, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8509278297424316, "num_tokens": 162491414.0, "step": 4276 }, { "epoch": 0.5440783615316117, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4206976890563965, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8643679618835449, "num_tokens": 162535265.0, "step": 4277 }, { "epoch": 0.5442055718102022, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.59599232673645, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8470656871795654, "num_tokens": 162577198.0, "step": 4278 }, { "epoch": 0.5443327820887928, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.427635908126831, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8626487255096436, "num_tokens": 162609557.0, "step": 4279 }, { "epoch": 0.5444599923673833, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.598775625228882, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8624984622001648, "num_tokens": 162642943.0, "step": 4280 }, { "epoch": 0.5445872026459738, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.425657033920288, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8591181039810181, "num_tokens": 162676444.0, "step": 4281 }, { "epoch": 0.5447144129245644, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.583275318145752, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8534859418869019, "num_tokens": 162715445.0, "step": 4282 }, { "epoch": 0.5448416232031548, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.422492504119873, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8421746492385864, "num_tokens": 162751306.0, "step": 4283 }, { "epoch": 0.5449688334817453, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5758609771728516, "learning_rate": 1e-05, "loss": 0.5638, "mean_token_accuracy": 0.8275703191757202, "num_tokens": 162793587.0, "step": 4284 }, { "epoch": 0.5450960437603358, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.421149492263794, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8417937755584717, "num_tokens": 162831000.0, "step": 4285 }, { "epoch": 0.5452232540389264, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.57071590423584, "learning_rate": 1e-05, "loss": 0.3886, "mean_token_accuracy": 0.8820297122001648, "num_tokens": 162863913.0, "step": 4286 }, { "epoch": 0.5453504643175169, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4143569469451904, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.861855149269104, "num_tokens": 162899135.0, "step": 4287 }, { "epoch": 0.5454776745961074, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5821926593780518, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8695420026779175, "num_tokens": 162932534.0, "step": 4288 }, { "epoch": 0.5456048848746978, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3987033367156982, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8596085906028748, "num_tokens": 162973199.0, "step": 4289 }, { "epoch": 0.5457320951532884, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6081583499908447, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8541387915611267, "num_tokens": 163007018.0, "step": 4290 }, { "epoch": 0.5458593054318789, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.394012451171875, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8539462089538574, "num_tokens": 163044381.0, "step": 4291 }, { "epoch": 0.5459865157104694, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.617614984512329, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8574275970458984, "num_tokens": 163079606.0, "step": 4292 }, { "epoch": 0.5461137259890599, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.384176731109619, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8539618253707886, "num_tokens": 163120421.0, "step": 4293 }, { "epoch": 0.5462409362676505, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.62808895111084, "learning_rate": 1e-05, "loss": 0.5313, "mean_token_accuracy": 0.8395530581474304, "num_tokens": 163164761.0, "step": 4294 }, { "epoch": 0.5463681465462409, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.389662981033325, "learning_rate": 1e-05, "loss": 0.6168, "mean_token_accuracy": 0.8139549493789673, "num_tokens": 163212432.0, "step": 4295 }, { "epoch": 0.5464953568248314, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6346895694732666, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8741260766983032, "num_tokens": 163251333.0, "step": 4296 }, { "epoch": 0.5466225671034219, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4023821353912354, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8553732633590698, "num_tokens": 163286428.0, "step": 4297 }, { "epoch": 0.5467497773820125, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.652432680130005, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8682473301887512, "num_tokens": 163318974.0, "step": 4298 }, { "epoch": 0.546876987660603, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.426206350326538, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8506931662559509, "num_tokens": 163358621.0, "step": 4299 }, { "epoch": 0.5470041979391935, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.622631549835205, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8656858801841736, "num_tokens": 163401062.0, "step": 4300 }, { "epoch": 0.5471314082177839, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4206676483154297, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8531405329704285, "num_tokens": 163436194.0, "step": 4301 }, { "epoch": 0.5472586184963745, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6230828762054443, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8509693145751953, "num_tokens": 163476403.0, "step": 4302 }, { "epoch": 0.547385828774965, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.412851572036743, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8535467386245728, "num_tokens": 163515822.0, "step": 4303 }, { "epoch": 0.5475130390535555, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6310062408447266, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8462381958961487, "num_tokens": 163553611.0, "step": 4304 }, { "epoch": 0.5476402493321461, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4108121395111084, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8637347221374512, "num_tokens": 163588591.0, "step": 4305 }, { "epoch": 0.5477674596107366, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6178016662597656, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8582309484481812, "num_tokens": 163630857.0, "step": 4306 }, { "epoch": 0.5478946698893271, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4131298065185547, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8476229310035706, "num_tokens": 163666684.0, "step": 4307 }, { "epoch": 0.5480218801679175, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6146199703216553, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8542284965515137, "num_tokens": 163707409.0, "step": 4308 }, { "epoch": 0.5481490904465081, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4040820598602295, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8630868196487427, "num_tokens": 163745190.0, "step": 4309 }, { "epoch": 0.5482763007250986, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6180408000946045, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8563381433486938, "num_tokens": 163781538.0, "step": 4310 }, { "epoch": 0.5484035110036891, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4086525440216064, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8555657267570496, "num_tokens": 163816412.0, "step": 4311 }, { "epoch": 0.5485307212822796, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6201229095458984, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8653637170791626, "num_tokens": 163853202.0, "step": 4312 }, { "epoch": 0.5486579315608702, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.395742177963257, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8541090488433838, "num_tokens": 163893669.0, "step": 4313 }, { "epoch": 0.5487851418394606, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.620783567428589, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8595259189605713, "num_tokens": 163932372.0, "step": 4314 }, { "epoch": 0.5489123521180511, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.398876667022705, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8695718050003052, "num_tokens": 163969431.0, "step": 4315 }, { "epoch": 0.5490395623966416, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6361706256866455, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8566699028015137, "num_tokens": 164006776.0, "step": 4316 }, { "epoch": 0.5491667726752322, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3926455974578857, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.871125340461731, "num_tokens": 164042178.0, "step": 4317 }, { "epoch": 0.5492939829538227, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.635986328125, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8488026857376099, "num_tokens": 164089028.0, "step": 4318 }, { "epoch": 0.5494211932324132, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.39585018157959, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.85638028383255, "num_tokens": 164130311.0, "step": 4319 }, { "epoch": 0.5495484035110036, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6578354835510254, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8551104068756104, "num_tokens": 164165398.0, "step": 4320 }, { "epoch": 0.5496756137895942, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.401754140853882, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8681563138961792, "num_tokens": 164202724.0, "step": 4321 }, { "epoch": 0.5498028240681847, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.679835319519043, "learning_rate": 1e-05, "loss": 0.5269, "mean_token_accuracy": 0.8415555953979492, "num_tokens": 164247666.0, "step": 4322 }, { "epoch": 0.5499300343467752, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.392136812210083, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8555395007133484, "num_tokens": 164292323.0, "step": 4323 }, { "epoch": 0.5500572446253658, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6778547763824463, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8722615838050842, "num_tokens": 164331160.0, "step": 4324 }, { "epoch": 0.5501844549039563, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.410449743270874, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8728233575820923, "num_tokens": 164365457.0, "step": 4325 }, { "epoch": 0.5503116651825467, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6736655235290527, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8507933020591736, "num_tokens": 164403731.0, "step": 4326 }, { "epoch": 0.5504388754611372, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4238905906677246, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8617594242095947, "num_tokens": 164436616.0, "step": 4327 }, { "epoch": 0.5505660857397278, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.661956548690796, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8614035844802856, "num_tokens": 164472453.0, "step": 4328 }, { "epoch": 0.5506932960183183, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.436788320541382, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8499854207038879, "num_tokens": 164506551.0, "step": 4329 }, { "epoch": 0.5508205062969088, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6444435119628906, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8715864419937134, "num_tokens": 164542638.0, "step": 4330 }, { "epoch": 0.5509477165754993, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4349794387817383, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8590613007545471, "num_tokens": 164580071.0, "step": 4331 }, { "epoch": 0.5510749268540898, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6367883682250977, "learning_rate": 1e-05, "loss": 0.5196, "mean_token_accuracy": 0.8433825373649597, "num_tokens": 164617894.0, "step": 4332 }, { "epoch": 0.5512021371326803, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4262070655822754, "learning_rate": 1e-05, "loss": 0.5463, "mean_token_accuracy": 0.8354077339172363, "num_tokens": 164660509.0, "step": 4333 }, { "epoch": 0.5513293474112708, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6155335903167725, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.869315505027771, "num_tokens": 164699683.0, "step": 4334 }, { "epoch": 0.5514565576898613, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4374120235443115, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8558909296989441, "num_tokens": 164733134.0, "step": 4335 }, { "epoch": 0.5515837679684519, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6157116889953613, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8514217138290405, "num_tokens": 164771911.0, "step": 4336 }, { "epoch": 0.5517109782470424, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4267427921295166, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8480000495910645, "num_tokens": 164810579.0, "step": 4337 }, { "epoch": 0.5518381885256328, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.592471122741699, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8667193055152893, "num_tokens": 164850215.0, "step": 4338 }, { "epoch": 0.5519653988042234, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4174532890319824, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.852362871170044, "num_tokens": 164889523.0, "step": 4339 }, { "epoch": 0.5520926090828139, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.593513250350952, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8731726408004761, "num_tokens": 164925946.0, "step": 4340 }, { "epoch": 0.5522198193614044, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.5674099922180176, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8655083179473877, "num_tokens": 164961954.0, "step": 4341 }, { "epoch": 0.5523470296399949, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.49381947517395, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8604305982589722, "num_tokens": 165008013.0, "step": 4342 }, { "epoch": 0.5524742399185855, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.496917724609375, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.862675666809082, "num_tokens": 165037723.0, "step": 4343 }, { "epoch": 0.5526014501971759, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.472368001937866, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8631571531295776, "num_tokens": 165071432.0, "step": 4344 }, { "epoch": 0.5527286604757664, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4749157428741455, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8655148148536682, "num_tokens": 165118784.0, "step": 4345 }, { "epoch": 0.5528558707543569, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4593939781188965, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8414716720581055, "num_tokens": 165163281.0, "step": 4346 }, { "epoch": 0.5529830810329475, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.493316411972046, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8509364128112793, "num_tokens": 165195615.0, "step": 4347 }, { "epoch": 0.553110291311538, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4576785564422607, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8546346426010132, "num_tokens": 165233710.0, "step": 4348 }, { "epoch": 0.5532375015901285, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4784810543060303, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8549342751502991, "num_tokens": 165275788.0, "step": 4349 }, { "epoch": 0.5533647118687189, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4503393173217773, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8663558959960938, "num_tokens": 165316627.0, "step": 4350 }, { "epoch": 0.5534919221473095, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4829342365264893, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8584465384483337, "num_tokens": 165356643.0, "step": 4351 }, { "epoch": 0.5536191324259, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4425885677337646, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8522933721542358, "num_tokens": 165396897.0, "step": 4352 }, { "epoch": 0.5537463427044905, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4862966537475586, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8630639314651489, "num_tokens": 165434409.0, "step": 4353 }, { "epoch": 0.553873552983081, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.443133592605591, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8574659824371338, "num_tokens": 165474674.0, "step": 4354 }, { "epoch": 0.5540007632616716, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5012500286102295, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.846786379814148, "num_tokens": 165511865.0, "step": 4355 }, { "epoch": 0.554127973540262, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4407222270965576, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8558353781700134, "num_tokens": 165548589.0, "step": 4356 }, { "epoch": 0.5542551838188525, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001220703125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4875144958496094, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8625601530075073, "num_tokens": 165589553.0, "step": 4357 }, { "epoch": 0.554382394097443, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4542503356933594, "learning_rate": 1e-05, "loss": 0.5628, "mean_token_accuracy": 0.832983136177063, "num_tokens": 165625499.0, "step": 4358 }, { "epoch": 0.5545096043760336, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.480897903442383, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8589669466018677, "num_tokens": 165663967.0, "step": 4359 }, { "epoch": 0.5546368146546241, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.433194398880005, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8526716232299805, "num_tokens": 165706382.0, "step": 4360 }, { "epoch": 0.5547640249332146, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.496025323867798, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.860201358795166, "num_tokens": 165742473.0, "step": 4361 }, { "epoch": 0.5548912352118052, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012302398681640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4429686069488525, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8523000478744507, "num_tokens": 165776329.0, "step": 4362 }, { "epoch": 0.5550184454903956, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.499361991882324, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8528632521629333, "num_tokens": 165812887.0, "step": 4363 }, { "epoch": 0.5551456557689861, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001239776611328125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.427626848220825, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8682279586791992, "num_tokens": 165848933.0, "step": 4364 }, { "epoch": 0.5552728660475766, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5179429054260254, "learning_rate": 1e-05, "loss": 0.5189, "mean_token_accuracy": 0.8435996770858765, "num_tokens": 165887158.0, "step": 4365 }, { "epoch": 0.5554000763261672, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4130825996398926, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8564456105232239, "num_tokens": 165928214.0, "step": 4366 }, { "epoch": 0.5555272866047577, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5536410808563232, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8560988306999207, "num_tokens": 165964838.0, "step": 4367 }, { "epoch": 0.5556544968833482, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.4112184047698975, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8481049537658691, "num_tokens": 165996097.0, "step": 4368 }, { "epoch": 0.5557817071619386, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.566770553588867, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8563315868377686, "num_tokens": 166028660.0, "step": 4369 }, { "epoch": 0.5559089174405292, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.391295909881592, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8557084798812866, "num_tokens": 166065259.0, "step": 4370 }, { "epoch": 0.5560361277191197, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5774269104003906, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8534107804298401, "num_tokens": 166105757.0, "step": 4371 }, { "epoch": 0.5561633379977102, "ewc_loss": 0.04736328125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.4809112548828125e-05, "grad_norm": 2.38850998878479, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8419051170349121, "num_tokens": 166142362.0, "step": 4372 }, { "epoch": 0.5562905482763008, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5969722270965576, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8573127388954163, "num_tokens": 166179437.0, "step": 4373 }, { "epoch": 0.5564177585548913, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3788540363311768, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.850346565246582, "num_tokens": 166218555.0, "step": 4374 }, { "epoch": 0.5565449688334817, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6267905235290527, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8662606477737427, "num_tokens": 166252297.0, "step": 4375 }, { "epoch": 0.5566721791120722, "ewc_loss": 0.047119140625, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.457069396972656e-05, "grad_norm": 2.3724472522735596, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8614516854286194, "num_tokens": 166289597.0, "step": 4376 }, { "epoch": 0.5567993893906628, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6599173545837402, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8470133543014526, "num_tokens": 166328655.0, "step": 4377 }, { "epoch": 0.5569265996692533, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3551065921783447, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8545123338699341, "num_tokens": 166371825.0, "step": 4378 }, { "epoch": 0.5570538099478438, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6942896842956543, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8562902212142944, "num_tokens": 166407084.0, "step": 4379 }, { "epoch": 0.5571810202264343, "ewc_loss": 0.047607421875, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.373246431350708, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8637255430221558, "num_tokens": 166442272.0, "step": 4380 }, { "epoch": 0.5573082305050248, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.731576919555664, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8487224578857422, "num_tokens": 166475015.0, "step": 4381 }, { "epoch": 0.5574354407836153, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.374969005584717, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8589301705360413, "num_tokens": 166512421.0, "step": 4382 }, { "epoch": 0.5575626510622058, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7507240772247314, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8493463397026062, "num_tokens": 166554248.0, "step": 4383 }, { "epoch": 0.5576898613407963, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.399311065673828, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8739852905273438, "num_tokens": 166590664.0, "step": 4384 }, { "epoch": 0.5578170716193869, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.77644419670105, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8553823828697205, "num_tokens": 166625615.0, "step": 4385 }, { "epoch": 0.5579442818979774, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4300003051757812, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8611676692962646, "num_tokens": 166662699.0, "step": 4386 }, { "epoch": 0.5580714921765678, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7578680515289307, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8637183904647827, "num_tokens": 166701218.0, "step": 4387 }, { "epoch": 0.5581987024551583, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4554076194763184, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8548521995544434, "num_tokens": 166739699.0, "step": 4388 }, { "epoch": 0.5583259127337489, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.721543073654175, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8648185729980469, "num_tokens": 166782275.0, "step": 4389 }, { "epoch": 0.5584531230123394, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4634201526641846, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8620516061782837, "num_tokens": 166820535.0, "step": 4390 }, { "epoch": 0.5585803332909299, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.696660041809082, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8429703712463379, "num_tokens": 166865450.0, "step": 4391 }, { "epoch": 0.5587075435695205, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012493133544921875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.461986541748047, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.86189866065979, "num_tokens": 166905838.0, "step": 4392 }, { "epoch": 0.5588347538481109, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6772167682647705, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.8437914848327637, "num_tokens": 166943437.0, "step": 4393 }, { "epoch": 0.5589619641267014, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4646575450897217, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8625290393829346, "num_tokens": 166980863.0, "step": 4394 }, { "epoch": 0.5590891744052919, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.639542579650879, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8734267950057983, "num_tokens": 167017861.0, "step": 4395 }, { "epoch": 0.5592163846838825, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4603655338287354, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8685610890388489, "num_tokens": 167051779.0, "step": 4396 }, { "epoch": 0.559343594962473, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.606348991394043, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8586278557777405, "num_tokens": 167095205.0, "step": 4397 }, { "epoch": 0.5594708052410635, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4536592960357666, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8516063690185547, "num_tokens": 167130594.0, "step": 4398 }, { "epoch": 0.5595980155196539, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.591280937194824, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8579165935516357, "num_tokens": 167170816.0, "step": 4399 }, { "epoch": 0.5597252257982445, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4460346698760986, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8683253526687622, "num_tokens": 167205301.0, "step": 4400 }, { "epoch": 0.559852436076835, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5889604091644287, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8420857191085815, "num_tokens": 167240356.0, "step": 4401 }, { "epoch": 0.5599796463554255, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.45009708404541, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8632760643959045, "num_tokens": 167269853.0, "step": 4402 }, { "epoch": 0.560106856634016, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.559784173965454, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8605129718780518, "num_tokens": 167315398.0, "step": 4403 }, { "epoch": 0.5602340669126066, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4282634258270264, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.860397458076477, "num_tokens": 167352715.0, "step": 4404 }, { "epoch": 0.560361277191197, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.000125885009765625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.564296245574951, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8439334034919739, "num_tokens": 167391646.0, "step": 4405 }, { "epoch": 0.5604884874697875, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4254913330078125, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8544727563858032, "num_tokens": 167431676.0, "step": 4406 }, { "epoch": 0.560615697748378, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.568967342376709, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8518776297569275, "num_tokens": 167469121.0, "step": 4407 }, { "epoch": 0.5607429080269686, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4060869216918945, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8626574277877808, "num_tokens": 167511721.0, "step": 4408 }, { "epoch": 0.5608701183055591, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.58186411857605, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8656700849533081, "num_tokens": 167546965.0, "step": 4409 }, { "epoch": 0.5609973285841496, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.413914203643799, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.845223069190979, "num_tokens": 167584968.0, "step": 4410 }, { "epoch": 0.5611245388627402, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5879030227661133, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8450913429260254, "num_tokens": 167623124.0, "step": 4411 }, { "epoch": 0.5612517491413306, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3950421810150146, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8536610007286072, "num_tokens": 167664877.0, "step": 4412 }, { "epoch": 0.5613789594199211, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5970847606658936, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8507407903671265, "num_tokens": 167706154.0, "step": 4413 }, { "epoch": 0.5615061696985116, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4065377712249756, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8544167280197144, "num_tokens": 167740081.0, "step": 4414 }, { "epoch": 0.5616333799771022, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.598268747329712, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8401352167129517, "num_tokens": 167783324.0, "step": 4415 }, { "epoch": 0.5617605902556927, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3885676860809326, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8607927560806274, "num_tokens": 167823216.0, "step": 4416 }, { "epoch": 0.5618878005342832, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6337316036224365, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.8398903608322144, "num_tokens": 167857452.0, "step": 4417 }, { "epoch": 0.5620150108128736, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.38097882270813, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8635608553886414, "num_tokens": 167896402.0, "step": 4418 }, { "epoch": 0.5621422210914642, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.641388177871704, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8600561022758484, "num_tokens": 167936441.0, "step": 4419 }, { "epoch": 0.5622694313700547, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.393388509750366, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8672120571136475, "num_tokens": 167970786.0, "step": 4420 }, { "epoch": 0.5623966416486452, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.659240245819092, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.855760931968689, "num_tokens": 168011004.0, "step": 4421 }, { "epoch": 0.5625238519272358, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3855597972869873, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8554449081420898, "num_tokens": 168053418.0, "step": 4422 }, { "epoch": 0.5626510622058263, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00012683868408203125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6847732067108154, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8626293540000916, "num_tokens": 168088156.0, "step": 4423 }, { "epoch": 0.5627782724844167, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.393667221069336, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8470768928527832, "num_tokens": 168132329.0, "step": 4424 }, { "epoch": 0.5629054827630072, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6948580741882324, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8609586954116821, "num_tokens": 168167882.0, "step": 4425 }, { "epoch": 0.5630326930415978, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.408398389816284, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8583360910415649, "num_tokens": 168203265.0, "step": 4426 }, { "epoch": 0.5631599033201883, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.690723180770874, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8701499104499817, "num_tokens": 168241410.0, "step": 4427 }, { "epoch": 0.5632871135987788, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.406106472015381, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8633221387863159, "num_tokens": 168285412.0, "step": 4428 }, { "epoch": 0.5634143238773693, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.692782402038574, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8517134189605713, "num_tokens": 168324628.0, "step": 4429 }, { "epoch": 0.5635415341559598, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4570152759552, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8513798713684082, "num_tokens": 168359759.0, "step": 4430 }, { "epoch": 0.5636687444345503, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6722779273986816, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8533871173858643, "num_tokens": 168393904.0, "step": 4431 }, { "epoch": 0.5637959547131408, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4424169063568115, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8526569604873657, "num_tokens": 168429718.0, "step": 4432 }, { "epoch": 0.5639231649917313, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6490609645843506, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8680524826049805, "num_tokens": 168462950.0, "step": 4433 }, { "epoch": 0.5640503752703219, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4271657466888428, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8567754030227661, "num_tokens": 168504095.0, "step": 4434 }, { "epoch": 0.5641775855489124, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6445977687835693, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8543187379837036, "num_tokens": 168539931.0, "step": 4435 }, { "epoch": 0.5643047958275028, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4382379055023193, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8621888160705566, "num_tokens": 168573064.0, "step": 4436 }, { "epoch": 0.5644320061060933, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6343231201171875, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8575384616851807, "num_tokens": 168607997.0, "step": 4437 }, { "epoch": 0.5645592163846839, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.418762683868408, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8776794672012329, "num_tokens": 168647590.0, "step": 4438 }, { "epoch": 0.5646864266632744, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.64162540435791, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8620657324790955, "num_tokens": 168681347.0, "step": 4439 }, { "epoch": 0.5648136369418649, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4226126670837402, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8562406301498413, "num_tokens": 168722830.0, "step": 4440 }, { "epoch": 0.5649408472204555, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6354587078094482, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.8403617143630981, "num_tokens": 168762381.0, "step": 4441 }, { "epoch": 0.5650680574990459, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4118919372558594, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8617227673530579, "num_tokens": 168802832.0, "step": 4442 }, { "epoch": 0.5651952677776364, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.643005132675171, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8556278944015503, "num_tokens": 168837780.0, "step": 4443 }, { "epoch": 0.5653224780562269, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4441802501678467, "learning_rate": 1e-05, "loss": 0.5192, "mean_token_accuracy": 0.8466426730155945, "num_tokens": 168871830.0, "step": 4444 }, { "epoch": 0.5654496883348175, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6124844551086426, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8630079627037048, "num_tokens": 168912477.0, "step": 4445 }, { "epoch": 0.565576898613408, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.424835443496704, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8488633036613464, "num_tokens": 168948976.0, "step": 4446 }, { "epoch": 0.5657041088919985, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.603564739227295, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8576599359512329, "num_tokens": 168992454.0, "step": 4447 }, { "epoch": 0.5658313191705889, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4412264823913574, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8535372018814087, "num_tokens": 169023135.0, "step": 4448 }, { "epoch": 0.5659585294491795, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5853395462036133, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8577722311019897, "num_tokens": 169062066.0, "step": 4449 }, { "epoch": 0.56608573972777, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.430697202682495, "learning_rate": 1e-05, "loss": 0.5335, "mean_token_accuracy": 0.8397666215896606, "num_tokens": 169100317.0, "step": 4450 }, { "epoch": 0.5662129500063605, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.597755193710327, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8657245635986328, "num_tokens": 169128474.0, "step": 4451 }, { "epoch": 0.566340160284951, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4100801944732666, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8745601773262024, "num_tokens": 169164905.0, "step": 4452 }, { "epoch": 0.5664673705635416, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5948517322540283, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8536059856414795, "num_tokens": 169201153.0, "step": 4453 }, { "epoch": 0.566594580842132, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4024767875671387, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8717505931854248, "num_tokens": 169235325.0, "step": 4454 }, { "epoch": 0.5667217911207225, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6033761501312256, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8759206533432007, "num_tokens": 169276690.0, "step": 4455 }, { "epoch": 0.566849001399313, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4067838191986084, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.844023585319519, "num_tokens": 169315751.0, "step": 4456 }, { "epoch": 0.5669762116779036, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.603306770324707, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8586829304695129, "num_tokens": 169354626.0, "step": 4457 }, { "epoch": 0.5671034219564941, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.408172369003296, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8567638397216797, "num_tokens": 169390255.0, "step": 4458 }, { "epoch": 0.5672306322350846, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.599513530731201, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8493685722351074, "num_tokens": 169431912.0, "step": 4459 }, { "epoch": 0.5673578425136752, "ewc_loss": 0.0478515625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4004530906677246, "learning_rate": 1e-05, "loss": 0.3851, "mean_token_accuracy": 0.8821465969085693, "num_tokens": 169467803.0, "step": 4460 }, { "epoch": 0.5674850527922656, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.613762617111206, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8486446142196655, "num_tokens": 169512252.0, "step": 4461 }, { "epoch": 0.5676122630708561, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4078547954559326, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8623546361923218, "num_tokens": 169547632.0, "step": 4462 }, { "epoch": 0.5677394733494466, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6193928718566895, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8566805124282837, "num_tokens": 169589355.0, "step": 4463 }, { "epoch": 0.5678666836280372, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3999125957489014, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8652907609939575, "num_tokens": 169627441.0, "step": 4464 }, { "epoch": 0.5679938939066277, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6422691345214844, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8591587543487549, "num_tokens": 169659752.0, "step": 4465 }, { "epoch": 0.5681211041852182, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3914718627929688, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8749174475669861, "num_tokens": 169697197.0, "step": 4466 }, { "epoch": 0.5682483144638086, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.668133020401001, "learning_rate": 1e-05, "loss": 0.5356, "mean_token_accuracy": 0.8385591506958008, "num_tokens": 169732523.0, "step": 4467 }, { "epoch": 0.5683755247423992, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.397517204284668, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8545390367507935, "num_tokens": 169771494.0, "step": 4468 }, { "epoch": 0.5685027350209897, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6604726314544678, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8738771080970764, "num_tokens": 169809835.0, "step": 4469 }, { "epoch": 0.5686299452995802, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3971126079559326, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8518871068954468, "num_tokens": 169853276.0, "step": 4470 }, { "epoch": 0.5687571555781707, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6791157722473145, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.841001033782959, "num_tokens": 169892106.0, "step": 4471 }, { "epoch": 0.5688843658567613, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4108834266662598, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8496184945106506, "num_tokens": 169933635.0, "step": 4472 }, { "epoch": 0.5690115761353517, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.670400857925415, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8563879728317261, "num_tokens": 169971694.0, "step": 4473 }, { "epoch": 0.5691387864139422, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.41860032081604, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.840952455997467, "num_tokens": 170013285.0, "step": 4474 }, { "epoch": 0.5692659966925327, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.663093090057373, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8579615354537964, "num_tokens": 170052831.0, "step": 4475 }, { "epoch": 0.5693932069711233, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.430875301361084, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8430318832397461, "num_tokens": 170089859.0, "step": 4476 }, { "epoch": 0.5695204172497138, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6590681076049805, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8524466753005981, "num_tokens": 170126494.0, "step": 4477 }, { "epoch": 0.5696476275283043, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.422501802444458, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8497745990753174, "num_tokens": 170167744.0, "step": 4478 }, { "epoch": 0.5697748378068948, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.652989149093628, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8562322854995728, "num_tokens": 170203988.0, "step": 4479 }, { "epoch": 0.5699020480854853, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.440829277038574, "learning_rate": 1e-05, "loss": 0.5399, "mean_token_accuracy": 0.8349202275276184, "num_tokens": 170243231.0, "step": 4480 }, { "epoch": 0.5700292583640758, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6406514644622803, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8380236625671387, "num_tokens": 170280637.0, "step": 4481 }, { "epoch": 0.5701564686426663, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4387423992156982, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8484443426132202, "num_tokens": 170317106.0, "step": 4482 }, { "epoch": 0.5702836789212569, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.623663902282715, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8532642126083374, "num_tokens": 170353374.0, "step": 4483 }, { "epoch": 0.5704108891998474, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001277923583984375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.438429594039917, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8624951839447021, "num_tokens": 170386968.0, "step": 4484 }, { "epoch": 0.5705380994784378, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.610485076904297, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8625814914703369, "num_tokens": 170425386.0, "step": 4485 }, { "epoch": 0.5706653097570283, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4174342155456543, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8601914644241333, "num_tokens": 170467828.0, "step": 4486 }, { "epoch": 0.5707925200356189, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.620166063308716, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8552972674369812, "num_tokens": 170506333.0, "step": 4487 }, { "epoch": 0.5709197303142094, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4314985275268555, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8524157404899597, "num_tokens": 170542352.0, "step": 4488 }, { "epoch": 0.5710469405927999, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6133997440338135, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8715487718582153, "num_tokens": 170574251.0, "step": 4489 }, { "epoch": 0.5711741508713905, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.414128541946411, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.868209958076477, "num_tokens": 170610294.0, "step": 4490 }, { "epoch": 0.5713013611499809, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.612955093383789, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8612890243530273, "num_tokens": 170649245.0, "step": 4491 }, { "epoch": 0.5714285714285714, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4018936157226562, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8541425466537476, "num_tokens": 170692106.0, "step": 4492 }, { "epoch": 0.5715557817071619, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6254990100860596, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8643670082092285, "num_tokens": 170730319.0, "step": 4493 }, { "epoch": 0.5716829919857525, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4238662719726562, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8640890121459961, "num_tokens": 170763805.0, "step": 4494 }, { "epoch": 0.571810202264343, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6155312061309814, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8405236005783081, "num_tokens": 170804966.0, "step": 4495 }, { "epoch": 0.5719374125429335, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4101083278656006, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8617122173309326, "num_tokens": 170845652.0, "step": 4496 }, { "epoch": 0.5720646228215239, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6241965293884277, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8627582788467407, "num_tokens": 170883449.0, "step": 4497 }, { "epoch": 0.5721918331001145, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012874603271484375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4135262966156006, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8494727611541748, "num_tokens": 170920187.0, "step": 4498 }, { "epoch": 0.572319043378705, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.635118246078491, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8516433238983154, "num_tokens": 170952650.0, "step": 4499 }, { "epoch": 0.5724462536572955, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.40512752532959, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8519089221954346, "num_tokens": 170989130.0, "step": 4500 }, { "epoch": 0.572573463935886, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6152780055999756, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8602197170257568, "num_tokens": 171030713.0, "step": 4501 }, { "epoch": 0.5727006742144766, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.409777879714966, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8576333522796631, "num_tokens": 171066363.0, "step": 4502 }, { "epoch": 0.572827884493067, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.626727819442749, "learning_rate": 1e-05, "loss": 0.509, "mean_token_accuracy": 0.8517358899116516, "num_tokens": 171108290.0, "step": 4503 }, { "epoch": 0.5729550947716575, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4140214920043945, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8824285864830017, "num_tokens": 171141990.0, "step": 4504 }, { "epoch": 0.573082305050248, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6292762756347656, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8591086268424988, "num_tokens": 171180032.0, "step": 4505 }, { "epoch": 0.5732095153288386, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4191553592681885, "learning_rate": 1e-05, "loss": 0.5328, "mean_token_accuracy": 0.8391864895820618, "num_tokens": 171218918.0, "step": 4506 }, { "epoch": 0.5733367256074291, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.620150566101074, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8567643761634827, "num_tokens": 171261028.0, "step": 4507 }, { "epoch": 0.5734639358860196, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.407076358795166, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8597763180732727, "num_tokens": 171301477.0, "step": 4508 }, { "epoch": 0.5735911461646102, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.620260238647461, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.869540810585022, "num_tokens": 171338937.0, "step": 4509 }, { "epoch": 0.5737183564432006, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4176836013793945, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8556254506111145, "num_tokens": 171375058.0, "step": 4510 }, { "epoch": 0.5738455667217911, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6332037448883057, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8487077355384827, "num_tokens": 171410282.0, "step": 4511 }, { "epoch": 0.5739727770003816, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4232633113861084, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.855042576789856, "num_tokens": 171442864.0, "step": 4512 }, { "epoch": 0.5740999872789722, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.625123977661133, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8679907917976379, "num_tokens": 171476267.0, "step": 4513 }, { "epoch": 0.5742271975575627, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.411226749420166, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8550019860267639, "num_tokens": 171513173.0, "step": 4514 }, { "epoch": 0.5743544078361532, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6367461681365967, "learning_rate": 1e-05, "loss": 0.5594, "mean_token_accuracy": 0.8354296684265137, "num_tokens": 171555071.0, "step": 4515 }, { "epoch": 0.5744816181147436, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4097957611083984, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8563357591629028, "num_tokens": 171594263.0, "step": 4516 }, { "epoch": 0.5746088283933342, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6295382976531982, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8507973551750183, "num_tokens": 171635238.0, "step": 4517 }, { "epoch": 0.5747360386719247, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4051074981689453, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.851489782333374, "num_tokens": 171674537.0, "step": 4518 }, { "epoch": 0.5748632489505152, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.623408794403076, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8692225217819214, "num_tokens": 171718606.0, "step": 4519 }, { "epoch": 0.5749904592291057, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.409264326095581, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8609044551849365, "num_tokens": 171757092.0, "step": 4520 }, { "epoch": 0.5751176695076963, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6194677352905273, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8607645034790039, "num_tokens": 171799606.0, "step": 4521 }, { "epoch": 0.5752448797862867, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.423506259918213, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8649739027023315, "num_tokens": 171834686.0, "step": 4522 }, { "epoch": 0.5753720900648772, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.623220682144165, "learning_rate": 1e-05, "loss": 0.5436, "mean_token_accuracy": 0.8396115303039551, "num_tokens": 171870443.0, "step": 4523 }, { "epoch": 0.5754993003434677, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4111788272857666, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8620439767837524, "num_tokens": 171907615.0, "step": 4524 }, { "epoch": 0.5756265106220583, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.630824565887451, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8454891443252563, "num_tokens": 171943120.0, "step": 4525 }, { "epoch": 0.5757537209006488, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.406153678894043, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8645416498184204, "num_tokens": 171980637.0, "step": 4526 }, { "epoch": 0.5758809311792393, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.646352767944336, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8586177825927734, "num_tokens": 172014232.0, "step": 4527 }, { "epoch": 0.5760081414578297, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.402616262435913, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8561584949493408, "num_tokens": 172049639.0, "step": 4528 }, { "epoch": 0.5761353517364203, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.648887872695923, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8591576814651489, "num_tokens": 172086482.0, "step": 4529 }, { "epoch": 0.5762625620150108, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4259841442108154, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8642305135726929, "num_tokens": 172117220.0, "step": 4530 }, { "epoch": 0.5763897722936013, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6549971103668213, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8604354858398438, "num_tokens": 172149925.0, "step": 4531 }, { "epoch": 0.5765169825721919, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.41325306892395, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8688923716545105, "num_tokens": 172186903.0, "step": 4532 }, { "epoch": 0.5766441928507824, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.640754222869873, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8525115251541138, "num_tokens": 172226573.0, "step": 4533 }, { "epoch": 0.5767714031293728, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4247772693634033, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8625857830047607, "num_tokens": 172263639.0, "step": 4534 }, { "epoch": 0.5768986134079633, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.631208896636963, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8667973875999451, "num_tokens": 172297740.0, "step": 4535 }, { "epoch": 0.5770258236865539, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4183928966522217, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8488745093345642, "num_tokens": 172338243.0, "step": 4536 }, { "epoch": 0.5771530339651444, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6182358264923096, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8577708601951599, "num_tokens": 172383201.0, "step": 4537 }, { "epoch": 0.5772802442437349, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.428490161895752, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8684592247009277, "num_tokens": 172420587.0, "step": 4538 }, { "epoch": 0.5774074545223254, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6185786724090576, "learning_rate": 1e-05, "loss": 0.5206, "mean_token_accuracy": 0.8418608903884888, "num_tokens": 172459194.0, "step": 4539 }, { "epoch": 0.5775346648009159, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4411091804504395, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8387595415115356, "num_tokens": 172492463.0, "step": 4540 }, { "epoch": 0.5776618750795064, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6083407402038574, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8460124731063843, "num_tokens": 172531675.0, "step": 4541 }, { "epoch": 0.5777890853580969, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.41483998298645, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8579436540603638, "num_tokens": 172571666.0, "step": 4542 }, { "epoch": 0.5779162956366874, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.608907699584961, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8582956194877625, "num_tokens": 172608132.0, "step": 4543 }, { "epoch": 0.578043505915278, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.413836717605591, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8602254390716553, "num_tokens": 172650881.0, "step": 4544 }, { "epoch": 0.5781707161938685, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6238629817962646, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.861847996711731, "num_tokens": 172684632.0, "step": 4545 }, { "epoch": 0.5782979264724589, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3974335193634033, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.865719199180603, "num_tokens": 172724804.0, "step": 4546 }, { "epoch": 0.5784251367510495, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6318440437316895, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8549121618270874, "num_tokens": 172762687.0, "step": 4547 }, { "epoch": 0.57855234702964, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.393472194671631, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8661789894104004, "num_tokens": 172802885.0, "step": 4548 }, { "epoch": 0.5786795573082305, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6530568599700928, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.860133171081543, "num_tokens": 172836158.0, "step": 4549 }, { "epoch": 0.578806767586821, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4088175296783447, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8577463626861572, "num_tokens": 172872319.0, "step": 4550 }, { "epoch": 0.5789339778654116, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6617627143859863, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8627008199691772, "num_tokens": 172906803.0, "step": 4551 }, { "epoch": 0.579061188144002, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4038259983062744, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.859496533870697, "num_tokens": 172944095.0, "step": 4552 }, { "epoch": 0.5791883984225925, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.652111768722534, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8638756275177002, "num_tokens": 172980026.0, "step": 4553 }, { "epoch": 0.579315608701183, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4125425815582275, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8507281541824341, "num_tokens": 173019558.0, "step": 4554 }, { "epoch": 0.5794428189797736, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6522412300109863, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8619065880775452, "num_tokens": 173059067.0, "step": 4555 }, { "epoch": 0.5795700292583641, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.421818494796753, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8675016164779663, "num_tokens": 173096277.0, "step": 4556 }, { "epoch": 0.5796972395369546, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.653789758682251, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8450419902801514, "num_tokens": 173134746.0, "step": 4557 }, { "epoch": 0.5798244498155452, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4458940029144287, "learning_rate": 1e-05, "loss": 0.5398, "mean_token_accuracy": 0.8368027210235596, "num_tokens": 173164786.0, "step": 4558 }, { "epoch": 0.5799516600941356, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.629279375076294, "learning_rate": 1e-05, "loss": 0.4999, "mean_token_accuracy": 0.8511471748352051, "num_tokens": 173202948.0, "step": 4559 }, { "epoch": 0.5800788703727261, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.452519416809082, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8503952026367188, "num_tokens": 173235253.0, "step": 4560 }, { "epoch": 0.5802060806513166, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5911343097686768, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.853345513343811, "num_tokens": 173279878.0, "step": 4561 }, { "epoch": 0.5803332909299072, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.441315174102783, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.859779953956604, "num_tokens": 173316719.0, "step": 4562 }, { "epoch": 0.5804605012084977, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6046767234802246, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.863740086555481, "num_tokens": 173347077.0, "step": 4563 }, { "epoch": 0.5805877114870882, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4388270378112793, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.847444474697113, "num_tokens": 173386891.0, "step": 4564 }, { "epoch": 0.5807149217656786, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5785741806030273, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.858989953994751, "num_tokens": 173426919.0, "step": 4565 }, { "epoch": 0.5808421320442692, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.439591884613037, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8635321259498596, "num_tokens": 173461380.0, "step": 4566 }, { "epoch": 0.5809693423228597, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5642857551574707, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8588350415229797, "num_tokens": 173504680.0, "step": 4567 }, { "epoch": 0.5810965526014502, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.43874192237854, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8579319715499878, "num_tokens": 173541229.0, "step": 4568 }, { "epoch": 0.5812237628800407, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5681793689727783, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8553102612495422, "num_tokens": 173573805.0, "step": 4569 }, { "epoch": 0.5813509731586313, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4335532188415527, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8476150631904602, "num_tokens": 173613899.0, "step": 4570 }, { "epoch": 0.5814781834372217, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5595791339874268, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8613767623901367, "num_tokens": 173650590.0, "step": 4571 }, { "epoch": 0.5816053937158122, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4336931705474854, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8594056963920593, "num_tokens": 173684067.0, "step": 4572 }, { "epoch": 0.5817326039944027, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5601072311401367, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8626338243484497, "num_tokens": 173721846.0, "step": 4573 }, { "epoch": 0.5818598142729933, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.438985824584961, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8418920040130615, "num_tokens": 173754616.0, "step": 4574 }, { "epoch": 0.5819870245515838, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5546226501464844, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8543131351470947, "num_tokens": 173790208.0, "step": 4575 }, { "epoch": 0.5821142348301743, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.413633346557617, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.8473892211914062, "num_tokens": 173833132.0, "step": 4576 }, { "epoch": 0.5822414451087647, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5717294216156006, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8499464392662048, "num_tokens": 173870526.0, "step": 4577 }, { "epoch": 0.5823686553873553, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.429658889770508, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8533567190170288, "num_tokens": 173901561.0, "step": 4578 }, { "epoch": 0.5824958656659458, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5704619884490967, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8475453853607178, "num_tokens": 173941197.0, "step": 4579 }, { "epoch": 0.5826230759445363, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4082109928131104, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8541394472122192, "num_tokens": 173981755.0, "step": 4580 }, { "epoch": 0.5827502862231269, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5917820930480957, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8523109555244446, "num_tokens": 174015831.0, "step": 4581 }, { "epoch": 0.5828774965017174, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3995144367218018, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8633497953414917, "num_tokens": 174051623.0, "step": 4582 }, { "epoch": 0.5830047067803078, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6185033321380615, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8439240455627441, "num_tokens": 174086422.0, "step": 4583 }, { "epoch": 0.5831319170588983, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3885979652404785, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8517429828643799, "num_tokens": 174128557.0, "step": 4584 }, { "epoch": 0.5832591273374889, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6350419521331787, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8478793501853943, "num_tokens": 174169704.0, "step": 4585 }, { "epoch": 0.5833863376160794, "ewc_loss": 0.048095703125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3745875358581543, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8455795049667358, "num_tokens": 174213975.0, "step": 4586 }, { "epoch": 0.5835135478946699, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.663726806640625, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8554186224937439, "num_tokens": 174254425.0, "step": 4587 }, { "epoch": 0.5836407581732604, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3894011974334717, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8492612242698669, "num_tokens": 174292411.0, "step": 4588 }, { "epoch": 0.5837679684518509, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00012969970703125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.679335355758667, "learning_rate": 1e-05, "loss": 0.5362, "mean_token_accuracy": 0.8393427133560181, "num_tokens": 174335103.0, "step": 4589 }, { "epoch": 0.5838951787304414, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3829174041748047, "learning_rate": 1e-05, "loss": 0.5158, "mean_token_accuracy": 0.8474091291427612, "num_tokens": 174377984.0, "step": 4590 }, { "epoch": 0.5840223890090319, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.7192790508270264, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8547369241714478, "num_tokens": 174408211.0, "step": 4591 }, { "epoch": 0.5841495992876224, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4134275913238525, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8414921760559082, "num_tokens": 174444137.0, "step": 4592 }, { "epoch": 0.584276809566213, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.701467990875244, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8491197228431702, "num_tokens": 174488382.0, "step": 4593 }, { "epoch": 0.5844040198448035, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4109246730804443, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8714431524276733, "num_tokens": 174524436.0, "step": 4594 }, { "epoch": 0.5845312301233939, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7021822929382324, "learning_rate": 1e-05, "loss": 0.5287, "mean_token_accuracy": 0.8455580472946167, "num_tokens": 174563830.0, "step": 4595 }, { "epoch": 0.5846584404019844, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.429490089416504, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8555710911750793, "num_tokens": 174598600.0, "step": 4596 }, { "epoch": 0.584785650680575, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7100794315338135, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8661073446273804, "num_tokens": 174631234.0, "step": 4597 }, { "epoch": 0.5849128609591655, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4246010780334473, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8553089499473572, "num_tokens": 174673933.0, "step": 4598 }, { "epoch": 0.585040071237756, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6926844120025635, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8581669926643372, "num_tokens": 174715801.0, "step": 4599 }, { "epoch": 0.5851672815163466, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4306795597076416, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8554290533065796, "num_tokens": 174757309.0, "step": 4600 }, { "epoch": 0.585294491794937, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.694959878921509, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8732926249504089, "num_tokens": 174787942.0, "step": 4601 }, { "epoch": 0.5854217020735275, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.439622402191162, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8591545224189758, "num_tokens": 174827483.0, "step": 4602 }, { "epoch": 0.585548912352118, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6716465950012207, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8697681427001953, "num_tokens": 174862847.0, "step": 4603 }, { "epoch": 0.5856761226307086, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.441755771636963, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8479816913604736, "num_tokens": 174903327.0, "step": 4604 }, { "epoch": 0.5858033329092991, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.669034004211426, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8426604270935059, "num_tokens": 174939278.0, "step": 4605 }, { "epoch": 0.5859305431878896, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.455357789993286, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8559058308601379, "num_tokens": 174972671.0, "step": 4606 }, { "epoch": 0.5860577534664801, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.644338846206665, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8487525582313538, "num_tokens": 175009753.0, "step": 4607 }, { "epoch": 0.5861849637450706, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.450385808944702, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8728296160697937, "num_tokens": 175044271.0, "step": 4608 }, { "epoch": 0.5863121740236611, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.617371082305908, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8517661690711975, "num_tokens": 175086132.0, "step": 4609 }, { "epoch": 0.5864393843022516, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.452540874481201, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8720107674598694, "num_tokens": 175114902.0, "step": 4610 }, { "epoch": 0.5865665945808421, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6064817905426025, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8507810235023499, "num_tokens": 175152274.0, "step": 4611 }, { "epoch": 0.5866938048594327, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4201760292053223, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8617372512817383, "num_tokens": 175198449.0, "step": 4612 }, { "epoch": 0.5868210151380232, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6073129177093506, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8703950643539429, "num_tokens": 175236851.0, "step": 4613 }, { "epoch": 0.5869482254166136, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.431213617324829, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8592203855514526, "num_tokens": 175275694.0, "step": 4614 }, { "epoch": 0.5870754356952042, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5997602939605713, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8435794711112976, "num_tokens": 175321484.0, "step": 4615 }, { "epoch": 0.5872026459737947, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4222588539123535, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8601959943771362, "num_tokens": 175358387.0, "step": 4616 }, { "epoch": 0.5873298562523852, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00013065338134765625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.590921401977539, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.849851667881012, "num_tokens": 175402363.0, "step": 4617 }, { "epoch": 0.5874570665309757, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.422701120376587, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8546024560928345, "num_tokens": 175440655.0, "step": 4618 }, { "epoch": 0.5875842768095663, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6036157608032227, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8561992049217224, "num_tokens": 175476041.0, "step": 4619 }, { "epoch": 0.5877114870881567, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4107985496520996, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8765405416488647, "num_tokens": 175514673.0, "step": 4620 }, { "epoch": 0.5878386973667472, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.602660894393921, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8681936860084534, "num_tokens": 175555396.0, "step": 4621 }, { "epoch": 0.5879659076453377, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.404303789138794, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8747385740280151, "num_tokens": 175597588.0, "step": 4622 }, { "epoch": 0.5880931179239283, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.620239496231079, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.864510178565979, "num_tokens": 175636911.0, "step": 4623 }, { "epoch": 0.5882203282025188, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4297430515289307, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8418264985084534, "num_tokens": 175671778.0, "step": 4624 }, { "epoch": 0.5883475384811093, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.599609851837158, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.876103401184082, "num_tokens": 175711218.0, "step": 4625 }, { "epoch": 0.5884747487596997, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.434641122817993, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8589993715286255, "num_tokens": 175744630.0, "step": 4626 }, { "epoch": 0.5886019590382903, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.659381866455078, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8580145835876465, "num_tokens": 175783342.0, "step": 4627 }, { "epoch": 0.5887291693168808, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4130804538726807, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8470556139945984, "num_tokens": 175820361.0, "step": 4628 }, { "epoch": 0.5888563795954713, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6159911155700684, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8531572818756104, "num_tokens": 175860388.0, "step": 4629 }, { "epoch": 0.5889835898740619, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.413827896118164, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8569591641426086, "num_tokens": 175895725.0, "step": 4630 }, { "epoch": 0.5891108001526524, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.628260374069214, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8627988696098328, "num_tokens": 175933535.0, "step": 4631 }, { "epoch": 0.5892380104312428, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4266016483306885, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8498048186302185, "num_tokens": 175968681.0, "step": 4632 }, { "epoch": 0.5893652207098333, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6215133666992188, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.852340579032898, "num_tokens": 176006088.0, "step": 4633 }, { "epoch": 0.5894924309884239, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3981130123138428, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8726555705070496, "num_tokens": 176040855.0, "step": 4634 }, { "epoch": 0.5896196412670144, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6285240650177, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8586517572402954, "num_tokens": 176080015.0, "step": 4635 }, { "epoch": 0.5897468515456049, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4061341285705566, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8471298217773438, "num_tokens": 176120266.0, "step": 4636 }, { "epoch": 0.5898740618241954, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6518430709838867, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.864072322845459, "num_tokens": 176150436.0, "step": 4637 }, { "epoch": 0.5900012721027859, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.4078357219696045, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8482750654220581, "num_tokens": 176185340.0, "step": 4638 }, { "epoch": 0.5901284823813764, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.646486520767212, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8515257835388184, "num_tokens": 176225372.0, "step": 4639 }, { "epoch": 0.5902556926599669, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.387080669403076, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8592525124549866, "num_tokens": 176267114.0, "step": 4640 }, { "epoch": 0.5903829029385574, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6594297885894775, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8726249933242798, "num_tokens": 176302937.0, "step": 4641 }, { "epoch": 0.590510113217148, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4026646614074707, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.869165301322937, "num_tokens": 176340307.0, "step": 4642 }, { "epoch": 0.5906373234957385, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.664323329925537, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8504974842071533, "num_tokens": 176385420.0, "step": 4643 }, { "epoch": 0.5907645337743289, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.413278102874756, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8547879457473755, "num_tokens": 176424975.0, "step": 4644 }, { "epoch": 0.5908917440529194, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.664303779602051, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8406863808631897, "num_tokens": 176465471.0, "step": 4645 }, { "epoch": 0.59101895433151, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.431894302368164, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.861323356628418, "num_tokens": 176493717.0, "step": 4646 }, { "epoch": 0.5911461646101005, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6454741954803467, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8567168116569519, "num_tokens": 176533636.0, "step": 4647 }, { "epoch": 0.591273374888691, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4294333457946777, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8560669422149658, "num_tokens": 176573248.0, "step": 4648 }, { "epoch": 0.5914005851672816, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6260926723480225, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8714321851730347, "num_tokens": 176614607.0, "step": 4649 }, { "epoch": 0.591527795445872, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.438800096511841, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8527086973190308, "num_tokens": 176657482.0, "step": 4650 }, { "epoch": 0.5916550057244625, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6215367317199707, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8679945468902588, "num_tokens": 176694448.0, "step": 4651 }, { "epoch": 0.591782216003053, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.454991340637207, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8534045219421387, "num_tokens": 176726899.0, "step": 4652 }, { "epoch": 0.5919094262816436, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5907859802246094, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.852973222732544, "num_tokens": 176767495.0, "step": 4653 }, { "epoch": 0.5920366365602341, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.452871799468994, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8550246357917786, "num_tokens": 176802804.0, "step": 4654 }, { "epoch": 0.5921638468388246, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.557833671569824, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8682239651679993, "num_tokens": 176847224.0, "step": 4655 }, { "epoch": 0.592291057117415, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.45009446144104, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8550338745117188, "num_tokens": 176884423.0, "step": 4656 }, { "epoch": 0.5924182673960056, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5608389377593994, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8671884536743164, "num_tokens": 176917776.0, "step": 4657 }, { "epoch": 0.5925454776745961, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4348535537719727, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8669115304946899, "num_tokens": 176951387.0, "step": 4658 }, { "epoch": 0.5926726879531866, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.557199239730835, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8511049747467041, "num_tokens": 176986533.0, "step": 4659 }, { "epoch": 0.5927998982317771, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4294986724853516, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8695034980773926, "num_tokens": 177021541.0, "step": 4660 }, { "epoch": 0.5929271085103677, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5505685806274414, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.860487163066864, "num_tokens": 177062933.0, "step": 4661 }, { "epoch": 0.5930543187889582, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4254982471466064, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8398396372795105, "num_tokens": 177102549.0, "step": 4662 }, { "epoch": 0.5931815290675486, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.564755916595459, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8574122190475464, "num_tokens": 177139618.0, "step": 4663 }, { "epoch": 0.5933087393461391, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.427077293395996, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8507325053215027, "num_tokens": 177171912.0, "step": 4664 }, { "epoch": 0.5934359496247297, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5622174739837646, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8594158291816711, "num_tokens": 177214279.0, "step": 4665 }, { "epoch": 0.5935631599033202, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4289731979370117, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8556311130523682, "num_tokens": 177247277.0, "step": 4666 }, { "epoch": 0.5936903701819107, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5572292804718018, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.857291042804718, "num_tokens": 177287097.0, "step": 4667 }, { "epoch": 0.5938175804605013, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.414341688156128, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8698923587799072, "num_tokens": 177324400.0, "step": 4668 }, { "epoch": 0.5939447907390917, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5658130645751953, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8662647604942322, "num_tokens": 177362962.0, "step": 4669 }, { "epoch": 0.5940720010176822, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4055511951446533, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8576750755310059, "num_tokens": 177402708.0, "step": 4670 }, { "epoch": 0.5941992112962727, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5898256301879883, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8689877986907959, "num_tokens": 177433957.0, "step": 4671 }, { "epoch": 0.5943264215748633, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4112555980682373, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8631552457809448, "num_tokens": 177465265.0, "step": 4672 }, { "epoch": 0.5944536318534538, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.614011526107788, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8505370616912842, "num_tokens": 177495448.0, "step": 4673 }, { "epoch": 0.5945808421320443, "ewc_loss": 0.04833984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3909335136413574, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8639868497848511, "num_tokens": 177531582.0, "step": 4674 }, { "epoch": 0.5947080524106347, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.634617328643799, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8568127155303955, "num_tokens": 177565039.0, "step": 4675 }, { "epoch": 0.5948352626892253, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3808836936950684, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8575332760810852, "num_tokens": 177601872.0, "step": 4676 }, { "epoch": 0.5949624729678158, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.653076648712158, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8454880714416504, "num_tokens": 177645693.0, "step": 4677 }, { "epoch": 0.5950896832464063, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.382387399673462, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8630074858665466, "num_tokens": 177685334.0, "step": 4678 }, { "epoch": 0.5952168935249968, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6819186210632324, "learning_rate": 1e-05, "loss": 0.6057, "mean_token_accuracy": 0.8202551007270813, "num_tokens": 177727851.0, "step": 4679 }, { "epoch": 0.5953441038035874, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.3845598697662354, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8545160889625549, "num_tokens": 177769115.0, "step": 4680 }, { "epoch": 0.5954713140821778, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.701310873031616, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8522155284881592, "num_tokens": 177809094.0, "step": 4681 }, { "epoch": 0.5955985243607683, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.405613422393799, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8587190508842468, "num_tokens": 177842409.0, "step": 4682 }, { "epoch": 0.5957257346393589, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7114386558532715, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8531643152236938, "num_tokens": 177882525.0, "step": 4683 }, { "epoch": 0.5958529449179494, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4567079544067383, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8422422409057617, "num_tokens": 177924293.0, "step": 4684 }, { "epoch": 0.5959801551965399, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.676841974258423, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8611841201782227, "num_tokens": 177964716.0, "step": 4685 }, { "epoch": 0.5961073654751304, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4230306148529053, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.878696084022522, "num_tokens": 177999029.0, "step": 4686 }, { "epoch": 0.5962345757537209, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6739306449890137, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8713814616203308, "num_tokens": 178033872.0, "step": 4687 }, { "epoch": 0.5963617860323114, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4140639305114746, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8629170656204224, "num_tokens": 178077835.0, "step": 4688 }, { "epoch": 0.5964889963109019, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6676063537597656, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8653954863548279, "num_tokens": 178119205.0, "step": 4689 }, { "epoch": 0.5966162065894924, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4299709796905518, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8478842973709106, "num_tokens": 178157923.0, "step": 4690 }, { "epoch": 0.596743416868083, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6570394039154053, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8737732172012329, "num_tokens": 178200465.0, "step": 4691 }, { "epoch": 0.5968706271466735, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4611310958862305, "learning_rate": 1e-05, "loss": 0.552, "mean_token_accuracy": 0.8307634592056274, "num_tokens": 178234767.0, "step": 4692 }, { "epoch": 0.5969978374252639, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6461658477783203, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8487527966499329, "num_tokens": 178269569.0, "step": 4693 }, { "epoch": 0.5971250477038544, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4546866416931152, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8584107160568237, "num_tokens": 178300789.0, "step": 4694 }, { "epoch": 0.597252257982445, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6188695430755615, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.867456316947937, "num_tokens": 178333341.0, "step": 4695 }, { "epoch": 0.5973794682610355, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.443946361541748, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8443347215652466, "num_tokens": 178368743.0, "step": 4696 }, { "epoch": 0.597506678539626, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.593580961227417, "learning_rate": 1e-05, "loss": 0.3942, "mean_token_accuracy": 0.8776446580886841, "num_tokens": 178407432.0, "step": 4697 }, { "epoch": 0.5976338888182166, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.465862512588501, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8554373383522034, "num_tokens": 178440245.0, "step": 4698 }, { "epoch": 0.597761099096807, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.564171314239502, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8626044392585754, "num_tokens": 178484446.0, "step": 4699 }, { "epoch": 0.5978883093753975, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4476797580718994, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8606494069099426, "num_tokens": 178522390.0, "step": 4700 }, { "epoch": 0.598015519653988, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.552746295928955, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.864065408706665, "num_tokens": 178561671.0, "step": 4701 }, { "epoch": 0.5981427299325786, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4530487060546875, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.840079665184021, "num_tokens": 178605289.0, "step": 4702 }, { "epoch": 0.5982699402111691, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5399973392486572, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8663073778152466, "num_tokens": 178647947.0, "step": 4703 }, { "epoch": 0.5983971504897596, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4378247261047363, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8702514171600342, "num_tokens": 178686779.0, "step": 4704 }, { "epoch": 0.59852436076835, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5438454151153564, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8486796617507935, "num_tokens": 178730517.0, "step": 4705 }, { "epoch": 0.5986515710469406, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.459000587463379, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8578221797943115, "num_tokens": 178762521.0, "step": 4706 }, { "epoch": 0.5987787813255311, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.53558611869812, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8501330614089966, "num_tokens": 178804816.0, "step": 4707 }, { "epoch": 0.5989059916041216, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4532697200775146, "learning_rate": 1e-05, "loss": 0.5286, "mean_token_accuracy": 0.845522403717041, "num_tokens": 178845413.0, "step": 4708 }, { "epoch": 0.5990332018827121, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.512054681777954, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8568264245986938, "num_tokens": 178886887.0, "step": 4709 }, { "epoch": 0.5991604121613027, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4459242820739746, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8629564046859741, "num_tokens": 178922067.0, "step": 4710 }, { "epoch": 0.5992876224398932, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5479798316955566, "learning_rate": 1e-05, "loss": 0.5376, "mean_token_accuracy": 0.8377679586410522, "num_tokens": 178955645.0, "step": 4711 }, { "epoch": 0.5994148327184836, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4394571781158447, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8605867028236389, "num_tokens": 178992389.0, "step": 4712 }, { "epoch": 0.5995420429970741, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.535071611404419, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8471121788024902, "num_tokens": 179031140.0, "step": 4713 }, { "epoch": 0.5996692532756647, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.423640251159668, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8752715587615967, "num_tokens": 179065486.0, "step": 4714 }, { "epoch": 0.5997964635542552, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001316070556640625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5490832328796387, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8561167120933533, "num_tokens": 179104957.0, "step": 4715 }, { "epoch": 0.5999236738328457, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4079461097717285, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8581402897834778, "num_tokens": 179146022.0, "step": 4716 }, { "epoch": 0.6000508841114363, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.559969186782837, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8674731254577637, "num_tokens": 179184691.0, "step": 4717 }, { "epoch": 0.6001780943900267, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.404984474182129, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8612074851989746, "num_tokens": 179223734.0, "step": 4718 }, { "epoch": 0.6003053046686172, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5667059421539307, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8665372133255005, "num_tokens": 179265997.0, "step": 4719 }, { "epoch": 0.6004325149472077, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.00013256072998046875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4022469520568848, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8523236513137817, "num_tokens": 179302734.0, "step": 4720 }, { "epoch": 0.6005597252257983, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5894241333007812, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8520569801330566, "num_tokens": 179343779.0, "step": 4721 }, { "epoch": 0.6006869355043888, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3851237297058105, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8654695749282837, "num_tokens": 179384713.0, "step": 4722 }, { "epoch": 0.6008141457829793, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.621060609817505, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8641380667686462, "num_tokens": 179418005.0, "step": 4723 }, { "epoch": 0.6009413560615697, "ewc_loss": 0.048583984375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.3878238201141357, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8513869047164917, "num_tokens": 179459437.0, "step": 4724 }, { "epoch": 0.6010685663401603, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.655665874481201, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.861190915107727, "num_tokens": 179490534.0, "step": 4725 }, { "epoch": 0.6011957766187508, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3765273094177246, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8564906120300293, "num_tokens": 179533141.0, "step": 4726 }, { "epoch": 0.6013229868973413, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6955857276916504, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8500160574913025, "num_tokens": 179565438.0, "step": 4727 }, { "epoch": 0.6014501971759318, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3804543018341064, "learning_rate": 1e-05, "loss": 0.526, "mean_token_accuracy": 0.8404871225357056, "num_tokens": 179605184.0, "step": 4728 }, { "epoch": 0.6015774074545224, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7083964347839355, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8584551811218262, "num_tokens": 179643355.0, "step": 4729 }, { "epoch": 0.6017046177331128, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.39361572265625, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8632568717002869, "num_tokens": 179682471.0, "step": 4730 }, { "epoch": 0.6018318280117033, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7227184772491455, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8485939502716064, "num_tokens": 179722290.0, "step": 4731 }, { "epoch": 0.6019590382902938, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4072775840759277, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.843624472618103, "num_tokens": 179762409.0, "step": 4732 }, { "epoch": 0.6020862485688844, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.731289863586426, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8470430374145508, "num_tokens": 179802561.0, "step": 4733 }, { "epoch": 0.6022134588474749, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.417680263519287, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8601247668266296, "num_tokens": 179842923.0, "step": 4734 }, { "epoch": 0.6023406691260654, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.735670804977417, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8547464609146118, "num_tokens": 179882430.0, "step": 4735 }, { "epoch": 0.6024678794046558, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.434462070465088, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8509882688522339, "num_tokens": 179923974.0, "step": 4736 }, { "epoch": 0.6025950896832464, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7165229320526123, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8671577572822571, "num_tokens": 179966468.0, "step": 4737 }, { "epoch": 0.6027222999618369, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4412660598754883, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8721312284469604, "num_tokens": 180005356.0, "step": 4738 }, { "epoch": 0.6028495102404274, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7140421867370605, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8568769693374634, "num_tokens": 180044765.0, "step": 4739 }, { "epoch": 0.602976720519018, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.461594343185425, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8501490354537964, "num_tokens": 180082311.0, "step": 4740 }, { "epoch": 0.6031039307976085, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.678532123565674, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8628686666488647, "num_tokens": 180122889.0, "step": 4741 }, { "epoch": 0.6032311410761989, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.459200859069824, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.864879846572876, "num_tokens": 180157425.0, "step": 4742 }, { "epoch": 0.6033583513547894, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.647874355316162, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8657870888710022, "num_tokens": 180196099.0, "step": 4743 }, { "epoch": 0.60348556163338, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4534497261047363, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.8416427969932556, "num_tokens": 180238036.0, "step": 4744 }, { "epoch": 0.6036127719119705, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.633901834487915, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8602350950241089, "num_tokens": 180276193.0, "step": 4745 }, { "epoch": 0.603739982190561, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.451258659362793, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8562560081481934, "num_tokens": 180313470.0, "step": 4746 }, { "epoch": 0.6038671924691515, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6178879737854004, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8611921072006226, "num_tokens": 180354588.0, "step": 4747 }, { "epoch": 0.603994402747742, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.435636281967163, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8585066795349121, "num_tokens": 180395792.0, "step": 4748 }, { "epoch": 0.6041216130263325, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6108458042144775, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8502967357635498, "num_tokens": 180436823.0, "step": 4749 }, { "epoch": 0.604248823304923, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4370365142822266, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.867473304271698, "num_tokens": 180472643.0, "step": 4750 }, { "epoch": 0.6043760335835135, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6053836345672607, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8493033051490784, "num_tokens": 180511788.0, "step": 4751 }, { "epoch": 0.6045032438621041, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4292991161346436, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8716577291488647, "num_tokens": 180547871.0, "step": 4752 }, { "epoch": 0.6046304541406946, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.622976541519165, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8471827507019043, "num_tokens": 180580521.0, "step": 4753 }, { "epoch": 0.604757664419285, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.428950548171997, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8528934717178345, "num_tokens": 180617726.0, "step": 4754 }, { "epoch": 0.6048848746978756, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6085586547851562, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8570948243141174, "num_tokens": 180657705.0, "step": 4755 }, { "epoch": 0.6050120849764661, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.426563024520874, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8664174675941467, "num_tokens": 180693582.0, "step": 4756 }, { "epoch": 0.6051392952550566, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6136276721954346, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.851051926612854, "num_tokens": 180731159.0, "step": 4757 }, { "epoch": 0.6052665055336471, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4187381267547607, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8483322858810425, "num_tokens": 180769901.0, "step": 4758 }, { "epoch": 0.6053937158122377, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.599721908569336, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.861662745475769, "num_tokens": 180811763.0, "step": 4759 }, { "epoch": 0.6055209260908282, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4229884147644043, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8617888689041138, "num_tokens": 180846073.0, "step": 4760 }, { "epoch": 0.6056481363694186, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6066091060638428, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8672280311584473, "num_tokens": 180883609.0, "step": 4761 }, { "epoch": 0.6057753466480091, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4239518642425537, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8571311235427856, "num_tokens": 180921492.0, "step": 4762 }, { "epoch": 0.6059025569265997, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6090033054351807, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8442220687866211, "num_tokens": 180960298.0, "step": 4763 }, { "epoch": 0.6060297672051902, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4254775047302246, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.851598858833313, "num_tokens": 180997002.0, "step": 4764 }, { "epoch": 0.6061569774837807, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.613955020904541, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8390882611274719, "num_tokens": 181033575.0, "step": 4765 }, { "epoch": 0.6062841877623713, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.414375066757202, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8584209680557251, "num_tokens": 181068584.0, "step": 4766 }, { "epoch": 0.6064113980409617, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6264894008636475, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8630644083023071, "num_tokens": 181103017.0, "step": 4767 }, { "epoch": 0.6065386083195522, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4166855812072754, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8588377237319946, "num_tokens": 181137590.0, "step": 4768 }, { "epoch": 0.6066658185981427, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6221139430999756, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.857836902141571, "num_tokens": 181177581.0, "step": 4769 }, { "epoch": 0.6067930288767333, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.405071496963501, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8678164482116699, "num_tokens": 181212582.0, "step": 4770 }, { "epoch": 0.6069202391553238, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6381137371063232, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8498132824897766, "num_tokens": 181246373.0, "step": 4771 }, { "epoch": 0.6070474494339143, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.41076397895813, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8730195760726929, "num_tokens": 181280657.0, "step": 4772 }, { "epoch": 0.6071746597125047, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6236109733581543, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8674740195274353, "num_tokens": 181322165.0, "step": 4773 }, { "epoch": 0.6073018699910953, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.412092685699463, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8573802709579468, "num_tokens": 181358612.0, "step": 4774 }, { "epoch": 0.6074290802696858, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6406567096710205, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.871845006942749, "num_tokens": 181392928.0, "step": 4775 }, { "epoch": 0.6075562905482763, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.419968843460083, "learning_rate": 1e-05, "loss": 0.5342, "mean_token_accuracy": 0.8415302634239197, "num_tokens": 181431127.0, "step": 4776 }, { "epoch": 0.6076835008268668, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6378936767578125, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8589731454849243, "num_tokens": 181469830.0, "step": 4777 }, { "epoch": 0.6078107111054574, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4215445518493652, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8575470447540283, "num_tokens": 181505590.0, "step": 4778 }, { "epoch": 0.6079379213840478, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.63542103767395, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8732279539108276, "num_tokens": 181541675.0, "step": 4779 }, { "epoch": 0.6080651316626383, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.427628517150879, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8460162281990051, "num_tokens": 181579890.0, "step": 4780 }, { "epoch": 0.6081923419412288, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.633315324783325, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8562501668930054, "num_tokens": 181613821.0, "step": 4781 }, { "epoch": 0.6083195522198194, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4209182262420654, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8454781174659729, "num_tokens": 181656899.0, "step": 4782 }, { "epoch": 0.6084467624984099, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.645169973373413, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8422313332557678, "num_tokens": 181690044.0, "step": 4783 }, { "epoch": 0.6085739727770004, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.423363447189331, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8627566695213318, "num_tokens": 181725038.0, "step": 4784 }, { "epoch": 0.6087011830555908, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.633089780807495, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8598779439926147, "num_tokens": 181767914.0, "step": 4785 }, { "epoch": 0.6088283933341814, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.422701597213745, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8700493574142456, "num_tokens": 181798955.0, "step": 4786 }, { "epoch": 0.6089556036127719, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6466360092163086, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8452464938163757, "num_tokens": 181831967.0, "step": 4787 }, { "epoch": 0.6090828138913624, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4287145137786865, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.85822594165802, "num_tokens": 181865717.0, "step": 4788 }, { "epoch": 0.609210024169953, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6408114433288574, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8420367240905762, "num_tokens": 181902847.0, "step": 4789 }, { "epoch": 0.6093372344485435, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.421464681625366, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8703400492668152, "num_tokens": 181937129.0, "step": 4790 }, { "epoch": 0.6094644447271339, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.630854845046997, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8580124378204346, "num_tokens": 181979576.0, "step": 4791 }, { "epoch": 0.6095916550057244, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4163198471069336, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.858232855796814, "num_tokens": 182021510.0, "step": 4792 }, { "epoch": 0.609718865284315, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6294429302215576, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8595991730690002, "num_tokens": 182061178.0, "step": 4793 }, { "epoch": 0.6098460755629055, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4195828437805176, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.8434404134750366, "num_tokens": 182100575.0, "step": 4794 }, { "epoch": 0.609973285841496, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6430859565734863, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8558433651924133, "num_tokens": 182136142.0, "step": 4795 }, { "epoch": 0.6101004961200865, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4174435138702393, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8485101461410522, "num_tokens": 182173108.0, "step": 4796 }, { "epoch": 0.610227706398677, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.652456760406494, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8456135392189026, "num_tokens": 182209931.0, "step": 4797 }, { "epoch": 0.6103549166772675, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.430537700653076, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8619050979614258, "num_tokens": 182244128.0, "step": 4798 }, { "epoch": 0.610482126955858, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.642767906188965, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.857742428779602, "num_tokens": 182278112.0, "step": 4799 }, { "epoch": 0.6106093372344485, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4231927394866943, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8624254465103149, "num_tokens": 182313753.0, "step": 4800 }, { "epoch": 0.6107365475130391, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.635667085647583, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.879270076751709, "num_tokens": 182348758.0, "step": 4801 }, { "epoch": 0.6108637577916296, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.434004068374634, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8578383922576904, "num_tokens": 182381084.0, "step": 4802 }, { "epoch": 0.61099096807022, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.632929563522339, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8595835566520691, "num_tokens": 182420530.0, "step": 4803 }, { "epoch": 0.6111181783488105, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.417553424835205, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8578157424926758, "num_tokens": 182462286.0, "step": 4804 }, { "epoch": 0.6112453886274011, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.624204158782959, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8730767965316772, "num_tokens": 182497209.0, "step": 4805 }, { "epoch": 0.6113725989059916, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4288930892944336, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8711889386177063, "num_tokens": 182530609.0, "step": 4806 }, { "epoch": 0.6114998091845821, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.628786325454712, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8657400608062744, "num_tokens": 182566698.0, "step": 4807 }, { "epoch": 0.6116270194631727, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4121389389038086, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.864815354347229, "num_tokens": 182609524.0, "step": 4808 }, { "epoch": 0.6117542297417632, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.639329433441162, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8614274263381958, "num_tokens": 182649339.0, "step": 4809 }, { "epoch": 0.6118814400203536, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.414893627166748, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8603799343109131, "num_tokens": 182684142.0, "step": 4810 }, { "epoch": 0.6120086502989441, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.646216869354248, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.856411337852478, "num_tokens": 182722857.0, "step": 4811 }, { "epoch": 0.6121358605775347, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.420182466506958, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8521496057510376, "num_tokens": 182760196.0, "step": 4812 }, { "epoch": 0.6122630708561252, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.640568256378174, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8548864722251892, "num_tokens": 182795807.0, "step": 4813 }, { "epoch": 0.6123902811347157, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.414872646331787, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8508562445640564, "num_tokens": 182837956.0, "step": 4814 }, { "epoch": 0.6125174914133062, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6494216918945312, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8648950457572937, "num_tokens": 182873374.0, "step": 4815 }, { "epoch": 0.6126447016918967, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.419140577316284, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8722567558288574, "num_tokens": 182908018.0, "step": 4816 }, { "epoch": 0.6127719119704872, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.642646074295044, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8565837144851685, "num_tokens": 182950314.0, "step": 4817 }, { "epoch": 0.6128991222490777, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4201836585998535, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8585246801376343, "num_tokens": 182988434.0, "step": 4818 }, { "epoch": 0.6130263325276682, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6446785926818848, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8656495213508606, "num_tokens": 183026355.0, "step": 4819 }, { "epoch": 0.6131535428062588, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4153144359588623, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8580681085586548, "num_tokens": 183063031.0, "step": 4820 }, { "epoch": 0.6132807530848493, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6506190299987793, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.846840500831604, "num_tokens": 183102082.0, "step": 4821 }, { "epoch": 0.6134079633634397, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.417714834213257, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8576467633247375, "num_tokens": 183137961.0, "step": 4822 }, { "epoch": 0.6135351736420303, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.642570734024048, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.870881199836731, "num_tokens": 183173779.0, "step": 4823 }, { "epoch": 0.6136623839206208, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.426025629043579, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.850536048412323, "num_tokens": 183215044.0, "step": 4824 }, { "epoch": 0.6137895941992113, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6450016498565674, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8654999732971191, "num_tokens": 183248793.0, "step": 4825 }, { "epoch": 0.6139168044778018, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4125633239746094, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8580650091171265, "num_tokens": 183291180.0, "step": 4826 }, { "epoch": 0.6140440147563924, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.655656099319458, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8588501214981079, "num_tokens": 183326214.0, "step": 4827 }, { "epoch": 0.6141712250349828, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4143600463867188, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8632216453552246, "num_tokens": 183362403.0, "step": 4828 }, { "epoch": 0.6142984353135733, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6509287357330322, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.865652322769165, "num_tokens": 183407419.0, "step": 4829 }, { "epoch": 0.6144256455921638, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.428623676300049, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8583437204360962, "num_tokens": 183443140.0, "step": 4830 }, { "epoch": 0.6145528558707544, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.65627384185791, "learning_rate": 1e-05, "loss": 0.5319, "mean_token_accuracy": 0.8431482315063477, "num_tokens": 183485986.0, "step": 4831 }, { "epoch": 0.6146800661493449, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.435361623764038, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8440167307853699, "num_tokens": 183525302.0, "step": 4832 }, { "epoch": 0.6148072764279354, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6444854736328125, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8528957962989807, "num_tokens": 183561575.0, "step": 4833 }, { "epoch": 0.6149344867065258, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4278409481048584, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8666621446609497, "num_tokens": 183596465.0, "step": 4834 }, { "epoch": 0.6150616969851164, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6451289653778076, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8367047905921936, "num_tokens": 183631733.0, "step": 4835 }, { "epoch": 0.6151889072637069, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4172916412353516, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8550113439559937, "num_tokens": 183676138.0, "step": 4836 }, { "epoch": 0.6153161175422974, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000133514404296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6463212966918945, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8586304783821106, "num_tokens": 183711004.0, "step": 4837 }, { "epoch": 0.615443327820888, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4254419803619385, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8791002631187439, "num_tokens": 183744945.0, "step": 4838 }, { "epoch": 0.6155705380994785, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6527626514434814, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8679082989692688, "num_tokens": 183779119.0, "step": 4839 }, { "epoch": 0.6156977483780689, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.425724744796753, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8668201565742493, "num_tokens": 183814136.0, "step": 4840 }, { "epoch": 0.6158249586566594, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.642446279525757, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.853222131729126, "num_tokens": 183850957.0, "step": 4841 }, { "epoch": 0.61595216893525, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4122936725616455, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8694498538970947, "num_tokens": 183890049.0, "step": 4842 }, { "epoch": 0.6160793792138405, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6419084072113037, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8588895797729492, "num_tokens": 183933055.0, "step": 4843 }, { "epoch": 0.616206589492431, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.416482448577881, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8584643602371216, "num_tokens": 183975059.0, "step": 4844 }, { "epoch": 0.6163337997710215, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6500744819641113, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8518543243408203, "num_tokens": 184015130.0, "step": 4845 }, { "epoch": 0.616461010049612, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4406380653381348, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8477407097816467, "num_tokens": 184048416.0, "step": 4846 }, { "epoch": 0.6165882203282025, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.645158052444458, "learning_rate": 1e-05, "loss": 0.5043, "mean_token_accuracy": 0.849519670009613, "num_tokens": 184081591.0, "step": 4847 }, { "epoch": 0.616715430606793, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.415348768234253, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.878849446773529, "num_tokens": 184118078.0, "step": 4848 }, { "epoch": 0.6168426408853835, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6437432765960693, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8627650737762451, "num_tokens": 184154860.0, "step": 4849 }, { "epoch": 0.6169698511639741, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.437760353088379, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8710892796516418, "num_tokens": 184186425.0, "step": 4850 }, { "epoch": 0.6170970614425646, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6317763328552246, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8524453043937683, "num_tokens": 184224891.0, "step": 4851 }, { "epoch": 0.617224271721155, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.419433355331421, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.870307445526123, "num_tokens": 184266648.0, "step": 4852 }, { "epoch": 0.6173514819997455, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.648118734359741, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8475149869918823, "num_tokens": 184301453.0, "step": 4853 }, { "epoch": 0.6174786922783361, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4259138107299805, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8582467436790466, "num_tokens": 184344271.0, "step": 4854 }, { "epoch": 0.6176059025569266, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6335325241088867, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8648620843887329, "num_tokens": 184378212.0, "step": 4855 }, { "epoch": 0.6177331128355171, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.420952320098877, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8650990724563599, "num_tokens": 184414177.0, "step": 4856 }, { "epoch": 0.6178603231141077, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.63278865814209, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8477529883384705, "num_tokens": 184455088.0, "step": 4857 }, { "epoch": 0.6179875333926982, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4238646030426025, "learning_rate": 1e-05, "loss": 0.5134, "mean_token_accuracy": 0.8488689064979553, "num_tokens": 184496574.0, "step": 4858 }, { "epoch": 0.6181147436712886, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.629599094390869, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8531233072280884, "num_tokens": 184542303.0, "step": 4859 }, { "epoch": 0.6182419539498791, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.412893056869507, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8588358163833618, "num_tokens": 184583337.0, "step": 4860 }, { "epoch": 0.6183691642284697, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6353752613067627, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8543983697891235, "num_tokens": 184622424.0, "step": 4861 }, { "epoch": 0.6184963745070602, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.428846597671509, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.843514621257782, "num_tokens": 184660137.0, "step": 4862 }, { "epoch": 0.6186235847856507, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.627067804336548, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.85582435131073, "num_tokens": 184698353.0, "step": 4863 }, { "epoch": 0.6187507950642412, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4192330837249756, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8510928153991699, "num_tokens": 184738734.0, "step": 4864 }, { "epoch": 0.6188780053428317, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.638383388519287, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.848991870880127, "num_tokens": 184774525.0, "step": 4865 }, { "epoch": 0.6190052156214222, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.420077323913574, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8624238967895508, "num_tokens": 184809028.0, "step": 4866 }, { "epoch": 0.6191324259000127, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.631063222885132, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8780068159103394, "num_tokens": 184845479.0, "step": 4867 }, { "epoch": 0.6192596361786032, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.424295663833618, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8427159786224365, "num_tokens": 184886596.0, "step": 4868 }, { "epoch": 0.6193868464571938, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.642688274383545, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8550132513046265, "num_tokens": 184923763.0, "step": 4869 }, { "epoch": 0.6195140567357843, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013446807861328125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.42433500289917, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.865092396736145, "num_tokens": 184959882.0, "step": 4870 }, { "epoch": 0.6196412670143747, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.629150867462158, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8740663528442383, "num_tokens": 184999063.0, "step": 4871 }, { "epoch": 0.6197684772929652, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.404881715774536, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8746911883354187, "num_tokens": 185040016.0, "step": 4872 }, { "epoch": 0.6198956875715558, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.653766393661499, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8572218418121338, "num_tokens": 185075085.0, "step": 4873 }, { "epoch": 0.6200228978501463, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.424586534500122, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8578376770019531, "num_tokens": 185111621.0, "step": 4874 }, { "epoch": 0.6201501081287368, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.658820390701294, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.852340817451477, "num_tokens": 185149681.0, "step": 4875 }, { "epoch": 0.6202773184073274, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4201500415802, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8655568361282349, "num_tokens": 185186449.0, "step": 4876 }, { "epoch": 0.6204045286859178, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.657264232635498, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8641970753669739, "num_tokens": 185219324.0, "step": 4877 }, { "epoch": 0.6205317389645083, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4242634773254395, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8564866781234741, "num_tokens": 185256049.0, "step": 4878 }, { "epoch": 0.6206589492430988, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.65463924407959, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8623244166374207, "num_tokens": 185293304.0, "step": 4879 }, { "epoch": 0.6207861595216894, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4220917224884033, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8622058033943176, "num_tokens": 185334007.0, "step": 4880 }, { "epoch": 0.6209133698002799, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.650876045227051, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8515223264694214, "num_tokens": 185372932.0, "step": 4881 }, { "epoch": 0.6210405800788704, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4207634925842285, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8659970760345459, "num_tokens": 185410267.0, "step": 4882 }, { "epoch": 0.6211677903574608, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6636710166931152, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8513344526290894, "num_tokens": 185445838.0, "step": 4883 }, { "epoch": 0.6212950006360514, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4199514389038086, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8622530698776245, "num_tokens": 185486002.0, "step": 4884 }, { "epoch": 0.6214222109146419, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6623103618621826, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8545116782188416, "num_tokens": 185524501.0, "step": 4885 }, { "epoch": 0.6215494211932324, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.420745849609375, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8563243746757507, "num_tokens": 185562105.0, "step": 4886 }, { "epoch": 0.621676631471823, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.658761501312256, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.868066668510437, "num_tokens": 185602821.0, "step": 4887 }, { "epoch": 0.6218038417504135, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.419832706451416, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8499385714530945, "num_tokens": 185643188.0, "step": 4888 }, { "epoch": 0.6219310520290039, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.663379430770874, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8445651531219482, "num_tokens": 185685770.0, "step": 4889 }, { "epoch": 0.6220582623075944, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4416987895965576, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8611536026000977, "num_tokens": 185718981.0, "step": 4890 }, { "epoch": 0.622185472586185, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.659419059753418, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8626676797866821, "num_tokens": 185752184.0, "step": 4891 }, { "epoch": 0.6223126828647755, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4283552169799805, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8539769649505615, "num_tokens": 185791405.0, "step": 4892 }, { "epoch": 0.622439893143366, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6579947471618652, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8384747505187988, "num_tokens": 185832227.0, "step": 4893 }, { "epoch": 0.6225671034219565, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4336891174316406, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8629247546195984, "num_tokens": 185867175.0, "step": 4894 }, { "epoch": 0.622694313700547, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6457254886627197, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8552793264389038, "num_tokens": 185907582.0, "step": 4895 }, { "epoch": 0.6228215239791375, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4383041858673096, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8469613790512085, "num_tokens": 185944625.0, "step": 4896 }, { "epoch": 0.622948734257728, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6439549922943115, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8701149225234985, "num_tokens": 185975579.0, "step": 4897 }, { "epoch": 0.6230759445363185, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.430549383163452, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8613824844360352, "num_tokens": 186016948.0, "step": 4898 }, { "epoch": 0.6232031548149091, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6219663619995117, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8519668579101562, "num_tokens": 186060700.0, "step": 4899 }, { "epoch": 0.6233303650934996, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.427673101425171, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8725517988204956, "num_tokens": 186099624.0, "step": 4900 }, { "epoch": 0.62345757537209, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6199164390563965, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8571624755859375, "num_tokens": 186142360.0, "step": 4901 }, { "epoch": 0.6235847856506805, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.420642852783203, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8695172667503357, "num_tokens": 186182642.0, "step": 4902 }, { "epoch": 0.6237119959292711, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6361145973205566, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8542228937149048, "num_tokens": 186218867.0, "step": 4903 }, { "epoch": 0.6238392062078616, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.421518564224243, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8711200952529907, "num_tokens": 186257424.0, "step": 4904 }, { "epoch": 0.6239664164864521, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.626944065093994, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8597822189331055, "num_tokens": 186298588.0, "step": 4905 }, { "epoch": 0.6240936267650427, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.418855905532837, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8667877316474915, "num_tokens": 186334638.0, "step": 4906 }, { "epoch": 0.6242208370436332, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.648956537246704, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8519436120986938, "num_tokens": 186368303.0, "step": 4907 }, { "epoch": 0.6243480473222236, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4255623817443848, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8728527426719666, "num_tokens": 186406133.0, "step": 4908 }, { "epoch": 0.6244752576008141, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6396126747131348, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8767669796943665, "num_tokens": 186440763.0, "step": 4909 }, { "epoch": 0.6246024678794047, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.424903631210327, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8579359650611877, "num_tokens": 186479662.0, "step": 4910 }, { "epoch": 0.6247296781579952, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6350512504577637, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8480358123779297, "num_tokens": 186518944.0, "step": 4911 }, { "epoch": 0.6248568884365857, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.419994592666626, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.869011640548706, "num_tokens": 186555902.0, "step": 4912 }, { "epoch": 0.6249840987151762, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.627382755279541, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8764150142669678, "num_tokens": 186591962.0, "step": 4913 }, { "epoch": 0.6251113089937667, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4274003505706787, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8517208099365234, "num_tokens": 186631841.0, "step": 4914 }, { "epoch": 0.6252385192723572, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6115968227386475, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8686649799346924, "num_tokens": 186677963.0, "step": 4915 }, { "epoch": 0.6253657295509477, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4289605617523193, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.857349693775177, "num_tokens": 186715658.0, "step": 4916 }, { "epoch": 0.6254929398295382, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6252801418304443, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8661113977432251, "num_tokens": 186749405.0, "step": 4917 }, { "epoch": 0.6256201501081288, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.433912992477417, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8500072956085205, "num_tokens": 186785667.0, "step": 4918 }, { "epoch": 0.6257473603867193, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6181411743164062, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8715511560440063, "num_tokens": 186817425.0, "step": 4919 }, { "epoch": 0.6258745706653097, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.415220022201538, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.867411732673645, "num_tokens": 186859693.0, "step": 4920 }, { "epoch": 0.6260017809439002, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.627261161804199, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8669066429138184, "num_tokens": 186897774.0, "step": 4921 }, { "epoch": 0.6261289912224908, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.416743516921997, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8612651824951172, "num_tokens": 186935530.0, "step": 4922 }, { "epoch": 0.6262562015010813, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.649860382080078, "learning_rate": 1e-05, "loss": 0.5082, "mean_token_accuracy": 0.8474817276000977, "num_tokens": 186970778.0, "step": 4923 }, { "epoch": 0.6263834117796718, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.414889335632324, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8587340712547302, "num_tokens": 187007338.0, "step": 4924 }, { "epoch": 0.6265106220582624, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6550676822662354, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8595473766326904, "num_tokens": 187042614.0, "step": 4925 }, { "epoch": 0.6266378323368528, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4116902351379395, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8536210060119629, "num_tokens": 187081753.0, "step": 4926 }, { "epoch": 0.6267650426154433, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6739871501922607, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8523522615432739, "num_tokens": 187116758.0, "step": 4927 }, { "epoch": 0.6268922528940338, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4724841117858887, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.850933849811554, "num_tokens": 187152296.0, "step": 4928 }, { "epoch": 0.6270194631726244, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6368749141693115, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8522613048553467, "num_tokens": 187187619.0, "step": 4929 }, { "epoch": 0.6271466734512149, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.429934501647949, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8595245480537415, "num_tokens": 187227872.0, "step": 4930 }, { "epoch": 0.6272738837298054, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6234545707702637, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8580202460289001, "num_tokens": 187265720.0, "step": 4931 }, { "epoch": 0.6274010940083958, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4366466999053955, "learning_rate": 1e-05, "loss": 0.5546, "mean_token_accuracy": 0.8356471061706543, "num_tokens": 187307119.0, "step": 4932 }, { "epoch": 0.6275283042869864, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6180527210235596, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8630292415618896, "num_tokens": 187342804.0, "step": 4933 }, { "epoch": 0.6276555145655769, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4214589595794678, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8603842258453369, "num_tokens": 187381540.0, "step": 4934 }, { "epoch": 0.6277827248441674, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6226449012756348, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8422542810440063, "num_tokens": 187424952.0, "step": 4935 }, { "epoch": 0.627909935122758, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.425804853439331, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8487560749053955, "num_tokens": 187463668.0, "step": 4936 }, { "epoch": 0.6280371454013485, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.62858247756958, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8551175594329834, "num_tokens": 187505962.0, "step": 4937 }, { "epoch": 0.6281643556799389, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.41721773147583, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8529508113861084, "num_tokens": 187546342.0, "step": 4938 }, { "epoch": 0.6282915659585294, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.630446672439575, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8579009175300598, "num_tokens": 187587378.0, "step": 4939 }, { "epoch": 0.62841877623712, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4340853691101074, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8603522181510925, "num_tokens": 187621972.0, "step": 4940 }, { "epoch": 0.6285459865157105, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6391754150390625, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8468128442764282, "num_tokens": 187657205.0, "step": 4941 }, { "epoch": 0.628673196794301, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4100074768066406, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8525829315185547, "num_tokens": 187698201.0, "step": 4942 }, { "epoch": 0.6288004070728915, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6390700340270996, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8693108558654785, "num_tokens": 187734883.0, "step": 4943 }, { "epoch": 0.628927617351482, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4000918865203857, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8720542192459106, "num_tokens": 187777459.0, "step": 4944 }, { "epoch": 0.6290548276300725, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6527390480041504, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8525838851928711, "num_tokens": 187819661.0, "step": 4945 }, { "epoch": 0.629182037908663, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.408116102218628, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.865481972694397, "num_tokens": 187857685.0, "step": 4946 }, { "epoch": 0.6293092481872535, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.675464391708374, "learning_rate": 1e-05, "loss": 0.549, "mean_token_accuracy": 0.8373026251792908, "num_tokens": 187896614.0, "step": 4947 }, { "epoch": 0.6294364584658441, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.610313653945923, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8514506816864014, "num_tokens": 187925480.0, "step": 4948 }, { "epoch": 0.6295636687444346, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5611820220947266, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.860256552696228, "num_tokens": 187965045.0, "step": 4949 }, { "epoch": 0.629690879023025, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.469356060028076, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8577610850334167, "num_tokens": 188001633.0, "step": 4950 }, { "epoch": 0.6298180893016155, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.525131940841675, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8560175895690918, "num_tokens": 188039931.0, "step": 4951 }, { "epoch": 0.6299452995802061, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4671905040740967, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.8430743217468262, "num_tokens": 188085353.0, "step": 4952 }, { "epoch": 0.6300725098587966, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.517411470413208, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8779167532920837, "num_tokens": 188118966.0, "step": 4953 }, { "epoch": 0.6301997201373871, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4600863456726074, "learning_rate": 1e-05, "loss": 0.3854, "mean_token_accuracy": 0.8829700946807861, "num_tokens": 188153680.0, "step": 4954 }, { "epoch": 0.6303269304159776, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5148608684539795, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8723946809768677, "num_tokens": 188189549.0, "step": 4955 }, { "epoch": 0.6304541406945681, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4551949501037598, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8512430191040039, "num_tokens": 188230469.0, "step": 4956 }, { "epoch": 0.6305813509731586, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.524477243423462, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8604031801223755, "num_tokens": 188265074.0, "step": 4957 }, { "epoch": 0.6307085612517491, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.453803777694702, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8517454266548157, "num_tokens": 188302759.0, "step": 4958 }, { "epoch": 0.6308357715303397, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.547950267791748, "learning_rate": 1e-05, "loss": 0.5643, "mean_token_accuracy": 0.8317571878433228, "num_tokens": 188337627.0, "step": 4959 }, { "epoch": 0.6309629818089302, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4835426807403564, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8571934103965759, "num_tokens": 188365094.0, "step": 4960 }, { "epoch": 0.6310901920875207, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5139129161834717, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.861594021320343, "num_tokens": 188403244.0, "step": 4961 }, { "epoch": 0.6312174023661112, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4433350563049316, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8676212430000305, "num_tokens": 188442785.0, "step": 4962 }, { "epoch": 0.6313446126447017, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.529021739959717, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.851736843585968, "num_tokens": 188479298.0, "step": 4963 }, { "epoch": 0.6314718229232922, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4403841495513916, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8416542410850525, "num_tokens": 188519304.0, "step": 4964 }, { "epoch": 0.6315990332018827, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5428550243377686, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8426399230957031, "num_tokens": 188559372.0, "step": 4965 }, { "epoch": 0.6317262434804732, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.428645610809326, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8740051984786987, "num_tokens": 188595337.0, "step": 4966 }, { "epoch": 0.6318534537590638, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.553560972213745, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8599564433097839, "num_tokens": 188632988.0, "step": 4967 }, { "epoch": 0.6319806640376543, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4218814373016357, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8722236156463623, "num_tokens": 188667205.0, "step": 4968 }, { "epoch": 0.6321078743162447, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5656464099884033, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8541218042373657, "num_tokens": 188707717.0, "step": 4969 }, { "epoch": 0.6322350845948352, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4138565063476562, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8666351437568665, "num_tokens": 188742811.0, "step": 4970 }, { "epoch": 0.6323622948734258, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5885281562805176, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8587561249732971, "num_tokens": 188781004.0, "step": 4971 }, { "epoch": 0.6324895051520163, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4233455657958984, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.854870080947876, "num_tokens": 188813923.0, "step": 4972 }, { "epoch": 0.6326167154306068, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.580007553100586, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8624276518821716, "num_tokens": 188855375.0, "step": 4973 }, { "epoch": 0.6327439257091974, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.401909112930298, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8752550482749939, "num_tokens": 188892807.0, "step": 4974 }, { "epoch": 0.6328711359877878, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6000704765319824, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8658969402313232, "num_tokens": 188932942.0, "step": 4975 }, { "epoch": 0.6329983462663783, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4066479206085205, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8578063249588013, "num_tokens": 188968206.0, "step": 4976 }, { "epoch": 0.6331255565449688, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6469430923461914, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8525531888008118, "num_tokens": 189001315.0, "step": 4977 }, { "epoch": 0.6332527668235594, "ewc_loss": 0.048828125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.397887706756592, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8413264751434326, "num_tokens": 189039729.0, "step": 4978 }, { "epoch": 0.6333799771021499, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.645297050476074, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8759791851043701, "num_tokens": 189077455.0, "step": 4979 }, { "epoch": 0.6335071873807404, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.382747173309326, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8501788377761841, "num_tokens": 189118417.0, "step": 4980 }, { "epoch": 0.6336343976593308, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.676764488220215, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.871364951133728, "num_tokens": 189155581.0, "step": 4981 }, { "epoch": 0.6337616079379214, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.394113302230835, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8588523864746094, "num_tokens": 189191192.0, "step": 4982 }, { "epoch": 0.6338888182165119, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6946048736572266, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8656939268112183, "num_tokens": 189230475.0, "step": 4983 }, { "epoch": 0.6340160284951024, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.406165361404419, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8590078353881836, "num_tokens": 189265949.0, "step": 4984 }, { "epoch": 0.6341432387736929, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.712273359298706, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8578976392745972, "num_tokens": 189302914.0, "step": 4985 }, { "epoch": 0.6342704490522835, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.395310163497925, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.870432436466217, "num_tokens": 189347741.0, "step": 4986 }, { "epoch": 0.6343976593308739, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7204129695892334, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8681859970092773, "num_tokens": 189389672.0, "step": 4987 }, { "epoch": 0.6345248696094644, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4201927185058594, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8698195219039917, "num_tokens": 189425418.0, "step": 4988 }, { "epoch": 0.6346520798880549, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7440507411956787, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8693028688430786, "num_tokens": 189457042.0, "step": 4989 }, { "epoch": 0.6347792901666455, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.439802885055542, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8483622074127197, "num_tokens": 189497135.0, "step": 4990 }, { "epoch": 0.634906500445236, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7209131717681885, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8788686990737915, "num_tokens": 189534829.0, "step": 4991 }, { "epoch": 0.6350337107238265, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.448768377304077, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8541942834854126, "num_tokens": 189570535.0, "step": 4992 }, { "epoch": 0.635160921002417, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7138137817382812, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8672059774398804, "num_tokens": 189606448.0, "step": 4993 }, { "epoch": 0.6352881312810075, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.439075469970703, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8608745336532593, "num_tokens": 189650880.0, "step": 4994 }, { "epoch": 0.635415341559598, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.698934316635132, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8446431756019592, "num_tokens": 189690866.0, "step": 4995 }, { "epoch": 0.6355425518381885, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4711849689483643, "learning_rate": 1e-05, "loss": 0.5557, "mean_token_accuracy": 0.8311632871627808, "num_tokens": 189726891.0, "step": 4996 }, { "epoch": 0.6356697621167791, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.67326283454895, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.860629677772522, "num_tokens": 189763893.0, "step": 4997 }, { "epoch": 0.6357969723953696, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4711451530456543, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8446915745735168, "num_tokens": 189804122.0, "step": 4998 }, { "epoch": 0.63592418267396, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.658036947250366, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8558719158172607, "num_tokens": 189836355.0, "step": 4999 }, { "epoch": 0.6360513929525505, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.467458724975586, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8628227710723877, "num_tokens": 189871839.0, "step": 5000 }, { "epoch": 0.6361786032311411, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6132302284240723, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8701403141021729, "num_tokens": 189915631.0, "step": 5001 }, { "epoch": 0.6363058135097316, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.46091628074646, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8528491258621216, "num_tokens": 189953635.0, "step": 5002 }, { "epoch": 0.6364330237883221, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.603106737136841, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8622901439666748, "num_tokens": 189999338.0, "step": 5003 }, { "epoch": 0.6365602340669126, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.506155252456665, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.861047625541687, "num_tokens": 190045804.0, "step": 5004 }, { "epoch": 0.6366874443455031, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5647592544555664, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.857003927230835, "num_tokens": 190083350.0, "step": 5005 }, { "epoch": 0.6368146546240936, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.468540906906128, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8512948751449585, "num_tokens": 190121910.0, "step": 5006 }, { "epoch": 0.6369418649026841, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.536625385284424, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.863164484500885, "num_tokens": 190163857.0, "step": 5007 }, { "epoch": 0.6370690751812746, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4638280868530273, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8531352877616882, "num_tokens": 190204337.0, "step": 5008 }, { "epoch": 0.6371962854598652, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.537302017211914, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.872093915939331, "num_tokens": 190240646.0, "step": 5009 }, { "epoch": 0.6373234957384557, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5060553550720215, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8549661636352539, "num_tokens": 190273507.0, "step": 5010 }, { "epoch": 0.6374507060170462, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5183796882629395, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8589153289794922, "num_tokens": 190306876.0, "step": 5011 }, { "epoch": 0.6375779162956366, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.48258638381958, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8493895530700684, "num_tokens": 190343126.0, "step": 5012 }, { "epoch": 0.6377051265742272, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.505610227584839, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8498645424842834, "num_tokens": 190382089.0, "step": 5013 }, { "epoch": 0.6378323368528177, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4707400798797607, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8464494943618774, "num_tokens": 190419577.0, "step": 5014 }, { "epoch": 0.6379595471314082, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5066018104553223, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8466652631759644, "num_tokens": 190459725.0, "step": 5015 }, { "epoch": 0.6380867574099988, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4774227142333984, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8481546640396118, "num_tokens": 190493575.0, "step": 5016 }, { "epoch": 0.6382139676885893, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4939124584198, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8608460426330566, "num_tokens": 190530326.0, "step": 5017 }, { "epoch": 0.6383411779671797, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4703140258789062, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8577090501785278, "num_tokens": 190565727.0, "step": 5018 }, { "epoch": 0.6384683882457702, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4967758655548096, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8485926389694214, "num_tokens": 190600163.0, "step": 5019 }, { "epoch": 0.6385955985243608, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4696686267852783, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8495447635650635, "num_tokens": 190640801.0, "step": 5020 }, { "epoch": 0.6387228088029513, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.485105037689209, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8787695169448853, "num_tokens": 190679886.0, "step": 5021 }, { "epoch": 0.6388500190815418, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.451094627380371, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8574967384338379, "num_tokens": 190726246.0, "step": 5022 }, { "epoch": 0.6389772293601323, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.491560220718384, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8585073947906494, "num_tokens": 190765496.0, "step": 5023 }, { "epoch": 0.6391044396387228, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.45692777633667, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8546345233917236, "num_tokens": 190808770.0, "step": 5024 }, { "epoch": 0.6392316499173133, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.506312370300293, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8676543235778809, "num_tokens": 190844242.0, "step": 5025 }, { "epoch": 0.6393588601959038, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.45516037940979, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.8404086232185364, "num_tokens": 190887658.0, "step": 5026 }, { "epoch": 0.6394860704744944, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.507310390472412, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.862907886505127, "num_tokens": 190922322.0, "step": 5027 }, { "epoch": 0.6396132807530849, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001354217529296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.446737051010132, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.853386640548706, "num_tokens": 190959085.0, "step": 5028 }, { "epoch": 0.6397404910316754, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.527271032333374, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8726075291633606, "num_tokens": 190991646.0, "step": 5029 }, { "epoch": 0.6398677013102658, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.422544002532959, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8593522310256958, "num_tokens": 191039489.0, "step": 5030 }, { "epoch": 0.6399949115888564, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5492844581604004, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8545655012130737, "num_tokens": 191078166.0, "step": 5031 }, { "epoch": 0.6401221218674469, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.423293352127075, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8386167287826538, "num_tokens": 191115981.0, "step": 5032 }, { "epoch": 0.6402493321460374, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.563495635986328, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8598170280456543, "num_tokens": 191153989.0, "step": 5033 }, { "epoch": 0.6403765424246279, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.410768985748291, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8596519231796265, "num_tokens": 191192799.0, "step": 5034 }, { "epoch": 0.6405037527032185, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5763328075408936, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8581446409225464, "num_tokens": 191231923.0, "step": 5035 }, { "epoch": 0.6406309629818089, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.464478015899658, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8589935302734375, "num_tokens": 191271650.0, "step": 5036 }, { "epoch": 0.6407581732603994, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5831046104431152, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8610751628875732, "num_tokens": 191300212.0, "step": 5037 }, { "epoch": 0.6408853835389899, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.410452127456665, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8541775941848755, "num_tokens": 191336163.0, "step": 5038 }, { "epoch": 0.6410125938175805, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.581763982772827, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8570563197135925, "num_tokens": 191374749.0, "step": 5039 }, { "epoch": 0.641139804096171, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4005086421966553, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8608502149581909, "num_tokens": 191413342.0, "step": 5040 }, { "epoch": 0.6412670143747615, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6087870597839355, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.865825355052948, "num_tokens": 191450332.0, "step": 5041 }, { "epoch": 0.6413942246533519, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.3931641578674316, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8622784614562988, "num_tokens": 191486969.0, "step": 5042 }, { "epoch": 0.6415214349319425, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6284921169281006, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8627189993858337, "num_tokens": 191525074.0, "step": 5043 }, { "epoch": 0.641648645210533, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.3816678524017334, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8569194078445435, "num_tokens": 191564324.0, "step": 5044 }, { "epoch": 0.6417758554891235, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6568822860717773, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8523085117340088, "num_tokens": 191604081.0, "step": 5045 }, { "epoch": 0.641903065767714, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.402390241622925, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.84535151720047, "num_tokens": 191639811.0, "step": 5046 }, { "epoch": 0.6420302760463046, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6635618209838867, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8670487999916077, "num_tokens": 191680941.0, "step": 5047 }, { "epoch": 0.642157486324895, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.401097536087036, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8527541756629944, "num_tokens": 191718039.0, "step": 5048 }, { "epoch": 0.6422846966034855, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.684069871902466, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8585104942321777, "num_tokens": 191759837.0, "step": 5049 }, { "epoch": 0.6424119068820761, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.391101121902466, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8540653586387634, "num_tokens": 191801791.0, "step": 5050 }, { "epoch": 0.6425391171606666, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.701969861984253, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8683469295501709, "num_tokens": 191838193.0, "step": 5051 }, { "epoch": 0.6426663274392571, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4162445068359375, "learning_rate": 1e-05, "loss": 0.5555, "mean_token_accuracy": 0.8319336771965027, "num_tokens": 191874811.0, "step": 5052 }, { "epoch": 0.6427935377178476, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.705402374267578, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8528410792350769, "num_tokens": 191921289.0, "step": 5053 }, { "epoch": 0.6429207479964381, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4110114574432373, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8603429198265076, "num_tokens": 191961796.0, "step": 5054 }, { "epoch": 0.6430479582750286, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.705852746963501, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8774988055229187, "num_tokens": 192000853.0, "step": 5055 }, { "epoch": 0.6431751685536191, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4433469772338867, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8618108630180359, "num_tokens": 192029126.0, "step": 5056 }, { "epoch": 0.6433023788322096, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7018818855285645, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8516124486923218, "num_tokens": 192070472.0, "step": 5057 }, { "epoch": 0.6434295891108002, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.442303419113159, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8484611511230469, "num_tokens": 192107844.0, "step": 5058 }, { "epoch": 0.6435567993893907, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6944453716278076, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8596246838569641, "num_tokens": 192140637.0, "step": 5059 }, { "epoch": 0.6436840096679812, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4410808086395264, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8668646812438965, "num_tokens": 192176177.0, "step": 5060 }, { "epoch": 0.6438112199465716, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6761152744293213, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8639547824859619, "num_tokens": 192218131.0, "step": 5061 }, { "epoch": 0.6439384302251622, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4355945587158203, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.862301230430603, "num_tokens": 192258720.0, "step": 5062 }, { "epoch": 0.6440656405037527, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.678162097930908, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8486230969429016, "num_tokens": 192303197.0, "step": 5063 }, { "epoch": 0.6441928507823432, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.464833974838257, "learning_rate": 1e-05, "loss": 0.5479, "mean_token_accuracy": 0.8362382650375366, "num_tokens": 192340882.0, "step": 5064 }, { "epoch": 0.6443200610609338, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.647311210632324, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8626776933670044, "num_tokens": 192383596.0, "step": 5065 }, { "epoch": 0.6444472713395243, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4518532752990723, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8583260774612427, "num_tokens": 192423861.0, "step": 5066 }, { "epoch": 0.6445744816181147, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6440324783325195, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8667681813240051, "num_tokens": 192456564.0, "step": 5067 }, { "epoch": 0.6447016918967052, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4454715251922607, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8553586006164551, "num_tokens": 192496917.0, "step": 5068 }, { "epoch": 0.6448289021752958, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6294045448303223, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.857927680015564, "num_tokens": 192539548.0, "step": 5069 }, { "epoch": 0.6449561124538863, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.454127073287964, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8477636575698853, "num_tokens": 192577430.0, "step": 5070 }, { "epoch": 0.6450833227324768, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6081292629241943, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8702927827835083, "num_tokens": 192617343.0, "step": 5071 }, { "epoch": 0.6452105330110673, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.432382583618164, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8612757921218872, "num_tokens": 192662274.0, "step": 5072 }, { "epoch": 0.6453377432896578, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6171696186065674, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8660207390785217, "num_tokens": 192703305.0, "step": 5073 }, { "epoch": 0.6454649535682483, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.449244737625122, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8545914888381958, "num_tokens": 192738325.0, "step": 5074 }, { "epoch": 0.6455921638468388, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6008858680725098, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8728014230728149, "num_tokens": 192778285.0, "step": 5075 }, { "epoch": 0.6457193741254293, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4253180027008057, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8632315397262573, "num_tokens": 192822822.0, "step": 5076 }, { "epoch": 0.6458465844040199, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.612086057662964, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8552629351615906, "num_tokens": 192857312.0, "step": 5077 }, { "epoch": 0.6459737946826104, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4201273918151855, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8657023906707764, "num_tokens": 192892363.0, "step": 5078 }, { "epoch": 0.6461010049612008, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6264288425445557, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8487492799758911, "num_tokens": 192931774.0, "step": 5079 }, { "epoch": 0.6462282152397913, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.426447629928589, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8669973611831665, "num_tokens": 192964463.0, "step": 5080 }, { "epoch": 0.6463554255183819, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6296181678771973, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8702505826950073, "num_tokens": 192999807.0, "step": 5081 }, { "epoch": 0.6464826357969724, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4148662090301514, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8533101081848145, "num_tokens": 193041408.0, "step": 5082 }, { "epoch": 0.6466098460755629, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6535816192626953, "learning_rate": 1e-05, "loss": 0.5554, "mean_token_accuracy": 0.8334660530090332, "num_tokens": 193078104.0, "step": 5083 }, { "epoch": 0.6467370563541535, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.00013637542724609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.409881353378296, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8418896198272705, "num_tokens": 193118821.0, "step": 5084 }, { "epoch": 0.6468642666327439, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.663456439971924, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8543883562088013, "num_tokens": 193152633.0, "step": 5085 }, { "epoch": 0.6469914769113344, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.402965784072876, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8631134033203125, "num_tokens": 193191398.0, "step": 5086 }, { "epoch": 0.6471186871899249, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6764912605285645, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8499691486358643, "num_tokens": 193229612.0, "step": 5087 }, { "epoch": 0.6472458974685155, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4093101024627686, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.844934344291687, "num_tokens": 193271993.0, "step": 5088 }, { "epoch": 0.647373107747106, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.693740129470825, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8499875068664551, "num_tokens": 193308146.0, "step": 5089 }, { "epoch": 0.6475003180256965, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.41200590133667, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8640695810317993, "num_tokens": 193348448.0, "step": 5090 }, { "epoch": 0.6476275283042869, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.711308717727661, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8597770929336548, "num_tokens": 193379279.0, "step": 5091 }, { "epoch": 0.6477547385828775, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.425879716873169, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8639244437217712, "num_tokens": 193415999.0, "step": 5092 }, { "epoch": 0.647881948861468, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7044448852539062, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8620498180389404, "num_tokens": 193450287.0, "step": 5093 }, { "epoch": 0.6480091591400585, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.430936574935913, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8620949983596802, "num_tokens": 193485980.0, "step": 5094 }, { "epoch": 0.648136369418649, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.707441806793213, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8543723821640015, "num_tokens": 193525179.0, "step": 5095 }, { "epoch": 0.6482635796972396, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.443912982940674, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8573212623596191, "num_tokens": 193561580.0, "step": 5096 }, { "epoch": 0.64839078997583, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6948845386505127, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8480277061462402, "num_tokens": 193598218.0, "step": 5097 }, { "epoch": 0.6485180002544205, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.459758758544922, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8476735353469849, "num_tokens": 193635256.0, "step": 5098 }, { "epoch": 0.648645210533011, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6709461212158203, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8498389720916748, "num_tokens": 193672957.0, "step": 5099 }, { "epoch": 0.6487724208116016, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.452162027359009, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8677294254302979, "num_tokens": 193710575.0, "step": 5100 }, { "epoch": 0.6488996310901921, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.652618885040283, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8672863245010376, "num_tokens": 193749006.0, "step": 5101 }, { "epoch": 0.6490268413687826, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4589223861694336, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8429791927337646, "num_tokens": 193788975.0, "step": 5102 }, { "epoch": 0.649154051647373, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.638572931289673, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8586515784263611, "num_tokens": 193822898.0, "step": 5103 }, { "epoch": 0.6492812619259636, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.442902088165283, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8731481432914734, "num_tokens": 193862439.0, "step": 5104 }, { "epoch": 0.6494084722045541, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.620835304260254, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8578455448150635, "num_tokens": 193906721.0, "step": 5105 }, { "epoch": 0.6495356824831446, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4521234035491943, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8623827695846558, "num_tokens": 193940521.0, "step": 5106 }, { "epoch": 0.6496628927617352, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6118526458740234, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8501217365264893, "num_tokens": 193979340.0, "step": 5107 }, { "epoch": 0.6497901030403257, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4460947513580322, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8587779402732849, "num_tokens": 194018562.0, "step": 5108 }, { "epoch": 0.6499173133189162, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6096510887145996, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8318277597427368, "num_tokens": 194056804.0, "step": 5109 }, { "epoch": 0.6500445235975066, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4421274662017822, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8599635362625122, "num_tokens": 194095194.0, "step": 5110 }, { "epoch": 0.6501717338760972, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6017403602600098, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8567574620246887, "num_tokens": 194135061.0, "step": 5111 }, { "epoch": 0.6502989441546877, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4331045150756836, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.861750602722168, "num_tokens": 194169017.0, "step": 5112 }, { "epoch": 0.6504261544332782, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5981056690216064, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8496065735816956, "num_tokens": 194213424.0, "step": 5113 }, { "epoch": 0.6505533647118688, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.433753728866577, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8695948123931885, "num_tokens": 194247465.0, "step": 5114 }, { "epoch": 0.6506805749904593, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5984060764312744, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8656460046768188, "num_tokens": 194282575.0, "step": 5115 }, { "epoch": 0.6508077852690497, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.452653646469116, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8601909279823303, "num_tokens": 194311822.0, "step": 5116 }, { "epoch": 0.6509349955476402, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5850584506988525, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8483316898345947, "num_tokens": 194346990.0, "step": 5117 }, { "epoch": 0.6510622058262308, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4321014881134033, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8611137866973877, "num_tokens": 194384836.0, "step": 5118 }, { "epoch": 0.6511894161048213, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5850155353546143, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8709949851036072, "num_tokens": 194418080.0, "step": 5119 }, { "epoch": 0.6513166263834118, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4229586124420166, "learning_rate": 1e-05, "loss": 0.5118, "mean_token_accuracy": 0.8467903137207031, "num_tokens": 194460111.0, "step": 5120 }, { "epoch": 0.6514438366620023, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.59566593170166, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8736174702644348, "num_tokens": 194498157.0, "step": 5121 }, { "epoch": 0.6515710469405928, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.423954963684082, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8611622452735901, "num_tokens": 194538928.0, "step": 5122 }, { "epoch": 0.6516982572191833, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5960614681243896, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8618410229682922, "num_tokens": 194582276.0, "step": 5123 }, { "epoch": 0.6518254674977738, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.420578718185425, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8598436117172241, "num_tokens": 194620197.0, "step": 5124 }, { "epoch": 0.6519526777763643, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.6148719787597656, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8665969967842102, "num_tokens": 194654634.0, "step": 5125 }, { "epoch": 0.6520798880549549, "ewc_loss": 0.049072265625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.40838623046875, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8536593914031982, "num_tokens": 194696400.0, "step": 5126 }, { "epoch": 0.6522070983335454, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6203174591064453, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8656400442123413, "num_tokens": 194738326.0, "step": 5127 }, { "epoch": 0.6523343086121358, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.402974843978882, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8623433113098145, "num_tokens": 194778550.0, "step": 5128 }, { "epoch": 0.6524615188907263, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6436140537261963, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8565046787261963, "num_tokens": 194815124.0, "step": 5129 }, { "epoch": 0.6525887291693169, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4117088317871094, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8596996068954468, "num_tokens": 194851372.0, "step": 5130 }, { "epoch": 0.6527159394479074, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6484410762786865, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8715246915817261, "num_tokens": 194886705.0, "step": 5131 }, { "epoch": 0.6528431497264979, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.3964884281158447, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8601129055023193, "num_tokens": 194930372.0, "step": 5132 }, { "epoch": 0.6529703600050885, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6684072017669678, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8628382682800293, "num_tokens": 194971869.0, "step": 5133 }, { "epoch": 0.6530975702836789, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.409261465072632, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8529926538467407, "num_tokens": 195011271.0, "step": 5134 }, { "epoch": 0.6532247805622694, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.690877676010132, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8613095283508301, "num_tokens": 195046328.0, "step": 5135 }, { "epoch": 0.6533519908408599, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4090487957000732, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8653600215911865, "num_tokens": 195086424.0, "step": 5136 }, { "epoch": 0.6534792011194505, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.703242778778076, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8519715666770935, "num_tokens": 195121751.0, "step": 5137 }, { "epoch": 0.653606411398041, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4277939796447754, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8619571328163147, "num_tokens": 195158231.0, "step": 5138 }, { "epoch": 0.6537336216766315, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.698915958404541, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.858054518699646, "num_tokens": 195195009.0, "step": 5139 }, { "epoch": 0.6538608319552219, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4210352897644043, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8695554137229919, "num_tokens": 195232892.0, "step": 5140 }, { "epoch": 0.6539880422338125, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6875691413879395, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8583657741546631, "num_tokens": 195283233.0, "step": 5141 }, { "epoch": 0.654115252512403, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.430630683898926, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8780938982963562, "num_tokens": 195319402.0, "step": 5142 }, { "epoch": 0.6542424627909935, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.697514772415161, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8652173280715942, "num_tokens": 195358330.0, "step": 5143 }, { "epoch": 0.654369673069584, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.426954984664917, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8737353086471558, "num_tokens": 195398074.0, "step": 5144 }, { "epoch": 0.6544968833481746, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.698848247528076, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8730469942092896, "num_tokens": 195435684.0, "step": 5145 }, { "epoch": 0.654624093626765, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4308218955993652, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8708906173706055, "num_tokens": 195475126.0, "step": 5146 }, { "epoch": 0.6547513039053555, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6965205669403076, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8571949601173401, "num_tokens": 195516398.0, "step": 5147 }, { "epoch": 0.654878514183946, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4410104751586914, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8643379211425781, "num_tokens": 195558777.0, "step": 5148 }, { "epoch": 0.6550057244625366, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6825203895568848, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8545026779174805, "num_tokens": 195598663.0, "step": 5149 }, { "epoch": 0.6551329347411271, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4771883487701416, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8458687663078308, "num_tokens": 195628594.0, "step": 5150 }, { "epoch": 0.6552601450197176, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.662012815475464, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8537783622741699, "num_tokens": 195662484.0, "step": 5151 }, { "epoch": 0.655387355298308, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4649367332458496, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8495931029319763, "num_tokens": 195697753.0, "step": 5152 }, { "epoch": 0.6555145655768986, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6203250885009766, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8710634708404541, "num_tokens": 195740931.0, "step": 5153 }, { "epoch": 0.6556417758554891, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.454319953918457, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.8457098007202148, "num_tokens": 195783056.0, "step": 5154 }, { "epoch": 0.6557689861340796, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6275596618652344, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8466162085533142, "num_tokens": 195819515.0, "step": 5155 }, { "epoch": 0.6558961964126702, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4506828784942627, "learning_rate": 1e-05, "loss": 0.5161, "mean_token_accuracy": 0.8448268175125122, "num_tokens": 195861039.0, "step": 5156 }, { "epoch": 0.6560234066912607, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6162233352661133, "learning_rate": 1e-05, "loss": 0.5209, "mean_token_accuracy": 0.8431566953659058, "num_tokens": 195897818.0, "step": 5157 }, { "epoch": 0.6561506169698512, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4396870136260986, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8745099306106567, "num_tokens": 195930752.0, "step": 5158 }, { "epoch": 0.6562778272484416, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.625615358352661, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8548913598060608, "num_tokens": 195965036.0, "step": 5159 }, { "epoch": 0.6564050375270322, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.439603090286255, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8580012321472168, "num_tokens": 195999293.0, "step": 5160 }, { "epoch": 0.6565322478056227, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6167497634887695, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8495595455169678, "num_tokens": 196039066.0, "step": 5161 }, { "epoch": 0.6566594580842132, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001373291015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4373772144317627, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8509542346000671, "num_tokens": 196077805.0, "step": 5162 }, { "epoch": 0.6567866683628037, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6261167526245117, "learning_rate": 1e-05, "loss": 0.5339, "mean_token_accuracy": 0.8371238708496094, "num_tokens": 196117460.0, "step": 5163 }, { "epoch": 0.6569138786413943, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4199273586273193, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8605303764343262, "num_tokens": 196154855.0, "step": 5164 }, { "epoch": 0.6570410889199847, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6288113594055176, "learning_rate": 1e-05, "loss": 0.521, "mean_token_accuracy": 0.8435510396957397, "num_tokens": 196197360.0, "step": 5165 }, { "epoch": 0.6571682991985752, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.420262098312378, "learning_rate": 1e-05, "loss": 0.5403, "mean_token_accuracy": 0.8375908136367798, "num_tokens": 196237419.0, "step": 5166 }, { "epoch": 0.6572955094771658, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.63956356048584, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8486138582229614, "num_tokens": 196275533.0, "step": 5167 }, { "epoch": 0.6574227197557563, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.418159246444702, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.865454375743866, "num_tokens": 196312269.0, "step": 5168 }, { "epoch": 0.6575499300343468, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6441829204559326, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8534775376319885, "num_tokens": 196349338.0, "step": 5169 }, { "epoch": 0.6576771403129373, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.411618709564209, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8466652631759644, "num_tokens": 196396000.0, "step": 5170 }, { "epoch": 0.6578043505915278, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6683413982391357, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8518145084381104, "num_tokens": 196428119.0, "step": 5171 }, { "epoch": 0.6579315608701183, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4073195457458496, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.845763087272644, "num_tokens": 196470466.0, "step": 5172 }, { "epoch": 0.6580587711487088, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6808998584747314, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8592169284820557, "num_tokens": 196508883.0, "step": 5173 }, { "epoch": 0.6581859814272993, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.409282684326172, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8466563820838928, "num_tokens": 196552762.0, "step": 5174 }, { "epoch": 0.6583131917058899, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.699608087539673, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8791561126708984, "num_tokens": 196581394.0, "step": 5175 }, { "epoch": 0.6584404019844804, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.410747528076172, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8641905188560486, "num_tokens": 196623237.0, "step": 5176 }, { "epoch": 0.6585676122630708, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7039945125579834, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8714538812637329, "num_tokens": 196659219.0, "step": 5177 }, { "epoch": 0.6586948225416613, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4279839992523193, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8533153533935547, "num_tokens": 196700102.0, "step": 5178 }, { "epoch": 0.6588220328202519, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6881980895996094, "learning_rate": 1e-05, "loss": 0.3893, "mean_token_accuracy": 0.8862389326095581, "num_tokens": 196740569.0, "step": 5179 }, { "epoch": 0.6589492430988424, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4461920261383057, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8530923128128052, "num_tokens": 196778634.0, "step": 5180 }, { "epoch": 0.6590764533774329, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.717696189880371, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8499539494514465, "num_tokens": 196805021.0, "step": 5181 }, { "epoch": 0.6592036636560235, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4476752281188965, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8622573614120483, "num_tokens": 196839676.0, "step": 5182 }, { "epoch": 0.6593308739346139, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6870224475860596, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8667277097702026, "num_tokens": 196877564.0, "step": 5183 }, { "epoch": 0.6594580842132044, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.460447311401367, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8517251014709473, "num_tokens": 196913819.0, "step": 5184 }, { "epoch": 0.6595852944917949, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6692676544189453, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8452799916267395, "num_tokens": 196951019.0, "step": 5185 }, { "epoch": 0.6597125047703855, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.446061134338379, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8572050333023071, "num_tokens": 196995112.0, "step": 5186 }, { "epoch": 0.659839715048976, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.64998722076416, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8628453612327576, "num_tokens": 197035334.0, "step": 5187 }, { "epoch": 0.6599669253275665, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4476161003112793, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8499240875244141, "num_tokens": 197074234.0, "step": 5188 }, { "epoch": 0.6600941356061569, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.654461145401001, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8537253737449646, "num_tokens": 197110977.0, "step": 5189 }, { "epoch": 0.6602213458847475, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4511818885803223, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8484656810760498, "num_tokens": 197148634.0, "step": 5190 }, { "epoch": 0.660348556163338, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6364076137542725, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8633624315261841, "num_tokens": 197186361.0, "step": 5191 }, { "epoch": 0.6604757664419285, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4468979835510254, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8564764261245728, "num_tokens": 197222642.0, "step": 5192 }, { "epoch": 0.660602976720519, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6261444091796875, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8739409446716309, "num_tokens": 197263320.0, "step": 5193 }, { "epoch": 0.6607301869991096, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.475126266479492, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8486522436141968, "num_tokens": 197294041.0, "step": 5194 }, { "epoch": 0.6608573972777, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6169259548187256, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8651527166366577, "num_tokens": 197324878.0, "step": 5195 }, { "epoch": 0.6609846075562905, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.446133613586426, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8693559169769287, "num_tokens": 197368318.0, "step": 5196 }, { "epoch": 0.661111817834881, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6037418842315674, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8613710403442383, "num_tokens": 197407665.0, "step": 5197 }, { "epoch": 0.6612390281134716, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4415290355682373, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8595694303512573, "num_tokens": 197445812.0, "step": 5198 }, { "epoch": 0.6613662383920621, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.603346347808838, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8684086799621582, "num_tokens": 197482733.0, "step": 5199 }, { "epoch": 0.6614934486706526, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.443790912628174, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8585507869720459, "num_tokens": 197520874.0, "step": 5200 }, { "epoch": 0.661620658949243, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5921132564544678, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8550277948379517, "num_tokens": 197563422.0, "step": 5201 }, { "epoch": 0.6617478692278336, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.43802547454834, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8525516986846924, "num_tokens": 197603632.0, "step": 5202 }, { "epoch": 0.6618750795064241, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5939149856567383, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8616143465042114, "num_tokens": 197642090.0, "step": 5203 }, { "epoch": 0.6620022897850146, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.444993257522583, "learning_rate": 1e-05, "loss": 0.5618, "mean_token_accuracy": 0.8310927748680115, "num_tokens": 197679294.0, "step": 5204 }, { "epoch": 0.6621295000636052, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5898149013519287, "learning_rate": 1e-05, "loss": 0.5475, "mean_token_accuracy": 0.8396180272102356, "num_tokens": 197721254.0, "step": 5205 }, { "epoch": 0.6622567103421957, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4093635082244873, "learning_rate": 1e-05, "loss": 0.4098, "mean_token_accuracy": 0.877729058265686, "num_tokens": 197764656.0, "step": 5206 }, { "epoch": 0.6623839206207861, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6077282428741455, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.860630214214325, "num_tokens": 197805900.0, "step": 5207 }, { "epoch": 0.6625111308993766, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4105236530303955, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8642704486846924, "num_tokens": 197842752.0, "step": 5208 }, { "epoch": 0.6626383411779672, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6156842708587646, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.869931697845459, "num_tokens": 197883531.0, "step": 5209 }, { "epoch": 0.6627655514565577, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.41265869140625, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8640637993812561, "num_tokens": 197920875.0, "step": 5210 }, { "epoch": 0.6628927617351482, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.643162965774536, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.861760139465332, "num_tokens": 197955194.0, "step": 5211 }, { "epoch": 0.6630199720137387, "ewc_loss": 0.04931640625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.528594970703125e-05, "grad_norm": 2.4011592864990234, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8706197738647461, "num_tokens": 197991730.0, "step": 5212 }, { "epoch": 0.6631471822923293, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.655646562576294, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8648241758346558, "num_tokens": 198030032.0, "step": 5213 }, { "epoch": 0.6632743925709197, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4052059650421143, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8630553483963013, "num_tokens": 198068274.0, "step": 5214 }, { "epoch": 0.6634016028495102, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6742024421691895, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8555010557174683, "num_tokens": 198105326.0, "step": 5215 }, { "epoch": 0.6635288131281007, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4025843143463135, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8728821873664856, "num_tokens": 198147485.0, "step": 5216 }, { "epoch": 0.6636560234066913, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6955084800720215, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8654163479804993, "num_tokens": 198184928.0, "step": 5217 }, { "epoch": 0.6637832336852818, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4198713302612305, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8618342876434326, "num_tokens": 198224641.0, "step": 5218 }, { "epoch": 0.6639104439638723, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6917102336883545, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.870569109916687, "num_tokens": 198263821.0, "step": 5219 }, { "epoch": 0.6640376542424627, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.42875599861145, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8627251386642456, "num_tokens": 198300315.0, "step": 5220 }, { "epoch": 0.6641648645210533, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6970930099487305, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8657330274581909, "num_tokens": 198336543.0, "step": 5221 }, { "epoch": 0.6642920747996438, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4339356422424316, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8629951477050781, "num_tokens": 198374728.0, "step": 5222 }, { "epoch": 0.6644192850782343, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6798081398010254, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8629041910171509, "num_tokens": 198414504.0, "step": 5223 }, { "epoch": 0.6645464953568249, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4398255348205566, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8724373579025269, "num_tokens": 198449976.0, "step": 5224 }, { "epoch": 0.6646737056354154, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6680822372436523, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8655035495758057, "num_tokens": 198489522.0, "step": 5225 }, { "epoch": 0.6648009159140058, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4409682750701904, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8698699474334717, "num_tokens": 198524460.0, "step": 5226 }, { "epoch": 0.6649281261925963, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.64701509475708, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8653765320777893, "num_tokens": 198569139.0, "step": 5227 }, { "epoch": 0.6650553364711869, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4622459411621094, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8790845274925232, "num_tokens": 198597166.0, "step": 5228 }, { "epoch": 0.6651825467497774, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.640962839126587, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.865114688873291, "num_tokens": 198631646.0, "step": 5229 }, { "epoch": 0.6653097570283679, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.440932035446167, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8499436378479004, "num_tokens": 198675743.0, "step": 5230 }, { "epoch": 0.6654369673069584, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6330795288085938, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8572394847869873, "num_tokens": 198715826.0, "step": 5231 }, { "epoch": 0.6655641775855489, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4321367740631104, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8593493700027466, "num_tokens": 198757320.0, "step": 5232 }, { "epoch": 0.6656913878641394, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6429929733276367, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8689925074577332, "num_tokens": 198791110.0, "step": 5233 }, { "epoch": 0.6658185981427299, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.425548553466797, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8660480976104736, "num_tokens": 198828506.0, "step": 5234 }, { "epoch": 0.6659458084213205, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.646101951599121, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8442525863647461, "num_tokens": 198870759.0, "step": 5235 }, { "epoch": 0.666073018699911, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.421936511993408, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.869406521320343, "num_tokens": 198908178.0, "step": 5236 }, { "epoch": 0.6662002289785015, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6611955165863037, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.860297679901123, "num_tokens": 198942847.0, "step": 5237 }, { "epoch": 0.6663274392570919, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.422740936279297, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8592742085456848, "num_tokens": 198981444.0, "step": 5238 }, { "epoch": 0.6664546495356825, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6622776985168457, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8685737252235413, "num_tokens": 199019159.0, "step": 5239 }, { "epoch": 0.666581859814273, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4280569553375244, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.853783130645752, "num_tokens": 199056696.0, "step": 5240 }, { "epoch": 0.6667090700928635, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6584203243255615, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8721972703933716, "num_tokens": 199092358.0, "step": 5241 }, { "epoch": 0.666836280371454, "ewc_loss": 0.049560546875, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4350600242614746, "learning_rate": 1e-05, "loss": 0.558, "mean_token_accuracy": 0.8316432237625122, "num_tokens": 199135527.0, "step": 5242 }, { "epoch": 0.6669634906500446, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001392364501953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6617848873138428, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8626009225845337, "num_tokens": 199171886.0, "step": 5243 }, { "epoch": 0.667090700928635, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4323179721832275, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8572301864624023, "num_tokens": 199209807.0, "step": 5244 }, { "epoch": 0.6672179112072255, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6635751724243164, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.857037365436554, "num_tokens": 199244878.0, "step": 5245 }, { "epoch": 0.667345121485816, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.444725751876831, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8562684059143066, "num_tokens": 199279445.0, "step": 5246 }, { "epoch": 0.6674723317644066, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.649503231048584, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8766988515853882, "num_tokens": 199315504.0, "step": 5247 }, { "epoch": 0.6675995420429971, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4430344104766846, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8430273532867432, "num_tokens": 199354345.0, "step": 5248 }, { "epoch": 0.6677267523215876, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6353135108947754, "learning_rate": 1e-05, "loss": 0.522, "mean_token_accuracy": 0.8424859642982483, "num_tokens": 199401530.0, "step": 5249 }, { "epoch": 0.667853962600178, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4385645389556885, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8561279773712158, "num_tokens": 199440428.0, "step": 5250 }, { "epoch": 0.6679811728787686, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6301541328430176, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8770204782485962, "num_tokens": 199478875.0, "step": 5251 }, { "epoch": 0.6681083831573591, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.461092472076416, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.8397237658500671, "num_tokens": 199517287.0, "step": 5252 }, { "epoch": 0.6682355934359496, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6272852420806885, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8556960821151733, "num_tokens": 199553723.0, "step": 5253 }, { "epoch": 0.6683628037145402, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.445915460586548, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8608558177947998, "num_tokens": 199592059.0, "step": 5254 }, { "epoch": 0.6684900139931307, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6049795150756836, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8571263551712036, "num_tokens": 199633005.0, "step": 5255 }, { "epoch": 0.6686172242717211, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4453532695770264, "learning_rate": 1e-05, "loss": 0.5534, "mean_token_accuracy": 0.8325625658035278, "num_tokens": 199677507.0, "step": 5256 }, { "epoch": 0.6687444345503116, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.602725028991699, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8609623908996582, "num_tokens": 199716040.0, "step": 5257 }, { "epoch": 0.6688716448289022, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.445774793624878, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8652746081352234, "num_tokens": 199750438.0, "step": 5258 }, { "epoch": 0.6689988551074927, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6036911010742188, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.859778881072998, "num_tokens": 199786889.0, "step": 5259 }, { "epoch": 0.6691260653860832, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.437730550765991, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8633806705474854, "num_tokens": 199821673.0, "step": 5260 }, { "epoch": 0.6692532756646737, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.62111759185791, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8469852209091187, "num_tokens": 199854313.0, "step": 5261 }, { "epoch": 0.6693804859432643, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4245996475219727, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.863667368888855, "num_tokens": 199892853.0, "step": 5262 }, { "epoch": 0.6695076962218547, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.617061138153076, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8585237264633179, "num_tokens": 199931754.0, "step": 5263 }, { "epoch": 0.6696349065004452, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.414062738418579, "learning_rate": 1e-05, "loss": 0.4017, "mean_token_accuracy": 0.8730278611183167, "num_tokens": 199965570.0, "step": 5264 }, { "epoch": 0.6697621167790357, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.616544723510742, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8736175894737244, "num_tokens": 200004931.0, "step": 5265 }, { "epoch": 0.6698893270576263, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.432929277420044, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8704420328140259, "num_tokens": 200040315.0, "step": 5266 }, { "epoch": 0.6700165373362168, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6180789470672607, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8559508323669434, "num_tokens": 200081206.0, "step": 5267 }, { "epoch": 0.6701437476148073, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.417717933654785, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8654095530509949, "num_tokens": 200119975.0, "step": 5268 }, { "epoch": 0.6702709578933977, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6498312950134277, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8385085463523865, "num_tokens": 200154956.0, "step": 5269 }, { "epoch": 0.6703981681719883, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4077625274658203, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8517876863479614, "num_tokens": 200194499.0, "step": 5270 }, { "epoch": 0.6705253784505788, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6672568321228027, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8653269410133362, "num_tokens": 200227142.0, "step": 5271 }, { "epoch": 0.6706525887291693, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4258458614349365, "learning_rate": 1e-05, "loss": 0.5201, "mean_token_accuracy": 0.8466479182243347, "num_tokens": 200262123.0, "step": 5272 }, { "epoch": 0.6707797990077599, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6672184467315674, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8552943468093872, "num_tokens": 200299648.0, "step": 5273 }, { "epoch": 0.6709070092863504, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4263947010040283, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8422806262969971, "num_tokens": 200337191.0, "step": 5274 }, { "epoch": 0.6710342195649408, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.664820671081543, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8672855496406555, "num_tokens": 200371487.0, "step": 5275 }, { "epoch": 0.6711614298435313, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.415332794189453, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8631861805915833, "num_tokens": 200412634.0, "step": 5276 }, { "epoch": 0.6712886401221219, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6784753799438477, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8648035526275635, "num_tokens": 200449663.0, "step": 5277 }, { "epoch": 0.6714158504007124, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4216458797454834, "learning_rate": 1e-05, "loss": 0.5208, "mean_token_accuracy": 0.8459360599517822, "num_tokens": 200494962.0, "step": 5278 }, { "epoch": 0.6715430606793029, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6847259998321533, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8731305003166199, "num_tokens": 200531101.0, "step": 5279 }, { "epoch": 0.6716702709578934, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.426476240158081, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8669131994247437, "num_tokens": 200568725.0, "step": 5280 }, { "epoch": 0.6717974812364839, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6843512058258057, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8554373979568481, "num_tokens": 200607319.0, "step": 5281 }, { "epoch": 0.6719246915150744, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4360997676849365, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.856654942035675, "num_tokens": 200654791.0, "step": 5282 }, { "epoch": 0.6720519017936649, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.685683012008667, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8633072376251221, "num_tokens": 200694175.0, "step": 5283 }, { "epoch": 0.6721791120722554, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4459927082061768, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8602716326713562, "num_tokens": 200726675.0, "step": 5284 }, { "epoch": 0.672306322350846, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6914877891540527, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.849740743637085, "num_tokens": 200759358.0, "step": 5285 }, { "epoch": 0.6724335326294365, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.440664052963257, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8619832396507263, "num_tokens": 200797735.0, "step": 5286 }, { "epoch": 0.6725607429080269, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.673861503601074, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8561853170394897, "num_tokens": 200835446.0, "step": 5287 }, { "epoch": 0.6726879531866174, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4325990676879883, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8646926879882812, "num_tokens": 200881020.0, "step": 5288 }, { "epoch": 0.672815163465208, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.665257453918457, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8630551099777222, "num_tokens": 200921549.0, "step": 5289 }, { "epoch": 0.6729423737437985, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4641339778900146, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8487009406089783, "num_tokens": 200953991.0, "step": 5290 }, { "epoch": 0.673069584022389, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6521353721618652, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8625733852386475, "num_tokens": 200990240.0, "step": 5291 }, { "epoch": 0.6731967943009796, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4515185356140137, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8749325275421143, "num_tokens": 201022433.0, "step": 5292 }, { "epoch": 0.67332400457957, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6535158157348633, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8601763248443604, "num_tokens": 201055801.0, "step": 5293 }, { "epoch": 0.6734512148581605, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.431034803390503, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8518185019493103, "num_tokens": 201102141.0, "step": 5294 }, { "epoch": 0.673578425136751, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.658740758895874, "learning_rate": 1e-05, "loss": 0.5379, "mean_token_accuracy": 0.8493081331253052, "num_tokens": 201139834.0, "step": 5295 }, { "epoch": 0.6737056354153416, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.454123020172119, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8692312240600586, "num_tokens": 201169650.0, "step": 5296 }, { "epoch": 0.6738328456939321, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.633469581604004, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8635352849960327, "num_tokens": 201208716.0, "step": 5297 }, { "epoch": 0.6739600559725226, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.43165922164917, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8623272180557251, "num_tokens": 201253976.0, "step": 5298 }, { "epoch": 0.674087266251113, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6431915760040283, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8517671227455139, "num_tokens": 201291478.0, "step": 5299 }, { "epoch": 0.6742144765297036, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4334208965301514, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.857966423034668, "num_tokens": 201328824.0, "step": 5300 }, { "epoch": 0.6743416868082941, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.653555393218994, "learning_rate": 1e-05, "loss": 0.5123, "mean_token_accuracy": 0.8465045690536499, "num_tokens": 201363859.0, "step": 5301 }, { "epoch": 0.6744688970868846, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4236483573913574, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8676811456680298, "num_tokens": 201402635.0, "step": 5302 }, { "epoch": 0.6745961073654752, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6536622047424316, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8522160649299622, "num_tokens": 201444203.0, "step": 5303 }, { "epoch": 0.6747233176440657, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4331629276275635, "learning_rate": 1e-05, "loss": 0.5237, "mean_token_accuracy": 0.8450806140899658, "num_tokens": 201486397.0, "step": 5304 }, { "epoch": 0.6748505279226561, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6505415439605713, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8521961569786072, "num_tokens": 201528262.0, "step": 5305 }, { "epoch": 0.6749777382012466, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.438450813293457, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8594712018966675, "num_tokens": 201566274.0, "step": 5306 }, { "epoch": 0.6751049484798372, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6429126262664795, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8543800115585327, "num_tokens": 201610196.0, "step": 5307 }, { "epoch": 0.6752321587584277, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4346747398376465, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8534412980079651, "num_tokens": 201650144.0, "step": 5308 }, { "epoch": 0.6753593690370182, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6410152912139893, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8594725131988525, "num_tokens": 201689550.0, "step": 5309 }, { "epoch": 0.6754865793156087, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.42230224609375, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8742784857749939, "num_tokens": 201729807.0, "step": 5310 }, { "epoch": 0.6756137895941993, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6645123958587646, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8639192581176758, "num_tokens": 201759540.0, "step": 5311 }, { "epoch": 0.6757409998727897, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.428243637084961, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8551269769668579, "num_tokens": 201800067.0, "step": 5312 }, { "epoch": 0.6758682101513802, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6716060638427734, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8615728616714478, "num_tokens": 201832688.0, "step": 5313 }, { "epoch": 0.6759954204299707, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4403131008148193, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8545910120010376, "num_tokens": 201876090.0, "step": 5314 }, { "epoch": 0.6761226307085613, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.662752389907837, "learning_rate": 1e-05, "loss": 0.5193, "mean_token_accuracy": 0.8436914086341858, "num_tokens": 201914339.0, "step": 5315 }, { "epoch": 0.6762498409871518, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4387433528900146, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8699349164962769, "num_tokens": 201945746.0, "step": 5316 }, { "epoch": 0.6763770512657423, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6681814193725586, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.865575909614563, "num_tokens": 201978773.0, "step": 5317 }, { "epoch": 0.6765042615443327, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4250407218933105, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8597151041030884, "num_tokens": 202019838.0, "step": 5318 }, { "epoch": 0.6766314718229233, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.7222065925598145, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8613274693489075, "num_tokens": 202060825.0, "step": 5319 }, { "epoch": 0.6767586821015138, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4343020915985107, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8757251501083374, "num_tokens": 202092656.0, "step": 5320 }, { "epoch": 0.6768858923801043, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6971940994262695, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8447255492210388, "num_tokens": 202126982.0, "step": 5321 }, { "epoch": 0.6770131026586949, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.472095489501953, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8632069826126099, "num_tokens": 202157550.0, "step": 5322 }, { "epoch": 0.6771403129372854, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6746950149536133, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8467745184898376, "num_tokens": 202192168.0, "step": 5323 }, { "epoch": 0.6772675232158758, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4491889476776123, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.869895339012146, "num_tokens": 202226887.0, "step": 5324 }, { "epoch": 0.6773947334944663, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.648681640625, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8467361927032471, "num_tokens": 202269631.0, "step": 5325 }, { "epoch": 0.6775219437730569, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4590208530426025, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8581544756889343, "num_tokens": 202306951.0, "step": 5326 }, { "epoch": 0.6776491540516474, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6328155994415283, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8698839545249939, "num_tokens": 202342922.0, "step": 5327 }, { "epoch": 0.6777763643302379, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.449350118637085, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.857925534248352, "num_tokens": 202382125.0, "step": 5328 }, { "epoch": 0.6779035746088284, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.619062662124634, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8607770800590515, "num_tokens": 202424914.0, "step": 5329 }, { "epoch": 0.6780307848874189, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4555976390838623, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8600280284881592, "num_tokens": 202460828.0, "step": 5330 }, { "epoch": 0.6781579951660094, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.628472089767456, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8646179437637329, "num_tokens": 202492484.0, "step": 5331 }, { "epoch": 0.6782852054445999, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4778919219970703, "learning_rate": 1e-05, "loss": 0.5574, "mean_token_accuracy": 0.8295149207115173, "num_tokens": 202529079.0, "step": 5332 }, { "epoch": 0.6784124157231904, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6138556003570557, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8469488620758057, "num_tokens": 202563024.0, "step": 5333 }, { "epoch": 0.678539626001781, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.445690393447876, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8607662916183472, "num_tokens": 202601264.0, "step": 5334 }, { "epoch": 0.6786668362803715, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6022019386291504, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8553810119628906, "num_tokens": 202642933.0, "step": 5335 }, { "epoch": 0.6787940465589619, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4481360912323, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8688057065010071, "num_tokens": 202674851.0, "step": 5336 }, { "epoch": 0.6789212568375524, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6026458740234375, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8699978590011597, "num_tokens": 202711117.0, "step": 5337 }, { "epoch": 0.679048467116143, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.430529832839966, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8698595762252808, "num_tokens": 202749025.0, "step": 5338 }, { "epoch": 0.6791756773947335, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.615483283996582, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.8416947722434998, "num_tokens": 202790733.0, "step": 5339 }, { "epoch": 0.679302887673324, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4328441619873047, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8548024892807007, "num_tokens": 202825753.0, "step": 5340 }, { "epoch": 0.6794300979519146, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014019012451171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6152758598327637, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8740774393081665, "num_tokens": 202862069.0, "step": 5341 }, { "epoch": 0.679557308230505, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.414674758911133, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8619416952133179, "num_tokens": 202904526.0, "step": 5342 }, { "epoch": 0.6796845185090955, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.63960599899292, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8482550382614136, "num_tokens": 202942342.0, "step": 5343 }, { "epoch": 0.679811728787686, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.429706335067749, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.851765513420105, "num_tokens": 202974644.0, "step": 5344 }, { "epoch": 0.6799389390662766, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.63924241065979, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8727476596832275, "num_tokens": 203011728.0, "step": 5345 }, { "epoch": 0.6800661493448671, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4080123901367188, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8554472923278809, "num_tokens": 203051768.0, "step": 5346 }, { "epoch": 0.6801933596234576, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6491401195526123, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8504469990730286, "num_tokens": 203094943.0, "step": 5347 }, { "epoch": 0.680320569902048, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4247803688049316, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8623372912406921, "num_tokens": 203130587.0, "step": 5348 }, { "epoch": 0.6804477801806386, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6558148860931396, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8561195731163025, "num_tokens": 203170030.0, "step": 5349 }, { "epoch": 0.6805749904592291, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.424186944961548, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8557987213134766, "num_tokens": 203209382.0, "step": 5350 }, { "epoch": 0.6807022007378196, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.666529417037964, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8579444289207458, "num_tokens": 203247750.0, "step": 5351 }, { "epoch": 0.6808294110164101, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.427725315093994, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8645682334899902, "num_tokens": 203284443.0, "step": 5352 }, { "epoch": 0.6809566212950007, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6593551635742188, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8699315190315247, "num_tokens": 203324525.0, "step": 5353 }, { "epoch": 0.6810838315735911, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.416743755340576, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8633208274841309, "num_tokens": 203369633.0, "step": 5354 }, { "epoch": 0.6812110418521816, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6716394424438477, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8590196371078491, "num_tokens": 203409549.0, "step": 5355 }, { "epoch": 0.6813382521307721, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4322168827056885, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8518716096878052, "num_tokens": 203446856.0, "step": 5356 }, { "epoch": 0.6814654624093627, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6796367168426514, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.882480263710022, "num_tokens": 203478575.0, "step": 5357 }, { "epoch": 0.6815926726879532, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4285175800323486, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.872870683670044, "num_tokens": 203515617.0, "step": 5358 }, { "epoch": 0.6817198829665437, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6957216262817383, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8546863794326782, "num_tokens": 203548636.0, "step": 5359 }, { "epoch": 0.6818470932451343, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4392495155334473, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8679304122924805, "num_tokens": 203581358.0, "step": 5360 }, { "epoch": 0.6819743035237247, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6868202686309814, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8773511648178101, "num_tokens": 203615052.0, "step": 5361 }, { "epoch": 0.6821015138023152, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4346487522125244, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8549474477767944, "num_tokens": 203655743.0, "step": 5362 }, { "epoch": 0.6822287240809057, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6994574069976807, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8480270504951477, "num_tokens": 203689370.0, "step": 5363 }, { "epoch": 0.6823559343594963, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4482219219207764, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8561661243438721, "num_tokens": 203723594.0, "step": 5364 }, { "epoch": 0.6824831446380868, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.675983190536499, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.8442447185516357, "num_tokens": 203761746.0, "step": 5365 }, { "epoch": 0.6826103549166773, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.453329086303711, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8596635460853577, "num_tokens": 203801904.0, "step": 5366 }, { "epoch": 0.6827375651952677, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6657516956329346, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8621355295181274, "num_tokens": 203837775.0, "step": 5367 }, { "epoch": 0.6828647754738583, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.437434673309326, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8788586854934692, "num_tokens": 203876456.0, "step": 5368 }, { "epoch": 0.6829919857524488, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6627907752990723, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.857223391532898, "num_tokens": 203917946.0, "step": 5369 }, { "epoch": 0.6831191960310393, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.435486078262329, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8639556169509888, "num_tokens": 203959544.0, "step": 5370 }, { "epoch": 0.6832464063096299, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6648149490356445, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8505938053131104, "num_tokens": 204000882.0, "step": 5371 }, { "epoch": 0.6833736165882204, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4519007205963135, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8657275438308716, "num_tokens": 204037715.0, "step": 5372 }, { "epoch": 0.6835008268668108, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6530795097351074, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8561000227928162, "num_tokens": 204073145.0, "step": 5373 }, { "epoch": 0.6836280371454013, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.440812349319458, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8542927503585815, "num_tokens": 204112781.0, "step": 5374 }, { "epoch": 0.6837552474239919, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6437275409698486, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8582894206047058, "num_tokens": 204151128.0, "step": 5375 }, { "epoch": 0.6838824577025824, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4435064792633057, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8671205639839172, "num_tokens": 204188666.0, "step": 5376 }, { "epoch": 0.6840096679811729, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.629958391189575, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8808543086051941, "num_tokens": 204225410.0, "step": 5377 }, { "epoch": 0.6841368782597634, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4415109157562256, "learning_rate": 1e-05, "loss": 0.5329, "mean_token_accuracy": 0.843654215335846, "num_tokens": 204271506.0, "step": 5378 }, { "epoch": 0.6842640885383539, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.628089427947998, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8572999835014343, "num_tokens": 204310569.0, "step": 5379 }, { "epoch": 0.6843912988169444, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4402964115142822, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8633451461791992, "num_tokens": 204349614.0, "step": 5380 }, { "epoch": 0.6845185090955349, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.646157741546631, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8462334871292114, "num_tokens": 204381985.0, "step": 5381 }, { "epoch": 0.6846457193741254, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.438556671142578, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8546261191368103, "num_tokens": 204418433.0, "step": 5382 }, { "epoch": 0.684772929652716, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6342341899871826, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8592181205749512, "num_tokens": 204457544.0, "step": 5383 }, { "epoch": 0.6849001399313065, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4221367835998535, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8611591458320618, "num_tokens": 204500626.0, "step": 5384 }, { "epoch": 0.6850273502098969, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6430952548980713, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.847123384475708, "num_tokens": 204540861.0, "step": 5385 }, { "epoch": 0.6851545604884874, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4230105876922607, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8717971444129944, "num_tokens": 204581555.0, "step": 5386 }, { "epoch": 0.685281770767078, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6602623462677, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8622188568115234, "num_tokens": 204614931.0, "step": 5387 }, { "epoch": 0.6854089810456685, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.437842607498169, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8582799434661865, "num_tokens": 204649678.0, "step": 5388 }, { "epoch": 0.685536191324259, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6512815952301025, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8530099391937256, "num_tokens": 204690617.0, "step": 5389 }, { "epoch": 0.6856634016028496, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.446160316467285, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8535048961639404, "num_tokens": 204725685.0, "step": 5390 }, { "epoch": 0.68579061188144, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6451923847198486, "learning_rate": 1e-05, "loss": 0.5336, "mean_token_accuracy": 0.8390021324157715, "num_tokens": 204768265.0, "step": 5391 }, { "epoch": 0.6859178221600305, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4327168464660645, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8627487421035767, "num_tokens": 204806211.0, "step": 5392 }, { "epoch": 0.686045032438621, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6431801319122314, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8613039255142212, "num_tokens": 204844982.0, "step": 5393 }, { "epoch": 0.6861722427172116, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4305031299591064, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8734858632087708, "num_tokens": 204881676.0, "step": 5394 }, { "epoch": 0.6862994529958021, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.64874005317688, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8575260639190674, "num_tokens": 204920316.0, "step": 5395 }, { "epoch": 0.6864266632743926, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.437793731689453, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8483367562294006, "num_tokens": 204960269.0, "step": 5396 }, { "epoch": 0.686553873552983, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.654646396636963, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8597186207771301, "num_tokens": 204993238.0, "step": 5397 }, { "epoch": 0.6866810838315736, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4355058670043945, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8508925437927246, "num_tokens": 205029550.0, "step": 5398 }, { "epoch": 0.6868082941101641, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6462013721466064, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8528172373771667, "num_tokens": 205073198.0, "step": 5399 }, { "epoch": 0.6869355043887546, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.441709518432617, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.864094614982605, "num_tokens": 205107581.0, "step": 5400 }, { "epoch": 0.6870627146673451, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.639939069747925, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8583061695098877, "num_tokens": 205144030.0, "step": 5401 }, { "epoch": 0.6871899249459357, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4438743591308594, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8563772439956665, "num_tokens": 205180583.0, "step": 5402 }, { "epoch": 0.6873171352245261, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6418874263763428, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.860689640045166, "num_tokens": 205215889.0, "step": 5403 }, { "epoch": 0.6874443455031166, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.000141143798828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.437589406967163, "learning_rate": 1e-05, "loss": 0.5162, "mean_token_accuracy": 0.841446042060852, "num_tokens": 205254725.0, "step": 5404 }, { "epoch": 0.6875715557817071, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.629873037338257, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8684532046318054, "num_tokens": 205289533.0, "step": 5405 }, { "epoch": 0.6876987660602977, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.443974494934082, "learning_rate": 1e-05, "loss": 0.5471, "mean_token_accuracy": 0.8381098508834839, "num_tokens": 205330462.0, "step": 5406 }, { "epoch": 0.6878259763388882, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6345162391662598, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8453917503356934, "num_tokens": 205371462.0, "step": 5407 }, { "epoch": 0.6879531866174787, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.430570363998413, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8805643916130066, "num_tokens": 205404629.0, "step": 5408 }, { "epoch": 0.6880803968960693, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6418569087982178, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8470746874809265, "num_tokens": 205440600.0, "step": 5409 }, { "epoch": 0.6882076071746597, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4344375133514404, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8623363375663757, "num_tokens": 205473696.0, "step": 5410 }, { "epoch": 0.6883348174532502, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6292226314544678, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8489538431167603, "num_tokens": 205519889.0, "step": 5411 }, { "epoch": 0.6884620277318407, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4411399364471436, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8740426898002625, "num_tokens": 205552303.0, "step": 5412 }, { "epoch": 0.6885892380104313, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6358237266540527, "learning_rate": 1e-05, "loss": 0.5067, "mean_token_accuracy": 0.845070481300354, "num_tokens": 205588918.0, "step": 5413 }, { "epoch": 0.6887164482890218, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4483938217163086, "learning_rate": 1e-05, "loss": 0.4878, "mean_token_accuracy": 0.8592371344566345, "num_tokens": 205624634.0, "step": 5414 }, { "epoch": 0.6888436585676123, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.610751152038574, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.864838182926178, "num_tokens": 205668700.0, "step": 5415 }, { "epoch": 0.6889708688462027, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.432889461517334, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8550654649734497, "num_tokens": 205711051.0, "step": 5416 }, { "epoch": 0.6890980791247933, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6194796562194824, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8637979030609131, "num_tokens": 205750349.0, "step": 5417 }, { "epoch": 0.6892252894033838, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.423194408416748, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8717265129089355, "num_tokens": 205789525.0, "step": 5418 }, { "epoch": 0.6893524996819743, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6242029666900635, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8588755130767822, "num_tokens": 205827266.0, "step": 5419 }, { "epoch": 0.6894797099605648, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4305431842803955, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8656360507011414, "num_tokens": 205866170.0, "step": 5420 }, { "epoch": 0.6896069202391554, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6291463375091553, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8670858144760132, "num_tokens": 205905613.0, "step": 5421 }, { "epoch": 0.6897341305177458, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4471628665924072, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8385261297225952, "num_tokens": 205943142.0, "step": 5422 }, { "epoch": 0.6898613407963363, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6234676837921143, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8590683937072754, "num_tokens": 205983321.0, "step": 5423 }, { "epoch": 0.6899885510749268, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4390671253204346, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8481830358505249, "num_tokens": 206019013.0, "step": 5424 }, { "epoch": 0.6901157613535174, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.617861032485962, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8584698438644409, "num_tokens": 206059958.0, "step": 5425 }, { "epoch": 0.6902429716321079, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4412102699279785, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8627195358276367, "num_tokens": 206096536.0, "step": 5426 }, { "epoch": 0.6903701819106984, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.615525722503662, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8536624312400818, "num_tokens": 206138399.0, "step": 5427 }, { "epoch": 0.6904973921892888, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.428882360458374, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8730317950248718, "num_tokens": 206171666.0, "step": 5428 }, { "epoch": 0.6906246024678794, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6219778060913086, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8657944202423096, "num_tokens": 206212439.0, "step": 5429 }, { "epoch": 0.6907518127464699, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.435274124145508, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8596217036247253, "num_tokens": 206245213.0, "step": 5430 }, { "epoch": 0.6908790230250604, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.617276430130005, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8606382608413696, "num_tokens": 206283875.0, "step": 5431 }, { "epoch": 0.691006233303651, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.436947822570801, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8537707328796387, "num_tokens": 206319442.0, "step": 5432 }, { "epoch": 0.6911334435822415, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6295361518859863, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8435889482498169, "num_tokens": 206358210.0, "step": 5433 }, { "epoch": 0.6912606538608319, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4317455291748047, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8716800212860107, "num_tokens": 206390900.0, "step": 5434 }, { "epoch": 0.6913878641394224, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6389272212982178, "learning_rate": 1e-05, "loss": 0.5294, "mean_token_accuracy": 0.8416670560836792, "num_tokens": 206426296.0, "step": 5435 }, { "epoch": 0.691515074418013, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4398534297943115, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.859014093875885, "num_tokens": 206457010.0, "step": 5436 }, { "epoch": 0.6916422846966035, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014209747314453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6259608268737793, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8657212257385254, "num_tokens": 206496324.0, "step": 5437 }, { "epoch": 0.691769494975194, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.436450719833374, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8532056212425232, "num_tokens": 206530507.0, "step": 5438 }, { "epoch": 0.6918967052537845, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.634657859802246, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8647143244743347, "num_tokens": 206565123.0, "step": 5439 }, { "epoch": 0.692023915532375, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4266738891601562, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8634604811668396, "num_tokens": 206602026.0, "step": 5440 }, { "epoch": 0.6921511258109655, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.63854718208313, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8553491830825806, "num_tokens": 206643371.0, "step": 5441 }, { "epoch": 0.692278336089556, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.429701089859009, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8712390661239624, "num_tokens": 206675822.0, "step": 5442 }, { "epoch": 0.6924055463681466, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6479976177215576, "learning_rate": 1e-05, "loss": 0.3574, "mean_token_accuracy": 0.8950448036193848, "num_tokens": 206703973.0, "step": 5443 }, { "epoch": 0.6925327566467371, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4301772117614746, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8540720343589783, "num_tokens": 206740645.0, "step": 5444 }, { "epoch": 0.6926599669253276, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6350672245025635, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.869017481803894, "num_tokens": 206785620.0, "step": 5445 }, { "epoch": 0.692787177203918, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.442617177963257, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8523338437080383, "num_tokens": 206818761.0, "step": 5446 }, { "epoch": 0.6929143874825086, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6458332538604736, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8687170743942261, "num_tokens": 206857523.0, "step": 5447 }, { "epoch": 0.6930415977610991, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4371707439422607, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8522161245346069, "num_tokens": 206893389.0, "step": 5448 }, { "epoch": 0.6931688080396896, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6474153995513916, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8540725708007812, "num_tokens": 206924997.0, "step": 5449 }, { "epoch": 0.6932960183182801, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4486889839172363, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8563399910926819, "num_tokens": 206957032.0, "step": 5450 }, { "epoch": 0.6934232285968707, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.631965398788452, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8625515699386597, "num_tokens": 206995407.0, "step": 5451 }, { "epoch": 0.6935504388754611, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4371728897094727, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8647331595420837, "num_tokens": 207031563.0, "step": 5452 }, { "epoch": 0.6936776491540516, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6420679092407227, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8636873960494995, "num_tokens": 207064641.0, "step": 5453 }, { "epoch": 0.6938048594326421, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4442272186279297, "learning_rate": 1e-05, "loss": 0.5351, "mean_token_accuracy": 0.8420356512069702, "num_tokens": 207099474.0, "step": 5454 }, { "epoch": 0.6939320697112327, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.628709316253662, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8707820773124695, "num_tokens": 207139250.0, "step": 5455 }, { "epoch": 0.6940592799898232, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4332945346832275, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8602912425994873, "num_tokens": 207179523.0, "step": 5456 }, { "epoch": 0.6941864902684137, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6338703632354736, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8661321401596069, "num_tokens": 207218143.0, "step": 5457 }, { "epoch": 0.6943137005470043, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4274003505706787, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8602976202964783, "num_tokens": 207256946.0, "step": 5458 }, { "epoch": 0.6944409108255947, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.634695529937744, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8731911778450012, "num_tokens": 207296793.0, "step": 5459 }, { "epoch": 0.6945681211041852, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.429873466491699, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8489304184913635, "num_tokens": 207333602.0, "step": 5460 }, { "epoch": 0.6946953313827757, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.644449234008789, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8491251468658447, "num_tokens": 207375806.0, "step": 5461 }, { "epoch": 0.6948225416613663, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.428553342819214, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.880347728729248, "num_tokens": 207407856.0, "step": 5462 }, { "epoch": 0.6949497519399568, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.651956081390381, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8456225395202637, "num_tokens": 207447000.0, "step": 5463 }, { "epoch": 0.6950769622185473, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.430130958557129, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8693768978118896, "num_tokens": 207479860.0, "step": 5464 }, { "epoch": 0.6952041724971377, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6462783813476562, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8729978799819946, "num_tokens": 207515454.0, "step": 5465 }, { "epoch": 0.6953313827757283, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4164490699768066, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8552750945091248, "num_tokens": 207559259.0, "step": 5466 }, { "epoch": 0.6954585930543188, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6740565299987793, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.850791871547699, "num_tokens": 207595909.0, "step": 5467 }, { "epoch": 0.6955858033329093, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4296998977661133, "learning_rate": 1e-05, "loss": 0.5624, "mean_token_accuracy": 0.8335068225860596, "num_tokens": 207636063.0, "step": 5468 }, { "epoch": 0.6957130136114998, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.662170886993408, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8632252216339111, "num_tokens": 207675990.0, "step": 5469 }, { "epoch": 0.6958402238900904, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4401562213897705, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8631911873817444, "num_tokens": 207708996.0, "step": 5470 }, { "epoch": 0.6959674341686808, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.690200090408325, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8432102203369141, "num_tokens": 207750433.0, "step": 5471 }, { "epoch": 0.6960946444472713, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4216935634613037, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8541192412376404, "num_tokens": 207792479.0, "step": 5472 }, { "epoch": 0.6962218547258618, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6722733974456787, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.860380232334137, "num_tokens": 207833420.0, "step": 5473 }, { "epoch": 0.6963490650044524, "ewc_loss": 0.0498046875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4192304611206055, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8717002868652344, "num_tokens": 207875933.0, "step": 5474 }, { "epoch": 0.6964762752830429, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.689075231552124, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8572109341621399, "num_tokens": 207915481.0, "step": 5475 }, { "epoch": 0.6966034855616334, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4372715950012207, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8570346236228943, "num_tokens": 207952024.0, "step": 5476 }, { "epoch": 0.6967306958402238, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6911871433258057, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8511257767677307, "num_tokens": 207990938.0, "step": 5477 }, { "epoch": 0.6968579061188144, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4318809509277344, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8587574362754822, "num_tokens": 208032492.0, "step": 5478 }, { "epoch": 0.6969851163974049, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7120437622070312, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.849186360836029, "num_tokens": 208068817.0, "step": 5479 }, { "epoch": 0.6971123266759954, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4505488872528076, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8637160062789917, "num_tokens": 208105090.0, "step": 5480 }, { "epoch": 0.697239536954586, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.7095394134521484, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.857501208782196, "num_tokens": 208136653.0, "step": 5481 }, { "epoch": 0.6973667472331765, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4499642848968506, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8608651161193848, "num_tokens": 208175556.0, "step": 5482 }, { "epoch": 0.6974939575117669, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.692924737930298, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8673542737960815, "num_tokens": 208214549.0, "step": 5483 }, { "epoch": 0.6976211677903574, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.471447467803955, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8759928941726685, "num_tokens": 208242582.0, "step": 5484 }, { "epoch": 0.697748378068948, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.670718193054199, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8540002107620239, "num_tokens": 208281639.0, "step": 5485 }, { "epoch": 0.6978755883475385, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.445685625076294, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.875843346118927, "num_tokens": 208323543.0, "step": 5486 }, { "epoch": 0.698002798626129, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6710093021392822, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.859246015548706, "num_tokens": 208359881.0, "step": 5487 }, { "epoch": 0.6981300089047195, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4469680786132812, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8493934869766235, "num_tokens": 208404541.0, "step": 5488 }, { "epoch": 0.69825721918331, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001430511474609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6698179244995117, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8645792007446289, "num_tokens": 208437731.0, "step": 5489 }, { "epoch": 0.6983844294619005, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.462888479232788, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8593299388885498, "num_tokens": 208475171.0, "step": 5490 }, { "epoch": 0.698511639740491, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6583733558654785, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8598059415817261, "num_tokens": 208511410.0, "step": 5491 }, { "epoch": 0.6986388500190815, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4474027156829834, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8635112047195435, "num_tokens": 208551133.0, "step": 5492 }, { "epoch": 0.6987660602976721, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.644535779953003, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8641253709793091, "num_tokens": 208593394.0, "step": 5493 }, { "epoch": 0.6988932705762626, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.466756582260132, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.850148618221283, "num_tokens": 208627160.0, "step": 5494 }, { "epoch": 0.699020480854853, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.621807813644409, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8586167097091675, "num_tokens": 208664896.0, "step": 5495 }, { "epoch": 0.6991476911334435, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4648821353912354, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8542576432228088, "num_tokens": 208702118.0, "step": 5496 }, { "epoch": 0.6992749014120341, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.632539749145508, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8516391515731812, "num_tokens": 208741243.0, "step": 5497 }, { "epoch": 0.6994021116906246, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.440023899078369, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.868774950504303, "num_tokens": 208778914.0, "step": 5498 }, { "epoch": 0.6995293219692151, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6208655834198, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8584740161895752, "num_tokens": 208815683.0, "step": 5499 }, { "epoch": 0.6996565322478057, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4351603984832764, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8602069616317749, "num_tokens": 208853812.0, "step": 5500 }, { "epoch": 0.6997837425263961, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.629699468612671, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8698532581329346, "num_tokens": 208894326.0, "step": 5501 }, { "epoch": 0.6999109528049866, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4366977214813232, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8674457669258118, "num_tokens": 208929644.0, "step": 5502 }, { "epoch": 0.7000381630835771, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.625861167907715, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8569531440734863, "num_tokens": 208972836.0, "step": 5503 }, { "epoch": 0.7001653733621677, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4247546195983887, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8600488305091858, "num_tokens": 209016255.0, "step": 5504 }, { "epoch": 0.7002925836407582, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.643092393875122, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8475387096405029, "num_tokens": 209054576.0, "step": 5505 }, { "epoch": 0.7004197939193487, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.407433271408081, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8687134981155396, "num_tokens": 209097881.0, "step": 5506 }, { "epoch": 0.7005470041979391, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.663992404937744, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8502273559570312, "num_tokens": 209137354.0, "step": 5507 }, { "epoch": 0.7006742144765297, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.432708740234375, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8561650514602661, "num_tokens": 209170505.0, "step": 5508 }, { "epoch": 0.7008014247551202, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6758415699005127, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.852438747882843, "num_tokens": 209207745.0, "step": 5509 }, { "epoch": 0.7009286350337107, "ewc_loss": 0.050048828125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4266111850738525, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8541762828826904, "num_tokens": 209246674.0, "step": 5510 }, { "epoch": 0.7010558453123013, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00014400482177734375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6679024696350098, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8620491027832031, "num_tokens": 209289019.0, "step": 5511 }, { "epoch": 0.7011830555908918, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4438769817352295, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8590775728225708, "num_tokens": 209323053.0, "step": 5512 }, { "epoch": 0.7013102658694823, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.664191961288452, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8600325584411621, "num_tokens": 209358810.0, "step": 5513 }, { "epoch": 0.7014374761480727, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4347825050354004, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8625298142433167, "num_tokens": 209401868.0, "step": 5514 }, { "epoch": 0.7015646864266633, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6729555130004883, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8482165932655334, "num_tokens": 209435212.0, "step": 5515 }, { "epoch": 0.7016918967052538, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4329543113708496, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8671764135360718, "num_tokens": 209472357.0, "step": 5516 }, { "epoch": 0.7018191069838443, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.672445058822632, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8447795510292053, "num_tokens": 209512687.0, "step": 5517 }, { "epoch": 0.7019463172624348, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4479668140411377, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8643614053726196, "num_tokens": 209547223.0, "step": 5518 }, { "epoch": 0.7020735275410254, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6695547103881836, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8617727756500244, "num_tokens": 209583299.0, "step": 5519 }, { "epoch": 0.7022007378196158, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.44987416267395, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8605765104293823, "num_tokens": 209618792.0, "step": 5520 }, { "epoch": 0.7023279480982063, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6660804748535156, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.855150580406189, "num_tokens": 209654324.0, "step": 5521 }, { "epoch": 0.7024551583767968, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.461812973022461, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8726323843002319, "num_tokens": 209684963.0, "step": 5522 }, { "epoch": 0.7025823686553874, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 3.013810157775879, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8746880292892456, "num_tokens": 209724197.0, "step": 5523 }, { "epoch": 0.7027095789339779, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.447239398956299, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8670795559883118, "num_tokens": 209759397.0, "step": 5524 }, { "epoch": 0.7028367892125684, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.780890703201294, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8623174428939819, "num_tokens": 209802504.0, "step": 5525 }, { "epoch": 0.7029639994911588, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4543495178222656, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8637329339981079, "num_tokens": 209838798.0, "step": 5526 }, { "epoch": 0.7030912097697494, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.790670871734619, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8506542444229126, "num_tokens": 209880935.0, "step": 5527 }, { "epoch": 0.7032184200483399, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.484221935272217, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8544249534606934, "num_tokens": 209921514.0, "step": 5528 }, { "epoch": 0.7033456303269304, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.749251365661621, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.868067741394043, "num_tokens": 209960082.0, "step": 5529 }, { "epoch": 0.703472840605521, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5104877948760986, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8612306118011475, "num_tokens": 209996338.0, "step": 5530 }, { "epoch": 0.7036000508841115, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7075798511505127, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.85040283203125, "num_tokens": 210039140.0, "step": 5531 }, { "epoch": 0.7037272611627019, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5142581462860107, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8638830184936523, "num_tokens": 210074166.0, "step": 5532 }, { "epoch": 0.7038544714412924, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.667692184448242, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8723455667495728, "num_tokens": 210110295.0, "step": 5533 }, { "epoch": 0.703981681719883, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4995553493499756, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8516237139701843, "num_tokens": 210151994.0, "step": 5534 }, { "epoch": 0.7041088919984735, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.644280195236206, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8634632229804993, "num_tokens": 210184769.0, "step": 5535 }, { "epoch": 0.704236102277064, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.486429452896118, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8556391596794128, "num_tokens": 210225561.0, "step": 5536 }, { "epoch": 0.7043633125556545, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6085643768310547, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8738511204719543, "num_tokens": 210265544.0, "step": 5537 }, { "epoch": 0.704490522834245, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4800727367401123, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.872705340385437, "num_tokens": 210298178.0, "step": 5538 }, { "epoch": 0.7046177331128355, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.596475839614868, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8528627157211304, "num_tokens": 210342391.0, "step": 5539 }, { "epoch": 0.704744943391426, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4606692790985107, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8656752109527588, "num_tokens": 210386673.0, "step": 5540 }, { "epoch": 0.7048721536700165, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5946953296661377, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8523275256156921, "num_tokens": 210424050.0, "step": 5541 }, { "epoch": 0.7049993639486071, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4553496837615967, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8684920072555542, "num_tokens": 210462593.0, "step": 5542 }, { "epoch": 0.7051265742271976, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6007118225097656, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8611404895782471, "num_tokens": 210496534.0, "step": 5543 }, { "epoch": 0.705253784505788, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4468343257904053, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8531379699707031, "num_tokens": 210534945.0, "step": 5544 }, { "epoch": 0.7053809947843785, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.594074010848999, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8705840706825256, "num_tokens": 210576742.0, "step": 5545 }, { "epoch": 0.7055082050629691, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4446523189544678, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8423280715942383, "num_tokens": 210616248.0, "step": 5546 }, { "epoch": 0.7056354153415596, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6026859283447266, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8543769121170044, "num_tokens": 210658229.0, "step": 5547 }, { "epoch": 0.7057626256201501, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4258267879486084, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.872259259223938, "num_tokens": 210696312.0, "step": 5548 }, { "epoch": 0.7058898358987407, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.616917133331299, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8553621768951416, "num_tokens": 210736916.0, "step": 5549 }, { "epoch": 0.7060170461773311, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.432892322540283, "learning_rate": 1e-05, "loss": 0.5369, "mean_token_accuracy": 0.8412039279937744, "num_tokens": 210777779.0, "step": 5550 }, { "epoch": 0.7061442564559216, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6204020977020264, "learning_rate": 1e-05, "loss": 0.5325, "mean_token_accuracy": 0.8405166864395142, "num_tokens": 210823134.0, "step": 5551 }, { "epoch": 0.7062714667345121, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.426220417022705, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8564658164978027, "num_tokens": 210862792.0, "step": 5552 }, { "epoch": 0.7063986770131027, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.615558624267578, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.866005539894104, "num_tokens": 210905411.0, "step": 5553 }, { "epoch": 0.7065258872916932, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4305124282836914, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8511447906494141, "num_tokens": 210944090.0, "step": 5554 }, { "epoch": 0.7066530975702837, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.637300491333008, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.857673168182373, "num_tokens": 210977372.0, "step": 5555 }, { "epoch": 0.7067803078488741, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4141011238098145, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.8506419062614441, "num_tokens": 211021800.0, "step": 5556 }, { "epoch": 0.7069075181274647, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.648606300354004, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8555569648742676, "num_tokens": 211063568.0, "step": 5557 }, { "epoch": 0.7070347284060552, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4161555767059326, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8622047305107117, "num_tokens": 211100422.0, "step": 5558 }, { "epoch": 0.7071619386846457, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6718740463256836, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8514713048934937, "num_tokens": 211138644.0, "step": 5559 }, { "epoch": 0.7072891489632362, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4237453937530518, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8566471338272095, "num_tokens": 211173033.0, "step": 5560 }, { "epoch": 0.7074163592418268, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.667032480239868, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8676403760910034, "num_tokens": 211212902.0, "step": 5561 }, { "epoch": 0.7075435695204173, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4181973934173584, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8380029201507568, "num_tokens": 211258812.0, "step": 5562 }, { "epoch": 0.7076707797990077, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6848063468933105, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8541136384010315, "num_tokens": 211296496.0, "step": 5563 }, { "epoch": 0.7077979900775982, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.419797658920288, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8561502695083618, "num_tokens": 211337081.0, "step": 5564 }, { "epoch": 0.7079252003561888, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7106029987335205, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8487670421600342, "num_tokens": 211373845.0, "step": 5565 }, { "epoch": 0.7080524106347793, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4278335571289062, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8571910262107849, "num_tokens": 211412994.0, "step": 5566 }, { "epoch": 0.7081796209133698, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.699152946472168, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8506379127502441, "num_tokens": 211456678.0, "step": 5567 }, { "epoch": 0.7083068311919604, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4372522830963135, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8526222109794617, "num_tokens": 211497040.0, "step": 5568 }, { "epoch": 0.7084340414705508, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.7076871395111084, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8485608100891113, "num_tokens": 211534113.0, "step": 5569 }, { "epoch": 0.7085612517491413, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.440854549407959, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8649753928184509, "num_tokens": 211569518.0, "step": 5570 }, { "epoch": 0.7086884620277318, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.699594497680664, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8706138134002686, "num_tokens": 211607918.0, "step": 5571 }, { "epoch": 0.7088156723063224, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.457914113998413, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8482994437217712, "num_tokens": 211647222.0, "step": 5572 }, { "epoch": 0.7089428825849129, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.680399179458618, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8670070171356201, "num_tokens": 211684584.0, "step": 5573 }, { "epoch": 0.7090700928635034, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.473006010055542, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8629971742630005, "num_tokens": 211718968.0, "step": 5574 }, { "epoch": 0.7091973031420938, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.657499313354492, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8619717955589294, "num_tokens": 211756230.0, "step": 5575 }, { "epoch": 0.7093245134206844, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.448058843612671, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8773661255836487, "num_tokens": 211794155.0, "step": 5576 }, { "epoch": 0.7094517236992749, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6481378078460693, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8583104610443115, "num_tokens": 211834483.0, "step": 5577 }, { "epoch": 0.7095789339778654, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5196051597595215, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8491024971008301, "num_tokens": 211873369.0, "step": 5578 }, { "epoch": 0.709706144256456, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.610642910003662, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8499984741210938, "num_tokens": 211911985.0, "step": 5579 }, { "epoch": 0.7098333545350465, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.460120677947998, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8631576895713806, "num_tokens": 211953739.0, "step": 5580 }, { "epoch": 0.7099605648136369, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.596489191055298, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.869053840637207, "num_tokens": 211988817.0, "step": 5581 }, { "epoch": 0.7100877750922274, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.461186647415161, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8484539985656738, "num_tokens": 212030891.0, "step": 5582 }, { "epoch": 0.710214985370818, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5937108993530273, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8719810247421265, "num_tokens": 212065727.0, "step": 5583 }, { "epoch": 0.7103421956494085, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.467930793762207, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8576774001121521, "num_tokens": 212098718.0, "step": 5584 }, { "epoch": 0.710469405927999, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5735204219818115, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8572408556938171, "num_tokens": 212142597.0, "step": 5585 }, { "epoch": 0.7105966162065895, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.461043357849121, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8364041447639465, "num_tokens": 212182878.0, "step": 5586 }, { "epoch": 0.71072382648518, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5788307189941406, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8660848736763, "num_tokens": 212216759.0, "step": 5587 }, { "epoch": 0.7108510367637705, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.465257406234741, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8520575761795044, "num_tokens": 212253559.0, "step": 5588 }, { "epoch": 0.710978247042361, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5620946884155273, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8626239895820618, "num_tokens": 212296821.0, "step": 5589 }, { "epoch": 0.7111054573209515, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.448004961013794, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.865540623664856, "num_tokens": 212334872.0, "step": 5590 }, { "epoch": 0.7112326675995421, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.580005407333374, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8625539541244507, "num_tokens": 212375692.0, "step": 5591 }, { "epoch": 0.7113598778781326, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4661805629730225, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8513160943984985, "num_tokens": 212407301.0, "step": 5592 }, { "epoch": 0.711487088156723, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5661134719848633, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8613080978393555, "num_tokens": 212443831.0, "step": 5593 }, { "epoch": 0.7116142984353135, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4497640132904053, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8679355382919312, "num_tokens": 212477595.0, "step": 5594 }, { "epoch": 0.7117415087139041, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.587651014328003, "learning_rate": 1e-05, "loss": 0.5311, "mean_token_accuracy": 0.841740608215332, "num_tokens": 212513786.0, "step": 5595 }, { "epoch": 0.7118687189924946, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.445988178253174, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8564362525939941, "num_tokens": 212548111.0, "step": 5596 }, { "epoch": 0.7119959292710851, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5939602851867676, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8536840677261353, "num_tokens": 212585395.0, "step": 5597 }, { "epoch": 0.7121231395496757, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4242520332336426, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8701040744781494, "num_tokens": 212625351.0, "step": 5598 }, { "epoch": 0.7122503498282661, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014495849609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6234681606292725, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8523836135864258, "num_tokens": 212657689.0, "step": 5599 }, { "epoch": 0.7123775601068566, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4051408767700195, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.873879075050354, "num_tokens": 212694741.0, "step": 5600 }, { "epoch": 0.7125047703854471, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6575353145599365, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8549183011054993, "num_tokens": 212727109.0, "step": 5601 }, { "epoch": 0.7126319806640377, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4081223011016846, "learning_rate": 1e-05, "loss": 0.5064, "mean_token_accuracy": 0.8480498790740967, "num_tokens": 212763031.0, "step": 5602 }, { "epoch": 0.7127591909426282, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6682136058807373, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8618994951248169, "num_tokens": 212803601.0, "step": 5603 }, { "epoch": 0.7128864012212187, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.3911867141723633, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8741686940193176, "num_tokens": 212848331.0, "step": 5604 }, { "epoch": 0.7130136114998091, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6959218978881836, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8570003509521484, "num_tokens": 212890364.0, "step": 5605 }, { "epoch": 0.7131408217783997, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.395116090774536, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8637194633483887, "num_tokens": 212931234.0, "step": 5606 }, { "epoch": 0.7132680320569902, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.725999355316162, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.860992968082428, "num_tokens": 212970103.0, "step": 5607 }, { "epoch": 0.7133952423355807, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.424041509628296, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.853378415107727, "num_tokens": 213004895.0, "step": 5608 }, { "epoch": 0.7135224526141712, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.736880302429199, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8622408509254456, "num_tokens": 213044263.0, "step": 5609 }, { "epoch": 0.7136496628927618, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.435666084289551, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.855422854423523, "num_tokens": 213081685.0, "step": 5610 }, { "epoch": 0.7137768731713523, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.742689609527588, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8626831769943237, "num_tokens": 213116986.0, "step": 5611 }, { "epoch": 0.7139040834499427, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4378647804260254, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8680388927459717, "num_tokens": 213156313.0, "step": 5612 }, { "epoch": 0.7140312937285332, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.761369228363037, "learning_rate": 1e-05, "loss": 0.5159, "mean_token_accuracy": 0.8484488725662231, "num_tokens": 213187216.0, "step": 5613 }, { "epoch": 0.7141585040071238, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.468691349029541, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8544602394104004, "num_tokens": 213223380.0, "step": 5614 }, { "epoch": 0.7142857142857143, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7305235862731934, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8737661838531494, "num_tokens": 213259100.0, "step": 5615 }, { "epoch": 0.7144129245643048, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4706180095672607, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8497370481491089, "num_tokens": 213300103.0, "step": 5616 }, { "epoch": 0.7145401348428954, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7092607021331787, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8688325881958008, "num_tokens": 213338283.0, "step": 5617 }, { "epoch": 0.7146673451214858, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.46502685546875, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.865609884262085, "num_tokens": 213382211.0, "step": 5618 }, { "epoch": 0.7147945554000763, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.7058181762695312, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8620771765708923, "num_tokens": 213418790.0, "step": 5619 }, { "epoch": 0.7149217656786668, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4691250324249268, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8600063323974609, "num_tokens": 213457843.0, "step": 5620 }, { "epoch": 0.7150489759572574, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6927125453948975, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8596092462539673, "num_tokens": 213496182.0, "step": 5621 }, { "epoch": 0.7151761862358479, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.460881233215332, "learning_rate": 1e-05, "loss": 0.3964, "mean_token_accuracy": 0.8762888312339783, "num_tokens": 213535254.0, "step": 5622 }, { "epoch": 0.7153033965144384, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.679579496383667, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8646520376205444, "num_tokens": 213574331.0, "step": 5623 }, { "epoch": 0.7154306067930288, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4724955558776855, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8588871955871582, "num_tokens": 213608185.0, "step": 5624 }, { "epoch": 0.7155578170716194, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6670215129852295, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8666973114013672, "num_tokens": 213642032.0, "step": 5625 }, { "epoch": 0.7156850273502099, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4757843017578125, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8650885224342346, "num_tokens": 213674652.0, "step": 5626 }, { "epoch": 0.7158122376288004, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.647383451461792, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8667687177658081, "num_tokens": 213711112.0, "step": 5627 }, { "epoch": 0.715939447907391, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.464902400970459, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8508634567260742, "num_tokens": 213750319.0, "step": 5628 }, { "epoch": 0.7160666581859815, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.632817506790161, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8664733171463013, "num_tokens": 213788408.0, "step": 5629 }, { "epoch": 0.7161938684645719, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.475666046142578, "learning_rate": 1e-05, "loss": 0.532, "mean_token_accuracy": 0.8403632044792175, "num_tokens": 213826822.0, "step": 5630 }, { "epoch": 0.7163210787431624, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.633324384689331, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.843252956867218, "num_tokens": 213864046.0, "step": 5631 }, { "epoch": 0.716448289021753, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.449686050415039, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.864565908908844, "num_tokens": 213907235.0, "step": 5632 }, { "epoch": 0.7165754993003435, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6264021396636963, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8629758358001709, "num_tokens": 213942126.0, "step": 5633 }, { "epoch": 0.716702709578934, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4509294033050537, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8600513339042664, "num_tokens": 213980486.0, "step": 5634 }, { "epoch": 0.7168299198575245, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6188364028930664, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8706172704696655, "num_tokens": 214016991.0, "step": 5635 }, { "epoch": 0.716957130136115, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.456442356109619, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8646354675292969, "num_tokens": 214049424.0, "step": 5636 }, { "epoch": 0.7170843404147055, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.60614013671875, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8609604239463806, "num_tokens": 214090422.0, "step": 5637 }, { "epoch": 0.717211550693296, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4512550830841064, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.869367241859436, "num_tokens": 214123075.0, "step": 5638 }, { "epoch": 0.7173387609718865, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.621795177459717, "learning_rate": 1e-05, "loss": 0.5539, "mean_token_accuracy": 0.8403680920600891, "num_tokens": 214158962.0, "step": 5639 }, { "epoch": 0.7174659712504771, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4310572147369385, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8659062385559082, "num_tokens": 214201867.0, "step": 5640 }, { "epoch": 0.7175931815290676, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.617863178253174, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8684459328651428, "num_tokens": 214241300.0, "step": 5641 }, { "epoch": 0.717720391807658, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4433374404907227, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8561999201774597, "num_tokens": 214276455.0, "step": 5642 }, { "epoch": 0.7178476020862485, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.623810052871704, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8638330698013306, "num_tokens": 214313737.0, "step": 5643 }, { "epoch": 0.7179748123648391, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.433490037918091, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8627270460128784, "num_tokens": 214351179.0, "step": 5644 }, { "epoch": 0.7181020226434296, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6256487369537354, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8741776943206787, "num_tokens": 214387606.0, "step": 5645 }, { "epoch": 0.7182292329220201, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4367618560791016, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8515515327453613, "num_tokens": 214428149.0, "step": 5646 }, { "epoch": 0.7183564432006107, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.639566421508789, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8675423860549927, "num_tokens": 214468878.0, "step": 5647 }, { "epoch": 0.7184836534792011, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.432358503341675, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8514827489852905, "num_tokens": 214508813.0, "step": 5648 }, { "epoch": 0.7186108637577916, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6323962211608887, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8620790243148804, "num_tokens": 214553896.0, "step": 5649 }, { "epoch": 0.7187380740363821, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.423260450363159, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8652781248092651, "num_tokens": 214593727.0, "step": 5650 }, { "epoch": 0.7188652843149727, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6660940647125244, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8629866242408752, "num_tokens": 214627029.0, "step": 5651 }, { "epoch": 0.7189924945935632, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4268229007720947, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8457823395729065, "num_tokens": 214665484.0, "step": 5652 }, { "epoch": 0.7191197048721537, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6599438190460205, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8688001036643982, "num_tokens": 214708654.0, "step": 5653 }, { "epoch": 0.7192469151507441, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.430072069168091, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8549277782440186, "num_tokens": 214746962.0, "step": 5654 }, { "epoch": 0.7193741254293347, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6677005290985107, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8598498106002808, "num_tokens": 214784691.0, "step": 5655 }, { "epoch": 0.7195013357079252, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4396560192108154, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8700060844421387, "num_tokens": 214817182.0, "step": 5656 }, { "epoch": 0.7196285459865157, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6752429008483887, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8456822633743286, "num_tokens": 214857053.0, "step": 5657 }, { "epoch": 0.7197557562651062, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4652011394500732, "learning_rate": 1e-05, "loss": 0.5704, "mean_token_accuracy": 0.8363045454025269, "num_tokens": 214894908.0, "step": 5658 }, { "epoch": 0.7198829665436968, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.647522211074829, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8577411770820618, "num_tokens": 214930235.0, "step": 5659 }, { "epoch": 0.7200101768222873, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.437727212905884, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.868848979473114, "num_tokens": 214967140.0, "step": 5660 }, { "epoch": 0.7201373871008777, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6513540744781494, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8596169948577881, "num_tokens": 215004039.0, "step": 5661 }, { "epoch": 0.7202645973794682, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4361965656280518, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8690980076789856, "num_tokens": 215039305.0, "step": 5662 }, { "epoch": 0.7203918076580588, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014591217041015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.65132737159729, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8695690631866455, "num_tokens": 215076455.0, "step": 5663 }, { "epoch": 0.7205190179366493, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.44338321685791, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8511500358581543, "num_tokens": 215116826.0, "step": 5664 }, { "epoch": 0.7206462282152398, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.645482063293457, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8508157730102539, "num_tokens": 215162302.0, "step": 5665 }, { "epoch": 0.7207734384938304, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4647605419158936, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8546209335327148, "num_tokens": 215192402.0, "step": 5666 }, { "epoch": 0.7209006487724208, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.646246910095215, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8544873595237732, "num_tokens": 215228521.0, "step": 5667 }, { "epoch": 0.7210278590510113, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.447617769241333, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8601967096328735, "num_tokens": 215264997.0, "step": 5668 }, { "epoch": 0.7211550693296018, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6272809505462646, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8810649514198303, "num_tokens": 215301997.0, "step": 5669 }, { "epoch": 0.7212822796081924, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4422781467437744, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8526449203491211, "num_tokens": 215343637.0, "step": 5670 }, { "epoch": 0.7214094898867829, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6415419578552246, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8617783784866333, "num_tokens": 215381966.0, "step": 5671 }, { "epoch": 0.7215367001653734, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4378600120544434, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8576042652130127, "num_tokens": 215422930.0, "step": 5672 }, { "epoch": 0.7216639104439638, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.635777235031128, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8720552921295166, "num_tokens": 215461204.0, "step": 5673 }, { "epoch": 0.7217911207225544, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4504055976867676, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8603814840316772, "num_tokens": 215493771.0, "step": 5674 }, { "epoch": 0.7219183310011449, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6428604125976562, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8602940440177917, "num_tokens": 215529398.0, "step": 5675 }, { "epoch": 0.7220455412797354, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4328086376190186, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8665197491645813, "num_tokens": 215565794.0, "step": 5676 }, { "epoch": 0.7221727515583259, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.651658535003662, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.85176682472229, "num_tokens": 215604005.0, "step": 5677 }, { "epoch": 0.7222999618369165, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4332478046417236, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8703866004943848, "num_tokens": 215635325.0, "step": 5678 }, { "epoch": 0.7224271721155069, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6645348072052, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8625565767288208, "num_tokens": 215667477.0, "step": 5679 }, { "epoch": 0.7225543823940974, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.430020809173584, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8513060212135315, "num_tokens": 215707087.0, "step": 5680 }, { "epoch": 0.722681592672688, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6632883548736572, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8640411496162415, "num_tokens": 215746182.0, "step": 5681 }, { "epoch": 0.7228088029512785, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.423555612564087, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8561946749687195, "num_tokens": 215785941.0, "step": 5682 }, { "epoch": 0.722936013229869, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.675361394882202, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8632453083992004, "num_tokens": 215823562.0, "step": 5683 }, { "epoch": 0.7230632235084595, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.45483136177063, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8398469090461731, "num_tokens": 215860454.0, "step": 5684 }, { "epoch": 0.72319043378705, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.682079315185547, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8674314022064209, "num_tokens": 215895088.0, "step": 5685 }, { "epoch": 0.7233176440656405, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.431140899658203, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8533797264099121, "num_tokens": 215935170.0, "step": 5686 }, { "epoch": 0.723444854344231, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6671831607818604, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8727378845214844, "num_tokens": 215977323.0, "step": 5687 }, { "epoch": 0.7235720646228215, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4301562309265137, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8718289732933044, "num_tokens": 216016729.0, "step": 5688 }, { "epoch": 0.7236992749014121, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6730544567108154, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8593182563781738, "num_tokens": 216058514.0, "step": 5689 }, { "epoch": 0.7238264851800026, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.444772481918335, "learning_rate": 1e-05, "loss": 0.5315, "mean_token_accuracy": 0.8399167060852051, "num_tokens": 216098633.0, "step": 5690 }, { "epoch": 0.723953695458593, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.665799379348755, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.8810938596725464, "num_tokens": 216136124.0, "step": 5691 }, { "epoch": 0.7240809057371835, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4424126148223877, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8569062352180481, "num_tokens": 216178799.0, "step": 5692 }, { "epoch": 0.7242081160157741, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.68472957611084, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.8450470566749573, "num_tokens": 216215233.0, "step": 5693 }, { "epoch": 0.7243353262943646, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4502382278442383, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8643353581428528, "num_tokens": 216252258.0, "step": 5694 }, { "epoch": 0.7244625365729551, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.674309492111206, "learning_rate": 1e-05, "loss": 0.5049, "mean_token_accuracy": 0.8476440906524658, "num_tokens": 216287845.0, "step": 5695 }, { "epoch": 0.7245897468515456, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4482741355895996, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8690169453620911, "num_tokens": 216325490.0, "step": 5696 }, { "epoch": 0.7247169571301361, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.673396110534668, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8530057668685913, "num_tokens": 216359353.0, "step": 5697 }, { "epoch": 0.7248441674087266, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4588332176208496, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8507201671600342, "num_tokens": 216393620.0, "step": 5698 }, { "epoch": 0.7249713776873171, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.66694974899292, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8647707104682922, "num_tokens": 216427400.0, "step": 5699 }, { "epoch": 0.7250985879659076, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4534945487976074, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.858161449432373, "num_tokens": 216463817.0, "step": 5700 }, { "epoch": 0.7252257982444982, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.660496950149536, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8585888147354126, "num_tokens": 216495596.0, "step": 5701 }, { "epoch": 0.7253530085230887, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.443878173828125, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8649386167526245, "num_tokens": 216533003.0, "step": 5702 }, { "epoch": 0.7254802188016791, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.656311273574829, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8640488386154175, "num_tokens": 216569280.0, "step": 5703 }, { "epoch": 0.7256074290802697, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4509289264678955, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8715943694114685, "num_tokens": 216603073.0, "step": 5704 }, { "epoch": 0.7257346393588602, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6538145542144775, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8561359643936157, "num_tokens": 216641971.0, "step": 5705 }, { "epoch": 0.7258618496374507, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4526901245117188, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8440333604812622, "num_tokens": 216683119.0, "step": 5706 }, { "epoch": 0.7259890599160412, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6620662212371826, "learning_rate": 1e-05, "loss": 0.541, "mean_token_accuracy": 0.8376734256744385, "num_tokens": 216722139.0, "step": 5707 }, { "epoch": 0.7261162701946318, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.452155351638794, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.870597243309021, "num_tokens": 216757712.0, "step": 5708 }, { "epoch": 0.7262434804732223, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6495845317840576, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8526381850242615, "num_tokens": 216794132.0, "step": 5709 }, { "epoch": 0.7263706907518127, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4372825622558594, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8562114238739014, "num_tokens": 216833937.0, "step": 5710 }, { "epoch": 0.7264979010304032, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6577799320220947, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8583000302314758, "num_tokens": 216872979.0, "step": 5711 }, { "epoch": 0.7266251113089938, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4262685775756836, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8700615167617798, "num_tokens": 216915369.0, "step": 5712 }, { "epoch": 0.7267523215875843, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6711602210998535, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8651703596115112, "num_tokens": 216948876.0, "step": 5713 }, { "epoch": 0.7268795318661748, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4329750537872314, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8613468408584595, "num_tokens": 216988254.0, "step": 5714 }, { "epoch": 0.7270067421447653, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6788597106933594, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8586691617965698, "num_tokens": 217024792.0, "step": 5715 }, { "epoch": 0.7271339524233558, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.440835475921631, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8591688275337219, "num_tokens": 217063228.0, "step": 5716 }, { "epoch": 0.7272611627019463, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6855928897857666, "learning_rate": 1e-05, "loss": 0.5442, "mean_token_accuracy": 0.8333128094673157, "num_tokens": 217102601.0, "step": 5717 }, { "epoch": 0.7273883729805368, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4434125423431396, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8657195568084717, "num_tokens": 217137839.0, "step": 5718 }, { "epoch": 0.7275155832591274, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6848044395446777, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8656037449836731, "num_tokens": 217179862.0, "step": 5719 }, { "epoch": 0.7276427935377179, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4595346450805664, "learning_rate": 1e-05, "loss": 0.5428, "mean_token_accuracy": 0.8368620872497559, "num_tokens": 217217728.0, "step": 5720 }, { "epoch": 0.7277700038163084, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.671504020690918, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8588401079177856, "num_tokens": 217255410.0, "step": 5721 }, { "epoch": 0.7278972140948988, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4416956901550293, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8516972064971924, "num_tokens": 217294365.0, "step": 5722 }, { "epoch": 0.7280244243734894, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6640877723693848, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8651832342147827, "num_tokens": 217333451.0, "step": 5723 }, { "epoch": 0.7281516346520799, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4508981704711914, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8558094501495361, "num_tokens": 217368737.0, "step": 5724 }, { "epoch": 0.7282788449306704, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.660396099090576, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.84273362159729, "num_tokens": 217409476.0, "step": 5725 }, { "epoch": 0.7284060552092609, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4456472396850586, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8683750629425049, "num_tokens": 217446481.0, "step": 5726 }, { "epoch": 0.7285332654878515, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.650284767150879, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8685073852539062, "num_tokens": 217482647.0, "step": 5727 }, { "epoch": 0.7286604757664419, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.453423500061035, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.856174111366272, "num_tokens": 217520783.0, "step": 5728 }, { "epoch": 0.7287876860450324, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6537058353424072, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8588611483573914, "num_tokens": 217557061.0, "step": 5729 }, { "epoch": 0.7289148963236229, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.458021879196167, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8707612752914429, "num_tokens": 217586833.0, "step": 5730 }, { "epoch": 0.7290421066022135, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6422059535980225, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8508179783821106, "num_tokens": 217622644.0, "step": 5731 }, { "epoch": 0.729169316880804, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.453585147857666, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8640865087509155, "num_tokens": 217659345.0, "step": 5732 }, { "epoch": 0.7292965271593945, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6252548694610596, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8676033616065979, "num_tokens": 217698134.0, "step": 5733 }, { "epoch": 0.7294237374379849, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4749484062194824, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8494042754173279, "num_tokens": 217730578.0, "step": 5734 }, { "epoch": 0.7295509477165755, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.606558084487915, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8758081793785095, "num_tokens": 217768090.0, "step": 5735 }, { "epoch": 0.729678157995166, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.454195976257324, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8601475954055786, "num_tokens": 217809350.0, "step": 5736 }, { "epoch": 0.7298053682737565, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6050634384155273, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8559787273406982, "num_tokens": 217852478.0, "step": 5737 }, { "epoch": 0.7299325785523471, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.458944082260132, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.867736279964447, "num_tokens": 217885463.0, "step": 5738 }, { "epoch": 0.7300597888309376, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6088523864746094, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8515133857727051, "num_tokens": 217925712.0, "step": 5739 }, { "epoch": 0.730186999109528, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4421701431274414, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8580787181854248, "num_tokens": 217969628.0, "step": 5740 }, { "epoch": 0.7303142093881185, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6114304065704346, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8618137836456299, "num_tokens": 218012027.0, "step": 5741 }, { "epoch": 0.7304414196667091, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4397079944610596, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8545448780059814, "num_tokens": 218054018.0, "step": 5742 }, { "epoch": 0.7305686299452996, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6094391345977783, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8473595976829529, "num_tokens": 218099181.0, "step": 5743 }, { "epoch": 0.7306958402238901, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4376299381256104, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8659136295318604, "num_tokens": 218135330.0, "step": 5744 }, { "epoch": 0.7308230505024806, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6347241401672363, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8438520431518555, "num_tokens": 218171785.0, "step": 5745 }, { "epoch": 0.7309502607810711, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4658617973327637, "learning_rate": 1e-05, "loss": 0.5372, "mean_token_accuracy": 0.8374070525169373, "num_tokens": 218204230.0, "step": 5746 }, { "epoch": 0.7310774710596616, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6195409297943115, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8726963400840759, "num_tokens": 218237755.0, "step": 5747 }, { "epoch": 0.7312046813382521, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.449571371078491, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8453329801559448, "num_tokens": 218271850.0, "step": 5748 }, { "epoch": 0.7313318916168426, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.607417583465576, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8643118143081665, "num_tokens": 218308466.0, "step": 5749 }, { "epoch": 0.7314591018954332, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4451487064361572, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8452016711235046, "num_tokens": 218346133.0, "step": 5750 }, { "epoch": 0.7315863121740237, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6202175617218018, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8589421510696411, "num_tokens": 218384341.0, "step": 5751 }, { "epoch": 0.7317135224526141, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.436032295227051, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8626240491867065, "num_tokens": 218418314.0, "step": 5752 }, { "epoch": 0.7318407327312046, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6361284255981445, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8512242436408997, "num_tokens": 218455258.0, "step": 5753 }, { "epoch": 0.7319679430097952, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4216227531433105, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8592405319213867, "num_tokens": 218498733.0, "step": 5754 }, { "epoch": 0.7320951532883857, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.650743007659912, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8531128168106079, "num_tokens": 218541365.0, "step": 5755 }, { "epoch": 0.7322223635669762, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4305074214935303, "learning_rate": 1e-05, "loss": 0.5012, "mean_token_accuracy": 0.8501822352409363, "num_tokens": 218578292.0, "step": 5756 }, { "epoch": 0.7323495738455668, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6464788913726807, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8598342537879944, "num_tokens": 218617988.0, "step": 5757 }, { "epoch": 0.7324767841241572, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4216148853302, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8565694093704224, "num_tokens": 218657759.0, "step": 5758 }, { "epoch": 0.7326039944027477, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00014781951904296875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.670980453491211, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.859375536441803, "num_tokens": 218693619.0, "step": 5759 }, { "epoch": 0.7327312046813382, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014781951904296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.427464008331299, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8651244640350342, "num_tokens": 218725978.0, "step": 5760 }, { "epoch": 0.7328584149599288, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00014781951904296875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6666383743286133, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8649161458015442, "num_tokens": 218769616.0, "step": 5761 }, { "epoch": 0.7329856252385193, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.00014781951904296875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.422006845474243, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8645716905593872, "num_tokens": 218806053.0, "step": 5762 }, { "epoch": 0.7331128355171098, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00014781951904296875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.685152769088745, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8506841659545898, "num_tokens": 218844650.0, "step": 5763 }, { "epoch": 0.7332400457957003, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4248385429382324, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8510019183158875, "num_tokens": 218885943.0, "step": 5764 }, { "epoch": 0.7333672560742908, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7004384994506836, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8596652746200562, "num_tokens": 218923939.0, "step": 5765 }, { "epoch": 0.7334944663528813, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4281535148620605, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8595079183578491, "num_tokens": 218964252.0, "step": 5766 }, { "epoch": 0.7336216766314718, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7017600536346436, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8546617031097412, "num_tokens": 219009815.0, "step": 5767 }, { "epoch": 0.7337488869100623, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.434572696685791, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8655412197113037, "num_tokens": 219047998.0, "step": 5768 }, { "epoch": 0.7338760971886529, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.724926471710205, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8444178104400635, "num_tokens": 219082447.0, "step": 5769 }, { "epoch": 0.7340033074672434, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.446599245071411, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.8484730124473572, "num_tokens": 219119393.0, "step": 5770 }, { "epoch": 0.7341305177458338, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7254321575164795, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8489534258842468, "num_tokens": 219155954.0, "step": 5771 }, { "epoch": 0.7342577280244243, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4480302333831787, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8582758903503418, "num_tokens": 219195960.0, "step": 5772 }, { "epoch": 0.7343849383030149, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.712479591369629, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.859255313873291, "num_tokens": 219238323.0, "step": 5773 }, { "epoch": 0.7345121485816054, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4501471519470215, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8639847040176392, "num_tokens": 219279044.0, "step": 5774 }, { "epoch": 0.7346393588601959, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7086102962493896, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.866804838180542, "num_tokens": 219312635.0, "step": 5775 }, { "epoch": 0.7347665691387865, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4604907035827637, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.874743640422821, "num_tokens": 219349424.0, "step": 5776 }, { "epoch": 0.7348937794173769, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6975924968719482, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8559325337409973, "num_tokens": 219389230.0, "step": 5777 }, { "epoch": 0.7350209896959674, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4843904972076416, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8598619699478149, "num_tokens": 219424082.0, "step": 5778 }, { "epoch": 0.7351481999745579, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.664015531539917, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.861994743347168, "num_tokens": 219463324.0, "step": 5779 }, { "epoch": 0.7352754102531485, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4641597270965576, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8788894414901733, "num_tokens": 219500653.0, "step": 5780 }, { "epoch": 0.735402620531739, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.651463031768799, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8697274327278137, "num_tokens": 219538906.0, "step": 5781 }, { "epoch": 0.7355298308103295, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4716897010803223, "learning_rate": 1e-05, "loss": 0.5289, "mean_token_accuracy": 0.8372442722320557, "num_tokens": 219578950.0, "step": 5782 }, { "epoch": 0.7356570410889199, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6386144161224365, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8735189437866211, "num_tokens": 219614298.0, "step": 5783 }, { "epoch": 0.7357842513675105, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4606330394744873, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8531243205070496, "num_tokens": 219655521.0, "step": 5784 }, { "epoch": 0.735911461646101, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.643176317214966, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8643431663513184, "num_tokens": 219689112.0, "step": 5785 }, { "epoch": 0.7360386719246915, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4805450439453125, "learning_rate": 1e-05, "loss": 0.547, "mean_token_accuracy": 0.8332133293151855, "num_tokens": 219727432.0, "step": 5786 }, { "epoch": 0.736165882203282, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6174468994140625, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8576545119285583, "num_tokens": 219767231.0, "step": 5787 }, { "epoch": 0.7362930924818726, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.470658779144287, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8656694889068604, "num_tokens": 219798999.0, "step": 5788 }, { "epoch": 0.736420302760463, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.589681386947632, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8689287900924683, "num_tokens": 219844867.0, "step": 5789 }, { "epoch": 0.7365475130390535, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.47469425201416, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8608911037445068, "num_tokens": 219882817.0, "step": 5790 }, { "epoch": 0.736674723317644, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5973665714263916, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8449141979217529, "num_tokens": 219916776.0, "step": 5791 }, { "epoch": 0.7368019335962346, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4563801288604736, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8791153430938721, "num_tokens": 219953471.0, "step": 5792 }, { "epoch": 0.7369291438748251, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6213743686676025, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.851213812828064, "num_tokens": 219983313.0, "step": 5793 }, { "epoch": 0.7370563541534156, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4426286220550537, "learning_rate": 1e-05, "loss": 0.5338, "mean_token_accuracy": 0.8398478627204895, "num_tokens": 220028222.0, "step": 5794 }, { "epoch": 0.737183564432006, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6141085624694824, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8592367172241211, "num_tokens": 220069005.0, "step": 5795 }, { "epoch": 0.7373107747105966, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4535505771636963, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8526427745819092, "num_tokens": 220103557.0, "step": 5796 }, { "epoch": 0.7374379849891871, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6183927059173584, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8588373064994812, "num_tokens": 220141405.0, "step": 5797 }, { "epoch": 0.7375651952677776, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.423656702041626, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8577590584754944, "num_tokens": 220184913.0, "step": 5798 }, { "epoch": 0.7376924055463682, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.631324052810669, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8561172485351562, "num_tokens": 220221886.0, "step": 5799 }, { "epoch": 0.7378196158249587, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.419429302215576, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8662815093994141, "num_tokens": 220260945.0, "step": 5800 }, { "epoch": 0.7379468261035491, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6500866413116455, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8630558252334595, "num_tokens": 220299580.0, "step": 5801 }, { "epoch": 0.7380740363821396, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4150729179382324, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8706387281417847, "num_tokens": 220338063.0, "step": 5802 }, { "epoch": 0.7382012466607302, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6572656631469727, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8674081563949585, "num_tokens": 220381212.0, "step": 5803 }, { "epoch": 0.7383284569393207, "ewc_loss": 0.05029296875, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4212543964385986, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8579415678977966, "num_tokens": 220422665.0, "step": 5804 }, { "epoch": 0.7384556672179112, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001468658447265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.670302629470825, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.868183970451355, "num_tokens": 220462044.0, "step": 5805 }, { "epoch": 0.7385828774965018, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.444878339767456, "learning_rate": 1e-05, "loss": 0.5296, "mean_token_accuracy": 0.8388176560401917, "num_tokens": 220497373.0, "step": 5806 }, { "epoch": 0.7387100877750922, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6627354621887207, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8725347518920898, "num_tokens": 220537612.0, "step": 5807 }, { "epoch": 0.7388372980536827, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4396214485168457, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8536942005157471, "num_tokens": 220574874.0, "step": 5808 }, { "epoch": 0.7389645083322732, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6617062091827393, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8719951510429382, "num_tokens": 220615699.0, "step": 5809 }, { "epoch": 0.7390917186108638, "ewc_loss": 0.050537109375, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.448664665222168, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8508822917938232, "num_tokens": 220654517.0, "step": 5810 }, { "epoch": 0.7392189288894543, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.667379140853882, "learning_rate": 1e-05, "loss": 0.5629, "mean_token_accuracy": 0.8333748579025269, "num_tokens": 220695915.0, "step": 5811 }, { "epoch": 0.7393461391680448, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4419381618499756, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8572943210601807, "num_tokens": 220734880.0, "step": 5812 }, { "epoch": 0.7394733494466353, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6817221641540527, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8551206588745117, "num_tokens": 220767800.0, "step": 5813 }, { "epoch": 0.7396005597252258, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.445012331008911, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.847997784614563, "num_tokens": 220812253.0, "step": 5814 }, { "epoch": 0.7397277700038163, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6692795753479004, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8738709688186646, "num_tokens": 220847777.0, "step": 5815 }, { "epoch": 0.7398549802824068, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4461121559143066, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8606017231941223, "num_tokens": 220884974.0, "step": 5816 }, { "epoch": 0.7399821905609973, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6670303344726562, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8576740622520447, "num_tokens": 220926236.0, "step": 5817 }, { "epoch": 0.7401094008395879, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4466757774353027, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8447955846786499, "num_tokens": 220967985.0, "step": 5818 }, { "epoch": 0.7402366111181784, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.676321268081665, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8569397330284119, "num_tokens": 221004308.0, "step": 5819 }, { "epoch": 0.7403638213967688, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4307010173797607, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8764655590057373, "num_tokens": 221043455.0, "step": 5820 }, { "epoch": 0.7404910316753593, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6877763271331787, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8483824133872986, "num_tokens": 221084634.0, "step": 5821 }, { "epoch": 0.7406182419539499, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4563286304473877, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8604507446289062, "num_tokens": 221117350.0, "step": 5822 }, { "epoch": 0.7407454522325404, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.680772066116333, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8602644801139832, "num_tokens": 221150864.0, "step": 5823 }, { "epoch": 0.7408726625111309, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4403066635131836, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8729394674301147, "num_tokens": 221186501.0, "step": 5824 }, { "epoch": 0.7409998727897215, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.688121795654297, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8402751684188843, "num_tokens": 221227734.0, "step": 5825 }, { "epoch": 0.7411270830683119, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.459080457687378, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8582751750946045, "num_tokens": 221260615.0, "step": 5826 }, { "epoch": 0.7412542933469024, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6843981742858887, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8503338098526001, "num_tokens": 221298553.0, "step": 5827 }, { "epoch": 0.7413815036254929, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.458921432495117, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8608472347259521, "num_tokens": 221338803.0, "step": 5828 }, { "epoch": 0.7415087139040835, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.000148773193359375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.661586284637451, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8620388507843018, "num_tokens": 221379557.0, "step": 5829 }, { "epoch": 0.741635924182674, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00014972686767578125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4699230194091797, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8593213558197021, "num_tokens": 221412242.0, "step": 5830 }, { "epoch": 0.7417631344612645, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00014972686767578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6764509677886963, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8757754564285278, "num_tokens": 221449382.0, "step": 5831 }, { "epoch": 0.7418903447398549, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00014972686767578125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4582912921905518, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8495129346847534, "num_tokens": 221489480.0, "step": 5832 }, { "epoch": 0.7420175550184455, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00014972686767578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6574113368988037, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.855256199836731, "num_tokens": 221525264.0, "step": 5833 }, { "epoch": 0.742144765297036, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00014972686767578125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.442160129547119, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8653463125228882, "num_tokens": 221568447.0, "step": 5834 }, { "epoch": 0.7422719755756265, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00014972686767578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6545610427856445, "learning_rate": 1e-05, "loss": 0.5491, "mean_token_accuracy": 0.8354806900024414, "num_tokens": 221612943.0, "step": 5835 }, { "epoch": 0.742399185854217, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00014972686767578125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4902989864349365, "learning_rate": 1e-05, "loss": 0.5631, "mean_token_accuracy": 0.8298096656799316, "num_tokens": 221645449.0, "step": 5836 }, { "epoch": 0.7425263961328076, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00014972686767578125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.62646746635437, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8723245859146118, "num_tokens": 221684041.0, "step": 5837 }, { "epoch": 0.742653606411398, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00014972686767578125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.465877056121826, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8685151934623718, "num_tokens": 221720200.0, "step": 5838 }, { "epoch": 0.7427808166899885, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00014972686767578125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6167490482330322, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8586801290512085, "num_tokens": 221758997.0, "step": 5839 }, { "epoch": 0.742908026968579, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00014972686767578125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.457271099090576, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8607405424118042, "num_tokens": 221800713.0, "step": 5840 }, { "epoch": 0.7430352372471696, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00014972686767578125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6251909732818604, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8471123576164246, "num_tokens": 221840336.0, "step": 5841 }, { "epoch": 0.7431624475257601, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.437685489654541, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.879406213760376, "num_tokens": 221876286.0, "step": 5842 }, { "epoch": 0.7432896578043506, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.634727954864502, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8533003926277161, "num_tokens": 221917596.0, "step": 5843 }, { "epoch": 0.743416868082941, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.449986696243286, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8668335676193237, "num_tokens": 221954045.0, "step": 5844 }, { "epoch": 0.7435440783615316, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6240406036376953, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8612249493598938, "num_tokens": 221996493.0, "step": 5845 }, { "epoch": 0.7436712886401221, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4388394355773926, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8568248748779297, "num_tokens": 222036608.0, "step": 5846 }, { "epoch": 0.7437984989187126, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.624537229537964, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.863474428653717, "num_tokens": 222080647.0, "step": 5847 }, { "epoch": 0.7439257091973032, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4532337188720703, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8505709171295166, "num_tokens": 222114039.0, "step": 5848 }, { "epoch": 0.7440529194758937, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6484923362731934, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8519531488418579, "num_tokens": 222149233.0, "step": 5849 }, { "epoch": 0.7441801297544841, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.43507981300354, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8727545142173767, "num_tokens": 222185842.0, "step": 5850 }, { "epoch": 0.7443073400330746, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6402900218963623, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8650721311569214, "num_tokens": 222233576.0, "step": 5851 }, { "epoch": 0.7444345503116652, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.444571018218994, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8617030382156372, "num_tokens": 222267822.0, "step": 5852 }, { "epoch": 0.7445617605902557, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6425280570983887, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8707246780395508, "num_tokens": 222304694.0, "step": 5853 }, { "epoch": 0.7446889708688462, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4446637630462646, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8522143363952637, "num_tokens": 222340793.0, "step": 5854 }, { "epoch": 0.7448161811474368, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6551263332366943, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8570817112922668, "num_tokens": 222380316.0, "step": 5855 }, { "epoch": 0.7449433914260272, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4269654750823975, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8556311130523682, "num_tokens": 222418867.0, "step": 5856 }, { "epoch": 0.7450706017046177, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6667566299438477, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8655476570129395, "num_tokens": 222456585.0, "step": 5857 }, { "epoch": 0.7451978119832082, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.430882692337036, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8712562322616577, "num_tokens": 222492894.0, "step": 5858 }, { "epoch": 0.7453250222617988, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.663499116897583, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8630961775779724, "num_tokens": 222538214.0, "step": 5859 }, { "epoch": 0.7454522325403893, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.427841901779175, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.860042929649353, "num_tokens": 222577643.0, "step": 5860 }, { "epoch": 0.7455794428189798, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6787056922912598, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8581506013870239, "num_tokens": 222615080.0, "step": 5861 }, { "epoch": 0.7457066530975703, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4275293350219727, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8704267740249634, "num_tokens": 222650358.0, "step": 5862 }, { "epoch": 0.7458338633761608, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6920032501220703, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8467434644699097, "num_tokens": 222691558.0, "step": 5863 }, { "epoch": 0.7459610736547513, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4176933765411377, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8585832118988037, "num_tokens": 222734632.0, "step": 5864 }, { "epoch": 0.7460882839333418, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.730699062347412, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8571330308914185, "num_tokens": 222769422.0, "step": 5865 }, { "epoch": 0.7462154942119323, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.427361488342285, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.87981116771698, "num_tokens": 222808667.0, "step": 5866 }, { "epoch": 0.7463427044905229, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7300922870635986, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8655682802200317, "num_tokens": 222844291.0, "step": 5867 }, { "epoch": 0.7464699147691134, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4546308517456055, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8446459770202637, "num_tokens": 222885812.0, "step": 5868 }, { "epoch": 0.7465971250477038, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7184340953826904, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8613986372947693, "num_tokens": 222920768.0, "step": 5869 }, { "epoch": 0.7467243353262943, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4506938457489014, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8647216558456421, "num_tokens": 222961973.0, "step": 5870 }, { "epoch": 0.7468515456048849, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7207717895507812, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8649311065673828, "num_tokens": 222997356.0, "step": 5871 }, { "epoch": 0.7469787558834754, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4555728435516357, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8673930764198303, "num_tokens": 223035933.0, "step": 5872 }, { "epoch": 0.7471059661620659, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7165122032165527, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8534877300262451, "num_tokens": 223074597.0, "step": 5873 }, { "epoch": 0.7472331764406565, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.460486888885498, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8732564449310303, "num_tokens": 223113736.0, "step": 5874 }, { "epoch": 0.7473603867192469, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7082886695861816, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8541643619537354, "num_tokens": 223153974.0, "step": 5875 }, { "epoch": 0.7474875969978374, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4913086891174316, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8675623536109924, "num_tokens": 223187043.0, "step": 5876 }, { "epoch": 0.7476148072764279, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6841537952423096, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8720049858093262, "num_tokens": 223221265.0, "step": 5877 }, { "epoch": 0.7477420175550185, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.474832057952881, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8576536774635315, "num_tokens": 223262167.0, "step": 5878 }, { "epoch": 0.747869227833609, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.667355537414551, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8511435985565186, "num_tokens": 223302333.0, "step": 5879 }, { "epoch": 0.7479964381121995, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.0001506805419921875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4716968536376953, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8698497414588928, "num_tokens": 223343337.0, "step": 5880 }, { "epoch": 0.7481236483907899, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.650149345397949, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8744005560874939, "num_tokens": 223382905.0, "step": 5881 }, { "epoch": 0.7482508586693805, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4701988697052, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8639332056045532, "num_tokens": 223422177.0, "step": 5882 }, { "epoch": 0.748378068947971, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.638399124145508, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8598711490631104, "num_tokens": 223465725.0, "step": 5883 }, { "epoch": 0.7485052792265615, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4666225910186768, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.853283166885376, "num_tokens": 223507246.0, "step": 5884 }, { "epoch": 0.748632489505152, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6250452995300293, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8680785298347473, "num_tokens": 223550736.0, "step": 5885 }, { "epoch": 0.7487596997837426, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.46938157081604, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.846541702747345, "num_tokens": 223588096.0, "step": 5886 }, { "epoch": 0.748886910062333, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6226236820220947, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8614475727081299, "num_tokens": 223629414.0, "step": 5887 }, { "epoch": 0.7490141203409235, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4668846130371094, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8596048355102539, "num_tokens": 223664468.0, "step": 5888 }, { "epoch": 0.749141330619514, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6161446571350098, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8703083992004395, "num_tokens": 223699832.0, "step": 5889 }, { "epoch": 0.7492685408981046, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.451699733734131, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8585919141769409, "num_tokens": 223736774.0, "step": 5890 }, { "epoch": 0.7493957511766951, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.613328456878662, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8642569780349731, "num_tokens": 223774622.0, "step": 5891 }, { "epoch": 0.7495229614552856, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4411251544952393, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8566504120826721, "num_tokens": 223816845.0, "step": 5892 }, { "epoch": 0.749650171733876, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.626857042312622, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8762786388397217, "num_tokens": 223849964.0, "step": 5893 }, { "epoch": 0.7497773820124666, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.444274425506592, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8524478077888489, "num_tokens": 223887821.0, "step": 5894 }, { "epoch": 0.7499045922910571, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.616964340209961, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.876127302646637, "num_tokens": 223929420.0, "step": 5895 }, { "epoch": 0.7500318025696476, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.437269449234009, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8692535161972046, "num_tokens": 223964910.0, "step": 5896 }, { "epoch": 0.7501590128482382, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.634486436843872, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8427680730819702, "num_tokens": 224001131.0, "step": 5897 }, { "epoch": 0.7502862231268287, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4589385986328125, "learning_rate": 1e-05, "loss": 0.5212, "mean_token_accuracy": 0.8441884517669678, "num_tokens": 224038969.0, "step": 5898 }, { "epoch": 0.7504134334054191, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6230995655059814, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8492845296859741, "num_tokens": 224081044.0, "step": 5899 }, { "epoch": 0.7505406436840096, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.446218967437744, "learning_rate": 1e-05, "loss": 0.5543, "mean_token_accuracy": 0.8352078199386597, "num_tokens": 224117967.0, "step": 5900 }, { "epoch": 0.7506678539626002, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.645504951477051, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8456627130508423, "num_tokens": 224148890.0, "step": 5901 }, { "epoch": 0.7507950642411907, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.454015016555786, "learning_rate": 1e-05, "loss": 0.5202, "mean_token_accuracy": 0.8439807891845703, "num_tokens": 224183592.0, "step": 5902 }, { "epoch": 0.7509222745197812, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6258647441864014, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8630568385124207, "num_tokens": 224225983.0, "step": 5903 }, { "epoch": 0.7510494847983717, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.440561056137085, "learning_rate": 1e-05, "loss": 0.5145, "mean_token_accuracy": 0.8478310108184814, "num_tokens": 224267068.0, "step": 5904 }, { "epoch": 0.7511766950769622, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6175267696380615, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8796769976615906, "num_tokens": 224313060.0, "step": 5905 }, { "epoch": 0.7513039053555527, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.438276767730713, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8542133569717407, "num_tokens": 224355066.0, "step": 5906 }, { "epoch": 0.7514311156341432, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.659940004348755, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8711661100387573, "num_tokens": 224383930.0, "step": 5907 }, { "epoch": 0.7515583259127337, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4467902183532715, "learning_rate": 1e-05, "loss": 0.5172, "mean_token_accuracy": 0.8440930843353271, "num_tokens": 224418462.0, "step": 5908 }, { "epoch": 0.7516855361913243, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6509783267974854, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8638990521430969, "num_tokens": 224453672.0, "step": 5909 }, { "epoch": 0.7518127464699148, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4212284088134766, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8649305105209351, "num_tokens": 224494093.0, "step": 5910 }, { "epoch": 0.7519399567485053, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6692652702331543, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8648499250411987, "num_tokens": 224531829.0, "step": 5911 }, { "epoch": 0.7520671670270958, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4396724700927734, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8503689169883728, "num_tokens": 224569403.0, "step": 5912 }, { "epoch": 0.7521943773056863, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00015163421630859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.710552453994751, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8547711372375488, "num_tokens": 224605189.0, "step": 5913 }, { "epoch": 0.7523215875842768, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4413390159606934, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8579582571983337, "num_tokens": 224638729.0, "step": 5914 }, { "epoch": 0.7524487978628673, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6931962966918945, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.862981915473938, "num_tokens": 224673442.0, "step": 5915 }, { "epoch": 0.7525760081414579, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4480133056640625, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8460034132003784, "num_tokens": 224705945.0, "step": 5916 }, { "epoch": 0.7527032184200484, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6919732093811035, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8618679046630859, "num_tokens": 224740035.0, "step": 5917 }, { "epoch": 0.7528304286986388, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4336163997650146, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8732380270957947, "num_tokens": 224782047.0, "step": 5918 }, { "epoch": 0.7529576389772293, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.690241575241089, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8523248434066772, "num_tokens": 224821755.0, "step": 5919 }, { "epoch": 0.7530848492558199, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4533181190490723, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8699604272842407, "num_tokens": 224854822.0, "step": 5920 }, { "epoch": 0.7532120595344104, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6739120483398438, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8655179738998413, "num_tokens": 224899570.0, "step": 5921 }, { "epoch": 0.7533392698130009, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4585821628570557, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8640002608299255, "num_tokens": 224934897.0, "step": 5922 }, { "epoch": 0.7534664800915915, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.667288064956665, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8717387914657593, "num_tokens": 224972866.0, "step": 5923 }, { "epoch": 0.7535936903701819, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4567525386810303, "learning_rate": 1e-05, "loss": 0.5181, "mean_token_accuracy": 0.8452528119087219, "num_tokens": 225015075.0, "step": 5924 }, { "epoch": 0.7537209006487724, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6756927967071533, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8616316318511963, "num_tokens": 225046729.0, "step": 5925 }, { "epoch": 0.7538481109273629, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4740538597106934, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8557754755020142, "num_tokens": 225080117.0, "step": 5926 }, { "epoch": 0.7539753212059535, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.652836561203003, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8602832555770874, "num_tokens": 225116854.0, "step": 5927 }, { "epoch": 0.754102531484544, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.450688600540161, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8683832883834839, "num_tokens": 225154044.0, "step": 5928 }, { "epoch": 0.7542297417631345, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.652043581008911, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8815892934799194, "num_tokens": 225191393.0, "step": 5929 }, { "epoch": 0.7543569520417249, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4538052082061768, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8505712747573853, "num_tokens": 225232577.0, "step": 5930 }, { "epoch": 0.7544841623203155, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.650705337524414, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8480278253555298, "num_tokens": 225270458.0, "step": 5931 }, { "epoch": 0.754611372598906, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.453662633895874, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8590681552886963, "num_tokens": 225306777.0, "step": 5932 }, { "epoch": 0.7547385828774965, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6377265453338623, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8677512407302856, "num_tokens": 225345602.0, "step": 5933 }, { "epoch": 0.754865793156087, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.457261800765991, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8686921000480652, "num_tokens": 225378885.0, "step": 5934 }, { "epoch": 0.7549930034346776, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6514387130737305, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8630377054214478, "num_tokens": 225408421.0, "step": 5935 }, { "epoch": 0.755120213713268, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.451927900314331, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8628457188606262, "num_tokens": 225441428.0, "step": 5936 }, { "epoch": 0.7552474239918585, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6517252922058105, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8516319990158081, "num_tokens": 225474912.0, "step": 5937 }, { "epoch": 0.755374634270449, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.454133987426758, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8573561906814575, "num_tokens": 225511876.0, "step": 5938 }, { "epoch": 0.7555018445490396, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6399333477020264, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8579641580581665, "num_tokens": 225548754.0, "step": 5939 }, { "epoch": 0.7556290548276301, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4523847103118896, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8501287698745728, "num_tokens": 225585554.0, "step": 5940 }, { "epoch": 0.7557562651062206, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.631876230239868, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8657835721969604, "num_tokens": 225622727.0, "step": 5941 }, { "epoch": 0.755883475384811, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4395110607147217, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8712388277053833, "num_tokens": 225662163.0, "step": 5942 }, { "epoch": 0.7560106856634016, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6455111503601074, "learning_rate": 1e-05, "loss": 0.4964, "mean_token_accuracy": 0.8480552434921265, "num_tokens": 225698932.0, "step": 5943 }, { "epoch": 0.7561378959419921, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.445727586746216, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8714842796325684, "num_tokens": 225735587.0, "step": 5944 }, { "epoch": 0.7562651062205826, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.634625196456909, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.872671365737915, "num_tokens": 225777013.0, "step": 5945 }, { "epoch": 0.7563923164991732, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4581139087677, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8652235269546509, "num_tokens": 225808496.0, "step": 5946 }, { "epoch": 0.7565195267777637, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.630195140838623, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8677293062210083, "num_tokens": 225848629.0, "step": 5947 }, { "epoch": 0.7566467370563541, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4473133087158203, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8702520132064819, "num_tokens": 225882284.0, "step": 5948 }, { "epoch": 0.7567739473349446, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.642329216003418, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.850569486618042, "num_tokens": 225919294.0, "step": 5949 }, { "epoch": 0.7569011576135352, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4531078338623047, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8479591608047485, "num_tokens": 225954742.0, "step": 5950 }, { "epoch": 0.7570283678921257, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.622659921646118, "learning_rate": 1e-05, "loss": 0.3932, "mean_token_accuracy": 0.8848294019699097, "num_tokens": 225992532.0, "step": 5951 }, { "epoch": 0.7571555781707162, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.440218210220337, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.862125039100647, "num_tokens": 226034272.0, "step": 5952 }, { "epoch": 0.7572827884493067, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6451897621154785, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8551726341247559, "num_tokens": 226066607.0, "step": 5953 }, { "epoch": 0.7574099987278972, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.442009687423706, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8533199429512024, "num_tokens": 226103077.0, "step": 5954 }, { "epoch": 0.7575372090064877, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6384806632995605, "learning_rate": 1e-05, "loss": 0.4021, "mean_token_accuracy": 0.8766471743583679, "num_tokens": 226140488.0, "step": 5955 }, { "epoch": 0.7576644192850782, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4310362339019775, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.876995861530304, "num_tokens": 226177139.0, "step": 5956 }, { "epoch": 0.7577916295636687, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6475236415863037, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8651214838027954, "num_tokens": 226216340.0, "step": 5957 }, { "epoch": 0.7579188398422593, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.43641996383667, "learning_rate": 1e-05, "loss": 0.4033, "mean_token_accuracy": 0.8792455792427063, "num_tokens": 226253507.0, "step": 5958 }, { "epoch": 0.7580460501208498, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6531999111175537, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.855944812297821, "num_tokens": 226296617.0, "step": 5959 }, { "epoch": 0.7581732603994403, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4420557022094727, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8570055961608887, "num_tokens": 226335483.0, "step": 5960 }, { "epoch": 0.7583004706780307, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6597702503204346, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8499398231506348, "num_tokens": 226373773.0, "step": 5961 }, { "epoch": 0.7584276809566213, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.430140733718872, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8653503060340881, "num_tokens": 226410333.0, "step": 5962 }, { "epoch": 0.7585548912352118, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.659210443496704, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8626571297645569, "num_tokens": 226451058.0, "step": 5963 }, { "epoch": 0.7586821015138023, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.428507089614868, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8658362627029419, "num_tokens": 226487099.0, "step": 5964 }, { "epoch": 0.7588093117923929, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.672947883605957, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8678220510482788, "num_tokens": 226521462.0, "step": 5965 }, { "epoch": 0.7589365220709834, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4392876625061035, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8607604503631592, "num_tokens": 226560573.0, "step": 5966 }, { "epoch": 0.7590637323495738, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6718342304229736, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8582572937011719, "num_tokens": 226603717.0, "step": 5967 }, { "epoch": 0.7591909426281643, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.428513288497925, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8668015599250793, "num_tokens": 226647049.0, "step": 5968 }, { "epoch": 0.7593181529067549, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.685685634613037, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8707574605941772, "num_tokens": 226682459.0, "step": 5969 }, { "epoch": 0.7594453631853454, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4270808696746826, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8561562299728394, "num_tokens": 226727381.0, "step": 5970 }, { "epoch": 0.7595725734639359, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.7020022869110107, "learning_rate": 1e-05, "loss": 0.5205, "mean_token_accuracy": 0.8428810834884644, "num_tokens": 226766732.0, "step": 5971 }, { "epoch": 0.7596997837425264, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4422969818115234, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8568739891052246, "num_tokens": 226808165.0, "step": 5972 }, { "epoch": 0.7598269940211169, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7013416290283203, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8541353940963745, "num_tokens": 226848113.0, "step": 5973 }, { "epoch": 0.7599542042997074, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4557836055755615, "learning_rate": 1e-05, "loss": 0.5323, "mean_token_accuracy": 0.8401951789855957, "num_tokens": 226889390.0, "step": 5974 }, { "epoch": 0.7600814145782979, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6982312202453613, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8507477641105652, "num_tokens": 226929111.0, "step": 5975 }, { "epoch": 0.7602086248568884, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.446934938430786, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8754318356513977, "num_tokens": 226966677.0, "step": 5976 }, { "epoch": 0.760335835135479, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.696803331375122, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8573459386825562, "num_tokens": 227004083.0, "step": 5977 }, { "epoch": 0.7604630454140695, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4454643726348877, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8587611317634583, "num_tokens": 227049954.0, "step": 5978 }, { "epoch": 0.7605902556926599, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.695913314819336, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8600384593009949, "num_tokens": 227088532.0, "step": 5979 }, { "epoch": 0.7607174659712505, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4502334594726562, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8768784999847412, "num_tokens": 227129471.0, "step": 5980 }, { "epoch": 0.760844676249841, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7019619941711426, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8580548763275146, "num_tokens": 227167443.0, "step": 5981 }, { "epoch": 0.7609718865284315, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4495370388031006, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8708541393280029, "num_tokens": 227207577.0, "step": 5982 }, { "epoch": 0.761099096807022, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6990678310394287, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8615450263023376, "num_tokens": 227247424.0, "step": 5983 }, { "epoch": 0.7612263070856126, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4528310298919678, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8610144257545471, "num_tokens": 227286361.0, "step": 5984 }, { "epoch": 0.761353517364203, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6892147064208984, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.868643045425415, "num_tokens": 227326786.0, "step": 5985 }, { "epoch": 0.7614807276427935, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.454524040222168, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8506908416748047, "num_tokens": 227370237.0, "step": 5986 }, { "epoch": 0.761607937921384, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.697162389755249, "learning_rate": 1e-05, "loss": 0.4014, "mean_token_accuracy": 0.8816455602645874, "num_tokens": 227399711.0, "step": 5987 }, { "epoch": 0.7617351481999746, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4682443141937256, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8686448335647583, "num_tokens": 227436468.0, "step": 5988 }, { "epoch": 0.7618623584785651, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.680555582046509, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8605443835258484, "num_tokens": 227476128.0, "step": 5989 }, { "epoch": 0.7619895687571556, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.474123001098633, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8634982705116272, "num_tokens": 227510499.0, "step": 5990 }, { "epoch": 0.762116779035746, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6784374713897705, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8591405749320984, "num_tokens": 227542886.0, "step": 5991 }, { "epoch": 0.7622439893143366, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4533777236938477, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8705196976661682, "num_tokens": 227583244.0, "step": 5992 }, { "epoch": 0.7623711995929271, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6632111072540283, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8687137365341187, "num_tokens": 227622690.0, "step": 5993 }, { "epoch": 0.7624984098715176, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4533867835998535, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8594034910202026, "num_tokens": 227663537.0, "step": 5994 }, { "epoch": 0.7626256201501082, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.668200731277466, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.863003134727478, "num_tokens": 227700679.0, "step": 5995 }, { "epoch": 0.7627528304286987, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.465712547302246, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8529843091964722, "num_tokens": 227738410.0, "step": 5996 }, { "epoch": 0.7628800407072891, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6501340866088867, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8659566044807434, "num_tokens": 227780849.0, "step": 5997 }, { "epoch": 0.7630072509858796, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.465848207473755, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8694286346435547, "num_tokens": 227815488.0, "step": 5998 }, { "epoch": 0.7631344612644702, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6569957733154297, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8684830665588379, "num_tokens": 227848627.0, "step": 5999 }, { "epoch": 0.7632616715430607, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4520328044891357, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8626767992973328, "num_tokens": 227888394.0, "step": 6000 }, { "epoch": 0.7633888818216512, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.647231101989746, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8695641756057739, "num_tokens": 227923785.0, "step": 6001 }, { "epoch": 0.7635160921002417, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4687836170196533, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8531756401062012, "num_tokens": 227958714.0, "step": 6002 }, { "epoch": 0.7636433023788322, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000152587890625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6436643600463867, "learning_rate": 1e-05, "loss": 0.5413, "mean_token_accuracy": 0.8358243703842163, "num_tokens": 228004414.0, "step": 6003 }, { "epoch": 0.7637705126574227, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.473113536834717, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8474668264389038, "num_tokens": 228044291.0, "step": 6004 }, { "epoch": 0.7638977229360132, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6340129375457764, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8682335615158081, "num_tokens": 228080656.0, "step": 6005 }, { "epoch": 0.7640249332146037, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4582176208496094, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.863544225692749, "num_tokens": 228117787.0, "step": 6006 }, { "epoch": 0.7641521434931943, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.642083168029785, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8553198575973511, "num_tokens": 228150620.0, "step": 6007 }, { "epoch": 0.7642793537717848, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4442834854125977, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8717842102050781, "num_tokens": 228188330.0, "step": 6008 }, { "epoch": 0.7644065640503753, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.625546932220459, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8689039349555969, "num_tokens": 228231369.0, "step": 6009 }, { "epoch": 0.7645337743289657, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.443601608276367, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8699084520339966, "num_tokens": 228269321.0, "step": 6010 }, { "epoch": 0.7646609846075563, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6444690227508545, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.870285153388977, "num_tokens": 228305240.0, "step": 6011 }, { "epoch": 0.7647881948861468, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4469714164733887, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8641716241836548, "num_tokens": 228345100.0, "step": 6012 }, { "epoch": 0.7649154051647373, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.652514696121216, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8666342496871948, "num_tokens": 228379215.0, "step": 6013 }, { "epoch": 0.7650426154433279, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.44521427154541, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8585186004638672, "num_tokens": 228415954.0, "step": 6014 }, { "epoch": 0.7651698257219184, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6545419692993164, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8470219373703003, "num_tokens": 228453569.0, "step": 6015 }, { "epoch": 0.7652970360005088, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4584884643554688, "learning_rate": 1e-05, "loss": 0.5266, "mean_token_accuracy": 0.8458861708641052, "num_tokens": 228491384.0, "step": 6016 }, { "epoch": 0.7654242462790993, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.659524440765381, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8443305492401123, "num_tokens": 228528118.0, "step": 6017 }, { "epoch": 0.7655514565576899, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.447654962539673, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.855851411819458, "num_tokens": 228568207.0, "step": 6018 }, { "epoch": 0.7656786668362804, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.659905433654785, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8599610328674316, "num_tokens": 228604313.0, "step": 6019 }, { "epoch": 0.7658058771148709, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4268527030944824, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8752953410148621, "num_tokens": 228645615.0, "step": 6020 }, { "epoch": 0.7659330873934614, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6818294525146484, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8572471737861633, "num_tokens": 228679137.0, "step": 6021 }, { "epoch": 0.7660602976720519, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4449219703674316, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.86829674243927, "num_tokens": 228717383.0, "step": 6022 }, { "epoch": 0.7661875079506424, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.688149929046631, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8728693723678589, "num_tokens": 228747260.0, "step": 6023 }, { "epoch": 0.7663147182292329, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4534149169921875, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8556692004203796, "num_tokens": 228780852.0, "step": 6024 }, { "epoch": 0.7664419285078234, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6709985733032227, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8662732839584351, "num_tokens": 228820220.0, "step": 6025 }, { "epoch": 0.766569138786414, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.440068006515503, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8669276237487793, "num_tokens": 228858903.0, "step": 6026 }, { "epoch": 0.7666963490650045, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6771974563598633, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8738712072372437, "num_tokens": 228894998.0, "step": 6027 }, { "epoch": 0.7668235593435949, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.442568302154541, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8667640686035156, "num_tokens": 228933333.0, "step": 6028 }, { "epoch": 0.7669507696221854, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.68552303314209, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.863945484161377, "num_tokens": 228972960.0, "step": 6029 }, { "epoch": 0.767077979900776, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015354156494140625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4529242515563965, "learning_rate": 1e-05, "loss": 0.4921, "mean_token_accuracy": 0.8524649739265442, "num_tokens": 229010967.0, "step": 6030 }, { "epoch": 0.7672051901793665, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.664734363555908, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8673548102378845, "num_tokens": 229052424.0, "step": 6031 }, { "epoch": 0.767332400457957, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4609644412994385, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8604884147644043, "num_tokens": 229088106.0, "step": 6032 }, { "epoch": 0.7674596107365476, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.669313430786133, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8660275936126709, "num_tokens": 229122340.0, "step": 6033 }, { "epoch": 0.767586821015138, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.468747138977051, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.854205310344696, "num_tokens": 229158299.0, "step": 6034 }, { "epoch": 0.7677140312937285, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6630215644836426, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8611458539962769, "num_tokens": 229187870.0, "step": 6035 }, { "epoch": 0.767841241572319, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4543087482452393, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8491522669792175, "num_tokens": 229226459.0, "step": 6036 }, { "epoch": 0.7679684518509096, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6667072772979736, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8638162612915039, "num_tokens": 229257973.0, "step": 6037 }, { "epoch": 0.7680956621295001, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4491162300109863, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8667123317718506, "num_tokens": 229297870.0, "step": 6038 }, { "epoch": 0.7682228724080906, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.677265167236328, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8553804159164429, "num_tokens": 229331295.0, "step": 6039 }, { "epoch": 0.768350082686681, "ewc_loss": 0.051025390625, "ewc_loss_diag": 0.0001544952392578125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.525287389755249, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8596874475479126, "num_tokens": 229369491.0, "step": 6040 }, { "epoch": 0.7684772929652716, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.61714506149292, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8549932241439819, "num_tokens": 229407836.0, "step": 6041 }, { "epoch": 0.7686045032438621, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4670357704162598, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8667513132095337, "num_tokens": 229444806.0, "step": 6042 }, { "epoch": 0.7687317135224526, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.598989248275757, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8671388030052185, "num_tokens": 229485588.0, "step": 6043 }, { "epoch": 0.7688589238010431, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4722204208374023, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8682476282119751, "num_tokens": 229521924.0, "step": 6044 }, { "epoch": 0.7689861340796337, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.601628541946411, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8663707375526428, "num_tokens": 229561511.0, "step": 6045 }, { "epoch": 0.7691133443582241, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4515740871429443, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8618521690368652, "num_tokens": 229607193.0, "step": 6046 }, { "epoch": 0.7692405546368146, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6035726070404053, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8707137107849121, "num_tokens": 229642775.0, "step": 6047 }, { "epoch": 0.7693677649154052, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4579660892486572, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8478457927703857, "num_tokens": 229680960.0, "step": 6048 }, { "epoch": 0.7694949751939957, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.615525960922241, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.87559974193573, "num_tokens": 229713473.0, "step": 6049 }, { "epoch": 0.7696221854725862, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.429713487625122, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8732426762580872, "num_tokens": 229753919.0, "step": 6050 }, { "epoch": 0.7697493957511767, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.65046763420105, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.848106861114502, "num_tokens": 229788635.0, "step": 6051 }, { "epoch": 0.7698766060297672, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.417834758758545, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8676739931106567, "num_tokens": 229831221.0, "step": 6052 }, { "epoch": 0.7700038163083577, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6762540340423584, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.850150465965271, "num_tokens": 229866762.0, "step": 6053 }, { "epoch": 0.7701310265869482, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.415477991104126, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8680861592292786, "num_tokens": 229904715.0, "step": 6054 }, { "epoch": 0.7702582368655387, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.688688039779663, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8655865788459778, "num_tokens": 229942938.0, "step": 6055 }, { "epoch": 0.7703854471441293, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.416027069091797, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8679514527320862, "num_tokens": 229982468.0, "step": 6056 }, { "epoch": 0.7705126574227198, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7153472900390625, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8551169037818909, "num_tokens": 230022386.0, "step": 6057 }, { "epoch": 0.7706398677013102, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.441941976547241, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.850114107131958, "num_tokens": 230054871.0, "step": 6058 }, { "epoch": 0.7707670779799007, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7130508422851562, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.871289074420929, "num_tokens": 230090996.0, "step": 6059 }, { "epoch": 0.7708942882584913, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.442122459411621, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8695640563964844, "num_tokens": 230124575.0, "step": 6060 }, { "epoch": 0.7710214985370818, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7046236991882324, "learning_rate": 1e-05, "loss": 0.5332, "mean_token_accuracy": 0.8426185846328735, "num_tokens": 230168225.0, "step": 6061 }, { "epoch": 0.7711487088156723, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.452016592025757, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8596612811088562, "num_tokens": 230208004.0, "step": 6062 }, { "epoch": 0.7712759190942629, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7114906311035156, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8548624515533447, "num_tokens": 230243247.0, "step": 6063 }, { "epoch": 0.7714031293728534, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4394285678863525, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8691167235374451, "num_tokens": 230286646.0, "step": 6064 }, { "epoch": 0.7715303396514438, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.733976364135742, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8426313400268555, "num_tokens": 230317705.0, "step": 6065 }, { "epoch": 0.7716575499300343, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.448856830596924, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8548554182052612, "num_tokens": 230362523.0, "step": 6066 }, { "epoch": 0.7717847602086249, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.722557306289673, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8535541892051697, "num_tokens": 230402678.0, "step": 6067 }, { "epoch": 0.7719119704872154, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.46333646774292, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8617210984230042, "num_tokens": 230440557.0, "step": 6068 }, { "epoch": 0.7720391807658059, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7239930629730225, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8620520234107971, "num_tokens": 230477602.0, "step": 6069 }, { "epoch": 0.7721663910443964, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4546120166778564, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8811103105545044, "num_tokens": 230518381.0, "step": 6070 }, { "epoch": 0.7722936013229869, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7130794525146484, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8537318110466003, "num_tokens": 230559615.0, "step": 6071 }, { "epoch": 0.7724208116015774, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.469792127609253, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8594304919242859, "num_tokens": 230595270.0, "step": 6072 }, { "epoch": 0.7725480218801679, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.712341547012329, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8511561155319214, "num_tokens": 230630405.0, "step": 6073 }, { "epoch": 0.7726752321587584, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4851231575012207, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8415403366088867, "num_tokens": 230670836.0, "step": 6074 }, { "epoch": 0.772802442437349, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6789684295654297, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8546041250228882, "num_tokens": 230709150.0, "step": 6075 }, { "epoch": 0.7729296527159395, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4935507774353027, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8583117723464966, "num_tokens": 230740306.0, "step": 6076 }, { "epoch": 0.7730568629945299, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.655146360397339, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8616706132888794, "num_tokens": 230776499.0, "step": 6077 }, { "epoch": 0.7731840732731204, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.480271100997925, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8628042340278625, "num_tokens": 230816614.0, "step": 6078 }, { "epoch": 0.773311283551711, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.622295618057251, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8545904159545898, "num_tokens": 230865271.0, "step": 6079 }, { "epoch": 0.7734384938303015, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4785261154174805, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8576859831809998, "num_tokens": 230903092.0, "step": 6080 }, { "epoch": 0.773565704108892, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6119272708892822, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8625668883323669, "num_tokens": 230942726.0, "step": 6081 }, { "epoch": 0.7736929143874826, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015544891357421875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.473478317260742, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8570736646652222, "num_tokens": 230978694.0, "step": 6082 }, { "epoch": 0.773820124666073, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.6138505935668945, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8575835227966309, "num_tokens": 231015466.0, "step": 6083 }, { "epoch": 0.7739473349446635, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4545536041259766, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8626487255096436, "num_tokens": 231054823.0, "step": 6084 }, { "epoch": 0.774074545223254, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6208441257476807, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8700357675552368, "num_tokens": 231087893.0, "step": 6085 }, { "epoch": 0.7742017555018446, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.454707384109497, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.86097252368927, "num_tokens": 231123446.0, "step": 6086 }, { "epoch": 0.7743289657804351, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.638533592224121, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8649705648422241, "num_tokens": 231151558.0, "step": 6087 }, { "epoch": 0.7744561760590256, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4341564178466797, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8647152185440063, "num_tokens": 231190715.0, "step": 6088 }, { "epoch": 0.774583386337616, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.646371603012085, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8710229396820068, "num_tokens": 231224634.0, "step": 6089 }, { "epoch": 0.7747105966162066, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.427928924560547, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8514541387557983, "num_tokens": 231265497.0, "step": 6090 }, { "epoch": 0.7748378068947971, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6562750339508057, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8637285828590393, "num_tokens": 231301886.0, "step": 6091 }, { "epoch": 0.7749650171733876, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4424819946289062, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8491629362106323, "num_tokens": 231339580.0, "step": 6092 }, { "epoch": 0.7750922274519781, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6638588905334473, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8615098595619202, "num_tokens": 231376176.0, "step": 6093 }, { "epoch": 0.7752194377305687, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.447313070297241, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8613440990447998, "num_tokens": 231409975.0, "step": 6094 }, { "epoch": 0.7753466480091591, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6599292755126953, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8566476106643677, "num_tokens": 231447998.0, "step": 6095 }, { "epoch": 0.7754738582877496, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.437922716140747, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8699280023574829, "num_tokens": 231481810.0, "step": 6096 }, { "epoch": 0.7756010685663401, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6573853492736816, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8709589838981628, "num_tokens": 231521866.0, "step": 6097 }, { "epoch": 0.7757282788449307, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.447937250137329, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.850302517414093, "num_tokens": 231558247.0, "step": 6098 }, { "epoch": 0.7758554891235212, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6658501625061035, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8649247884750366, "num_tokens": 231594183.0, "step": 6099 }, { "epoch": 0.7759826994021117, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4517111778259277, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8610692024230957, "num_tokens": 231628991.0, "step": 6100 }, { "epoch": 0.7761099096807021, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6623051166534424, "learning_rate": 1e-05, "loss": 0.5107, "mean_token_accuracy": 0.8476741909980774, "num_tokens": 231669413.0, "step": 6101 }, { "epoch": 0.7762371199592927, "ewc_loss": 0.05126953125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4437906742095947, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8624870777130127, "num_tokens": 231708765.0, "step": 6102 }, { "epoch": 0.7763643302378832, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.660316228866577, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8586775660514832, "num_tokens": 231747577.0, "step": 6103 }, { "epoch": 0.7764915405164737, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4430487155914307, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8553527593612671, "num_tokens": 231788256.0, "step": 6104 }, { "epoch": 0.7766187507950643, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6688387393951416, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8594106435775757, "num_tokens": 231825709.0, "step": 6105 }, { "epoch": 0.7767459610736548, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.442868709564209, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8651961088180542, "num_tokens": 231862240.0, "step": 6106 }, { "epoch": 0.7768731713522452, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.678577423095703, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8718871474266052, "num_tokens": 231897711.0, "step": 6107 }, { "epoch": 0.7770003816308357, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.453882932662964, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8548603057861328, "num_tokens": 231932383.0, "step": 6108 }, { "epoch": 0.7771275919094263, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6975960731506348, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8555311560630798, "num_tokens": 231960919.0, "step": 6109 }, { "epoch": 0.7772548021880168, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4384090900421143, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8797743916511536, "num_tokens": 231996435.0, "step": 6110 }, { "epoch": 0.7773820124666073, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.682560443878174, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.851625919342041, "num_tokens": 232036146.0, "step": 6111 }, { "epoch": 0.7775092227451978, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4426820278167725, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8539665937423706, "num_tokens": 232077656.0, "step": 6112 }, { "epoch": 0.7776364330237884, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.69631028175354, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8675602674484253, "num_tokens": 232113370.0, "step": 6113 }, { "epoch": 0.7777636433023788, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4361836910247803, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8674794435501099, "num_tokens": 232155450.0, "step": 6114 }, { "epoch": 0.7778908535809693, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6871354579925537, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.866683840751648, "num_tokens": 232198076.0, "step": 6115 }, { "epoch": 0.7780180638595598, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4416842460632324, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8551803827285767, "num_tokens": 232246472.0, "step": 6116 }, { "epoch": 0.7781452741381504, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.707571029663086, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8654946088790894, "num_tokens": 232280576.0, "step": 6117 }, { "epoch": 0.7782724844167409, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.450869083404541, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8677050471305847, "num_tokens": 232322172.0, "step": 6118 }, { "epoch": 0.7783996946953314, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.7141788005828857, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8621516227722168, "num_tokens": 232359317.0, "step": 6119 }, { "epoch": 0.7785269049739219, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.468916654586792, "learning_rate": 1e-05, "loss": 0.5114, "mean_token_accuracy": 0.845440149307251, "num_tokens": 232395209.0, "step": 6120 }, { "epoch": 0.7786541152525124, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.703162908554077, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8536483645439148, "num_tokens": 232433303.0, "step": 6121 }, { "epoch": 0.7787813255311029, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.468557596206665, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.8738812804222107, "num_tokens": 232464465.0, "step": 6122 }, { "epoch": 0.7789085358096934, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.675240993499756, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8769255876541138, "num_tokens": 232499846.0, "step": 6123 }, { "epoch": 0.779035746088284, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.484743118286133, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8737130165100098, "num_tokens": 232528714.0, "step": 6124 }, { "epoch": 0.7791629563668745, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.677813768386841, "learning_rate": 1e-05, "loss": 0.4944, "mean_token_accuracy": 0.8530228137969971, "num_tokens": 232561155.0, "step": 6125 }, { "epoch": 0.7792901666454649, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4819159507751465, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8756121397018433, "num_tokens": 232595685.0, "step": 6126 }, { "epoch": 0.7794173769240554, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.643815517425537, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.8484992980957031, "num_tokens": 232638683.0, "step": 6127 }, { "epoch": 0.779544587202646, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.467515230178833, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8628842830657959, "num_tokens": 232677492.0, "step": 6128 }, { "epoch": 0.7796717974812365, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6354706287384033, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8484736680984497, "num_tokens": 232717997.0, "step": 6129 }, { "epoch": 0.779799007759827, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.481153964996338, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8558686971664429, "num_tokens": 232756784.0, "step": 6130 }, { "epoch": 0.7799262180384176, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.618698835372925, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8600943088531494, "num_tokens": 232795194.0, "step": 6131 }, { "epoch": 0.780053428317008, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4859235286712646, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8654742240905762, "num_tokens": 232826505.0, "step": 6132 }, { "epoch": 0.7801806385955985, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.603008985519409, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8719192147254944, "num_tokens": 232860111.0, "step": 6133 }, { "epoch": 0.780307848874189, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4741158485412598, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8542814254760742, "num_tokens": 232895558.0, "step": 6134 }, { "epoch": 0.7804350591527796, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.59057354927063, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.868566632270813, "num_tokens": 232935737.0, "step": 6135 }, { "epoch": 0.7805622694313701, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4662625789642334, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8438916206359863, "num_tokens": 232977917.0, "step": 6136 }, { "epoch": 0.7806894797099606, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.593599319458008, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.874770998954773, "num_tokens": 233016748.0, "step": 6137 }, { "epoch": 0.780816689988551, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4634251594543457, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.863614559173584, "num_tokens": 233052814.0, "step": 6138 }, { "epoch": 0.7809439002671416, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.584944009780884, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8604214787483215, "num_tokens": 233088727.0, "step": 6139 }, { "epoch": 0.7810711105457321, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.464944839477539, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.857707142829895, "num_tokens": 233126402.0, "step": 6140 }, { "epoch": 0.7811983208243226, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.61767840385437, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.847683310508728, "num_tokens": 233161204.0, "step": 6141 }, { "epoch": 0.7813255311029131, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.453596353530884, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8619250059127808, "num_tokens": 233196527.0, "step": 6142 }, { "epoch": 0.7814527413815037, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.000156402587890625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5990798473358154, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8548749685287476, "num_tokens": 233236164.0, "step": 6143 }, { "epoch": 0.7815799516600941, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4664950370788574, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8546146154403687, "num_tokens": 233268762.0, "step": 6144 }, { "epoch": 0.7817071619386846, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.600294589996338, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8561400175094604, "num_tokens": 233307388.0, "step": 6145 }, { "epoch": 0.7818343722172751, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.440610647201538, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8657748699188232, "num_tokens": 233347902.0, "step": 6146 }, { "epoch": 0.7819615824958657, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6355814933776855, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8453389406204224, "num_tokens": 233378959.0, "step": 6147 }, { "epoch": 0.7820887927744562, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.428992748260498, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8569570779800415, "num_tokens": 233420580.0, "step": 6148 }, { "epoch": 0.7822160030530467, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6441526412963867, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8697920441627502, "num_tokens": 233451628.0, "step": 6149 }, { "epoch": 0.7823432133316371, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.419811964035034, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8539702892303467, "num_tokens": 233493994.0, "step": 6150 }, { "epoch": 0.7824704236102277, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.66137433052063, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8662274479866028, "num_tokens": 233534156.0, "step": 6151 }, { "epoch": 0.7825976338888182, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4167747497558594, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8604714870452881, "num_tokens": 233573179.0, "step": 6152 }, { "epoch": 0.7827248441674087, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6830527782440186, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8690708875656128, "num_tokens": 233606424.0, "step": 6153 }, { "epoch": 0.7828520544459993, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.417835235595703, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8557634353637695, "num_tokens": 233649155.0, "step": 6154 }, { "epoch": 0.7829792647245898, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00015735626220703125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.734041452407837, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8474565148353577, "num_tokens": 233677163.0, "step": 6155 }, { "epoch": 0.7831064750031802, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4302990436553955, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8502498865127563, "num_tokens": 233713120.0, "step": 6156 }, { "epoch": 0.7832336852817707, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7339937686920166, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8589143753051758, "num_tokens": 233751160.0, "step": 6157 }, { "epoch": 0.7833608955603613, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.439875602722168, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8594735860824585, "num_tokens": 233786474.0, "step": 6158 }, { "epoch": 0.7834881058389518, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7404255867004395, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.866769015789032, "num_tokens": 233818256.0, "step": 6159 }, { "epoch": 0.7836153161175423, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4562618732452393, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8514697551727295, "num_tokens": 233856519.0, "step": 6160 }, { "epoch": 0.7837425263961328, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.725600004196167, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8614088892936707, "num_tokens": 233896132.0, "step": 6161 }, { "epoch": 0.7838697366747234, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.464632749557495, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8611660003662109, "num_tokens": 233931668.0, "step": 6162 }, { "epoch": 0.7839969469533138, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.725749969482422, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8708289265632629, "num_tokens": 233965821.0, "step": 6163 }, { "epoch": 0.7841241572319043, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.469252586364746, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8625495433807373, "num_tokens": 234003488.0, "step": 6164 }, { "epoch": 0.7842513675104948, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7043564319610596, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8554143905639648, "num_tokens": 234044270.0, "step": 6165 }, { "epoch": 0.7843785777890854, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4662365913391113, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8804347515106201, "num_tokens": 234081754.0, "step": 6166 }, { "epoch": 0.7845057880676759, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.695340871810913, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8777562379837036, "num_tokens": 234118961.0, "step": 6167 }, { "epoch": 0.7846329983462664, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.477973461151123, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8454365134239197, "num_tokens": 234159084.0, "step": 6168 }, { "epoch": 0.7847602086248568, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6813032627105713, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8653379082679749, "num_tokens": 234201296.0, "step": 6169 }, { "epoch": 0.7848874189034474, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4698870182037354, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.867186963558197, "num_tokens": 234240661.0, "step": 6170 }, { "epoch": 0.7850146291820379, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6762704849243164, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8519954681396484, "num_tokens": 234277192.0, "step": 6171 }, { "epoch": 0.7851418394606284, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.465684175491333, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8585052490234375, "num_tokens": 234321527.0, "step": 6172 }, { "epoch": 0.785269049739219, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6846907138824463, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8492211103439331, "num_tokens": 234355578.0, "step": 6173 }, { "epoch": 0.7853962600178095, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4648163318634033, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.868684709072113, "num_tokens": 234395240.0, "step": 6174 }, { "epoch": 0.7855234702963999, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.674220561981201, "learning_rate": 1e-05, "loss": 0.5224, "mean_token_accuracy": 0.847550630569458, "num_tokens": 234436914.0, "step": 6175 }, { "epoch": 0.7856506805749904, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4591684341430664, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8604668378829956, "num_tokens": 234477811.0, "step": 6176 }, { "epoch": 0.785777890853581, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6668336391448975, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.865756630897522, "num_tokens": 234515489.0, "step": 6177 }, { "epoch": 0.7859051011321715, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4917948246002197, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8571580648422241, "num_tokens": 234544846.0, "step": 6178 }, { "epoch": 0.786032311410762, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6399130821228027, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8711543083190918, "num_tokens": 234581497.0, "step": 6179 }, { "epoch": 0.7861595216893525, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.564450979232788, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8543442487716675, "num_tokens": 234621837.0, "step": 6180 }, { "epoch": 0.786286731967943, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5790445804595947, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8660385608673096, "num_tokens": 234655049.0, "step": 6181 }, { "epoch": 0.7864139422465335, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.516470193862915, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8537107706069946, "num_tokens": 234689177.0, "step": 6182 }, { "epoch": 0.786541152525124, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5472803115844727, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8736534714698792, "num_tokens": 234722744.0, "step": 6183 }, { "epoch": 0.7866683628037145, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5078539848327637, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8659861087799072, "num_tokens": 234762998.0, "step": 6184 }, { "epoch": 0.7867955730823051, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.543879508972168, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8452563285827637, "num_tokens": 234799283.0, "step": 6185 }, { "epoch": 0.7869227833608956, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.502176523208618, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8520846962928772, "num_tokens": 234838019.0, "step": 6186 }, { "epoch": 0.787049993639486, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5493369102478027, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8546415567398071, "num_tokens": 234874449.0, "step": 6187 }, { "epoch": 0.7871772039180766, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.49043869972229, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8629153966903687, "num_tokens": 234913868.0, "step": 6188 }, { "epoch": 0.7873044141966671, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5377919673919678, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8802449703216553, "num_tokens": 234949060.0, "step": 6189 }, { "epoch": 0.7874316244752576, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4810473918914795, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8599045276641846, "num_tokens": 234987732.0, "step": 6190 }, { "epoch": 0.7875588347538481, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.542771816253662, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8700236082077026, "num_tokens": 235023030.0, "step": 6191 }, { "epoch": 0.7876860450324387, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4837520122528076, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.864730715751648, "num_tokens": 235061560.0, "step": 6192 }, { "epoch": 0.7878132553110291, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.544487476348877, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8669389486312866, "num_tokens": 235098826.0, "step": 6193 }, { "epoch": 0.7879404655896196, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.469438076019287, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8584765195846558, "num_tokens": 235142116.0, "step": 6194 }, { "epoch": 0.7880676758682101, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5503551959991455, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8559793829917908, "num_tokens": 235182920.0, "step": 6195 }, { "epoch": 0.7881948861468007, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4781064987182617, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8588812351226807, "num_tokens": 235214669.0, "step": 6196 }, { "epoch": 0.7883220964253912, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.554469585418701, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8630244135856628, "num_tokens": 235251723.0, "step": 6197 }, { "epoch": 0.7884493067039817, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4742772579193115, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8671295642852783, "num_tokens": 235285931.0, "step": 6198 }, { "epoch": 0.7885765169825721, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5589706897735596, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8691905736923218, "num_tokens": 235324571.0, "step": 6199 }, { "epoch": 0.7887037272611627, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4680981636047363, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8735896348953247, "num_tokens": 235359348.0, "step": 6200 }, { "epoch": 0.7888309375397532, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.559960126876831, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8733452558517456, "num_tokens": 235397998.0, "step": 6201 }, { "epoch": 0.7889581478183437, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4564108848571777, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8594716787338257, "num_tokens": 235436991.0, "step": 6202 }, { "epoch": 0.7890853580969343, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.55499005317688, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.878736138343811, "num_tokens": 235478129.0, "step": 6203 }, { "epoch": 0.7892125683755248, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4616899490356445, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8672926425933838, "num_tokens": 235514630.0, "step": 6204 }, { "epoch": 0.7893397786541152, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5726633071899414, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8728746175765991, "num_tokens": 235550212.0, "step": 6205 }, { "epoch": 0.7894669889327057, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4399120807647705, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8710188865661621, "num_tokens": 235592602.0, "step": 6206 }, { "epoch": 0.7895941992112963, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5906474590301514, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8733372092247009, "num_tokens": 235634199.0, "step": 6207 }, { "epoch": 0.7897214094898868, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5606849193573, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8708449602127075, "num_tokens": 235670086.0, "step": 6208 }, { "epoch": 0.7898486197684773, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.540583848953247, "learning_rate": 1e-05, "loss": 0.5255, "mean_token_accuracy": 0.8452768921852112, "num_tokens": 235709749.0, "step": 6209 }, { "epoch": 0.7899758300470678, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.487488031387329, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8641270399093628, "num_tokens": 235750574.0, "step": 6210 }, { "epoch": 0.7901030403256584, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.530623197555542, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8667398691177368, "num_tokens": 235785234.0, "step": 6211 }, { "epoch": 0.7902302506042488, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4957337379455566, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8499401807785034, "num_tokens": 235822931.0, "step": 6212 }, { "epoch": 0.7903574608828393, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.518718957901001, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8503644466400146, "num_tokens": 235863022.0, "step": 6213 }, { "epoch": 0.7904846711614298, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4964864253997803, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8634740114212036, "num_tokens": 235893275.0, "step": 6214 }, { "epoch": 0.7906118814400204, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.508368730545044, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8819769024848938, "num_tokens": 235933908.0, "step": 6215 }, { "epoch": 0.7907390917186109, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.498296022415161, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.853823721408844, "num_tokens": 235975296.0, "step": 6216 }, { "epoch": 0.7908663019972014, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5035502910614014, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8711191415786743, "num_tokens": 236012872.0, "step": 6217 }, { "epoch": 0.7909935122757918, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5016160011291504, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8647367358207703, "num_tokens": 236055762.0, "step": 6218 }, { "epoch": 0.7911207225543824, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4968461990356445, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8661892414093018, "num_tokens": 236099968.0, "step": 6219 }, { "epoch": 0.7912479328329729, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.5259861946105957, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8529582023620605, "num_tokens": 236134000.0, "step": 6220 }, { "epoch": 0.7913751431115634, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4859368801116943, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8677017092704773, "num_tokens": 236174427.0, "step": 6221 }, { "epoch": 0.791502353390154, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.55173921585083, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8539076447486877, "num_tokens": 236205643.0, "step": 6222 }, { "epoch": 0.7916295636687445, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.46942400932312, "learning_rate": 1e-05, "loss": 0.5154, "mean_token_accuracy": 0.8449335098266602, "num_tokens": 236252232.0, "step": 6223 }, { "epoch": 0.7917567739473349, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.544778347015381, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8584411144256592, "num_tokens": 236293694.0, "step": 6224 }, { "epoch": 0.7918839842259254, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4594666957855225, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8675905466079712, "num_tokens": 236335129.0, "step": 6225 }, { "epoch": 0.792011194504516, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5699117183685303, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8743164539337158, "num_tokens": 236370294.0, "step": 6226 }, { "epoch": 0.7921384047831065, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.443575859069824, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8719527721405029, "num_tokens": 236410728.0, "step": 6227 }, { "epoch": 0.792265615061697, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5782663822174072, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.862770676612854, "num_tokens": 236453053.0, "step": 6228 }, { "epoch": 0.7923928253402875, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4396252632141113, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8710812330245972, "num_tokens": 236491746.0, "step": 6229 }, { "epoch": 0.792520035618878, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6067090034484863, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8488566875457764, "num_tokens": 236530938.0, "step": 6230 }, { "epoch": 0.7926472458974685, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.419077157974243, "learning_rate": 1e-05, "loss": 0.3814, "mean_token_accuracy": 0.8881888389587402, "num_tokens": 236567765.0, "step": 6231 }, { "epoch": 0.792774456176059, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.650630474090576, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.860602855682373, "num_tokens": 236596522.0, "step": 6232 }, { "epoch": 0.7929016664546495, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.43766450881958, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8553289771080017, "num_tokens": 236637493.0, "step": 6233 }, { "epoch": 0.7930288767332401, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6642262935638428, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8593401312828064, "num_tokens": 236668405.0, "step": 6234 }, { "epoch": 0.7931560870118306, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4119491577148438, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.865851640701294, "num_tokens": 236703105.0, "step": 6235 }, { "epoch": 0.793283297290421, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6768674850463867, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8647999167442322, "num_tokens": 236747017.0, "step": 6236 }, { "epoch": 0.7934105075690115, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.3985755443573, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8693279027938843, "num_tokens": 236785659.0, "step": 6237 }, { "epoch": 0.7935377178476021, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.733215570449829, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8539661169052124, "num_tokens": 236819603.0, "step": 6238 }, { "epoch": 0.7936649281261926, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.402287244796753, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8673549890518188, "num_tokens": 236859440.0, "step": 6239 }, { "epoch": 0.7937921384047831, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7753658294677734, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8627368807792664, "num_tokens": 236892138.0, "step": 6240 }, { "epoch": 0.7939193486833737, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4263601303100586, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8453368544578552, "num_tokens": 236930559.0, "step": 6241 }, { "epoch": 0.7940465589619641, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.788759708404541, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8593347668647766, "num_tokens": 236969275.0, "step": 6242 }, { "epoch": 0.7941737692405546, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4305784702301025, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8708852529525757, "num_tokens": 237007662.0, "step": 6243 }, { "epoch": 0.7943009795191451, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.8116960525512695, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8458356857299805, "num_tokens": 237046553.0, "step": 6244 }, { "epoch": 0.7944281897977357, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.47772479057312, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8600653409957886, "num_tokens": 237082309.0, "step": 6245 }, { "epoch": 0.7945554000763262, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7929301261901855, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8707322478294373, "num_tokens": 237117997.0, "step": 6246 }, { "epoch": 0.7946826103549167, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.4911649227142334, "learning_rate": 1e-05, "loss": 0.5053, "mean_token_accuracy": 0.8478530645370483, "num_tokens": 237159169.0, "step": 6247 }, { "epoch": 0.7948098206335071, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7749736309051514, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8588925004005432, "num_tokens": 237200863.0, "step": 6248 }, { "epoch": 0.7949370309120977, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5127689838409424, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8492973446846008, "num_tokens": 237237318.0, "step": 6249 }, { "epoch": 0.7950642411906882, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.740941286087036, "learning_rate": 1e-05, "loss": 0.5155, "mean_token_accuracy": 0.8448801636695862, "num_tokens": 237278724.0, "step": 6250 }, { "epoch": 0.7951914514692787, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5207810401916504, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8665307760238647, "num_tokens": 237314190.0, "step": 6251 }, { "epoch": 0.7953186617478692, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.706751585006714, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8725836277008057, "num_tokens": 237348560.0, "step": 6252 }, { "epoch": 0.7954458720264598, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.507032632827759, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8637334108352661, "num_tokens": 237388421.0, "step": 6253 }, { "epoch": 0.7955730823050502, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6868374347686768, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8466742634773254, "num_tokens": 237424517.0, "step": 6254 }, { "epoch": 0.7957002925836407, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.48797607421875, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8658574819564819, "num_tokens": 237466195.0, "step": 6255 }, { "epoch": 0.7958275028622313, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.676882266998291, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8657253980636597, "num_tokens": 237500189.0, "step": 6256 }, { "epoch": 0.7959547131408218, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.503958225250244, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8562929630279541, "num_tokens": 237534186.0, "step": 6257 }, { "epoch": 0.7960819234194123, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.649857759475708, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8587902784347534, "num_tokens": 237572382.0, "step": 6258 }, { "epoch": 0.7962091336980028, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4781696796417236, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.859785795211792, "num_tokens": 237614300.0, "step": 6259 }, { "epoch": 0.7963363439765934, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.639125347137451, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8646174669265747, "num_tokens": 237650952.0, "step": 6260 }, { "epoch": 0.7964635542551838, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4749059677124023, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8533790111541748, "num_tokens": 237689901.0, "step": 6261 }, { "epoch": 0.7965907645337743, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.649350166320801, "learning_rate": 1e-05, "loss": 0.5072, "mean_token_accuracy": 0.8473919034004211, "num_tokens": 237724889.0, "step": 6262 }, { "epoch": 0.7967179748123648, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4632787704467773, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8663605451583862, "num_tokens": 237758853.0, "step": 6263 }, { "epoch": 0.7968451850909554, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6430559158325195, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8602450489997864, "num_tokens": 237795646.0, "step": 6264 }, { "epoch": 0.7969723953695459, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4634792804718018, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8494225144386292, "num_tokens": 237832229.0, "step": 6265 }, { "epoch": 0.7970996056481364, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6425976753234863, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8494892120361328, "num_tokens": 237870459.0, "step": 6266 }, { "epoch": 0.7972268159267268, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.443122625350952, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8574445843696594, "num_tokens": 237912367.0, "step": 6267 }, { "epoch": 0.7973540262053174, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6572773456573486, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.860883355140686, "num_tokens": 237950847.0, "step": 6268 }, { "epoch": 0.7974812364839079, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4515225887298584, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8681734800338745, "num_tokens": 237985325.0, "step": 6269 }, { "epoch": 0.7976084467624984, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.666933536529541, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.851616621017456, "num_tokens": 238021710.0, "step": 6270 }, { "epoch": 0.797735657041089, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4563989639282227, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.86716628074646, "num_tokens": 238053024.0, "step": 6271 }, { "epoch": 0.7978628673196795, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6565213203430176, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8679503202438354, "num_tokens": 238089732.0, "step": 6272 }, { "epoch": 0.7979900775982699, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.453233242034912, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8622317314147949, "num_tokens": 238127583.0, "step": 6273 }, { "epoch": 0.7981172878768604, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6550002098083496, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8606366515159607, "num_tokens": 238166328.0, "step": 6274 }, { "epoch": 0.798244498155451, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4518136978149414, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8523715734481812, "num_tokens": 238206111.0, "step": 6275 }, { "epoch": 0.7983717084340415, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.655837059020996, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8634873032569885, "num_tokens": 238243793.0, "step": 6276 }, { "epoch": 0.798498918712632, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.444668769836426, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8765349984169006, "num_tokens": 238284005.0, "step": 6277 }, { "epoch": 0.7986261289912225, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6665263175964355, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8566077351570129, "num_tokens": 238321180.0, "step": 6278 }, { "epoch": 0.798753339269813, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4301602840423584, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8671868443489075, "num_tokens": 238364383.0, "step": 6279 }, { "epoch": 0.7988805495484035, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6827244758605957, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8482685685157776, "num_tokens": 238405547.0, "step": 6280 }, { "epoch": 0.799007759826994, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4556655883789062, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8635261058807373, "num_tokens": 238436810.0, "step": 6281 }, { "epoch": 0.7991349701055845, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6944098472595215, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8531997203826904, "num_tokens": 238470062.0, "step": 6282 }, { "epoch": 0.7992621803841751, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.446290969848633, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8782594203948975, "num_tokens": 238502680.0, "step": 6283 }, { "epoch": 0.7993893906627656, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.703648805618286, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8530356884002686, "num_tokens": 238534355.0, "step": 6284 }, { "epoch": 0.799516600941356, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4448652267456055, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8595700263977051, "num_tokens": 238572225.0, "step": 6285 }, { "epoch": 0.7996438112199465, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.698216438293457, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8663971424102783, "num_tokens": 238608313.0, "step": 6286 }, { "epoch": 0.7997710214985371, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4446563720703125, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8459768295288086, "num_tokens": 238652973.0, "step": 6287 }, { "epoch": 0.7998982317771276, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7097508907318115, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8579339981079102, "num_tokens": 238688490.0, "step": 6288 }, { "epoch": 0.8000254420557181, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4462409019470215, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8701900243759155, "num_tokens": 238728641.0, "step": 6289 }, { "epoch": 0.8001526523343087, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7295689582824707, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8476958870887756, "num_tokens": 238764134.0, "step": 6290 }, { "epoch": 0.8002798626128991, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4528067111968994, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8671305775642395, "num_tokens": 238803705.0, "step": 6291 }, { "epoch": 0.8004070728914896, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.740104913711548, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8539284467697144, "num_tokens": 238836816.0, "step": 6292 }, { "epoch": 0.8005342831700801, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4637467861175537, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.869331419467926, "num_tokens": 238870006.0, "step": 6293 }, { "epoch": 0.8006614934486707, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.72402286529541, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8599192500114441, "num_tokens": 238907018.0, "step": 6294 }, { "epoch": 0.8007887037272612, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.484653949737549, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.847506582736969, "num_tokens": 238943319.0, "step": 6295 }, { "epoch": 0.8009159140058517, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.708343505859375, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8435532450675964, "num_tokens": 238978546.0, "step": 6296 }, { "epoch": 0.8010431242844421, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.489537477493286, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8574540615081787, "num_tokens": 239014785.0, "step": 6297 }, { "epoch": 0.8011703345630327, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7038087844848633, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8396475315093994, "num_tokens": 239048212.0, "step": 6298 }, { "epoch": 0.8012975448416232, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.482295274734497, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8493242263793945, "num_tokens": 239086946.0, "step": 6299 }, { "epoch": 0.8014247551202137, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.664508819580078, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8774083256721497, "num_tokens": 239125100.0, "step": 6300 }, { "epoch": 0.8015519653988042, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.470761775970459, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8529992699623108, "num_tokens": 239163118.0, "step": 6301 }, { "epoch": 0.8016791756773948, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6697216033935547, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8666712045669556, "num_tokens": 239196914.0, "step": 6302 }, { "epoch": 0.8018063859559852, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4790635108947754, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8612869381904602, "num_tokens": 239229609.0, "step": 6303 }, { "epoch": 0.8019335962345757, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6402957439422607, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8802319765090942, "num_tokens": 239269829.0, "step": 6304 }, { "epoch": 0.8020608065131662, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4705986976623535, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8763412237167358, "num_tokens": 239308563.0, "step": 6305 }, { "epoch": 0.8021880167917568, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001583099365234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6401941776275635, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8637831211090088, "num_tokens": 239348620.0, "step": 6306 }, { "epoch": 0.8023152270703473, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4877004623413086, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8600993156433105, "num_tokens": 239380706.0, "step": 6307 }, { "epoch": 0.8024424373489378, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.628824472427368, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8507986068725586, "num_tokens": 239417365.0, "step": 6308 }, { "epoch": 0.8025696476275284, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.501389980316162, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.858663022518158, "num_tokens": 239449926.0, "step": 6309 }, { "epoch": 0.8026968579061188, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5998921394348145, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8593734502792358, "num_tokens": 239485134.0, "step": 6310 }, { "epoch": 0.8028240681847093, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4757397174835205, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8699625134468079, "num_tokens": 239521353.0, "step": 6311 }, { "epoch": 0.8029512784632998, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.610783576965332, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8569853901863098, "num_tokens": 239554902.0, "step": 6312 }, { "epoch": 0.8030784887418904, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4624271392822266, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8505760431289673, "num_tokens": 239597436.0, "step": 6313 }, { "epoch": 0.8032056990204809, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6103098392486572, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8657540082931519, "num_tokens": 239635543.0, "step": 6314 }, { "epoch": 0.8033329092990714, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4514529705047607, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8717224597930908, "num_tokens": 239673956.0, "step": 6315 }, { "epoch": 0.8034601195776618, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.618722438812256, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.857135534286499, "num_tokens": 239712556.0, "step": 6316 }, { "epoch": 0.8035873298562524, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4477715492248535, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8749216794967651, "num_tokens": 239747034.0, "step": 6317 }, { "epoch": 0.8037145401348429, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.627495765686035, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8642150163650513, "num_tokens": 239787054.0, "step": 6318 }, { "epoch": 0.8038417504134334, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4548397064208984, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.84694904088974, "num_tokens": 239824796.0, "step": 6319 }, { "epoch": 0.803968960692024, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6377222537994385, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8535134792327881, "num_tokens": 239862526.0, "step": 6320 }, { "epoch": 0.8040961709706145, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.462594985961914, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8582338690757751, "num_tokens": 239897409.0, "step": 6321 }, { "epoch": 0.8042233812492049, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6277403831481934, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8683236837387085, "num_tokens": 239935202.0, "step": 6322 }, { "epoch": 0.8043505915277954, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4665889739990234, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8638330101966858, "num_tokens": 239967603.0, "step": 6323 }, { "epoch": 0.804477801806386, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.639341115951538, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.854042112827301, "num_tokens": 240000757.0, "step": 6324 }, { "epoch": 0.8046050120849765, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.446047306060791, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8536447286605835, "num_tokens": 240039747.0, "step": 6325 }, { "epoch": 0.804732222363567, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6512560844421387, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8564654588699341, "num_tokens": 240078951.0, "step": 6326 }, { "epoch": 0.8048594326421575, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.443211555480957, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8450028300285339, "num_tokens": 240120097.0, "step": 6327 }, { "epoch": 0.804986642920748, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.655787944793701, "learning_rate": 1e-05, "loss": 0.5361, "mean_token_accuracy": 0.842713475227356, "num_tokens": 240166510.0, "step": 6328 }, { "epoch": 0.8051138531993385, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.42933988571167, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8612830638885498, "num_tokens": 240207961.0, "step": 6329 }, { "epoch": 0.805241063477929, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6827592849731445, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8646714687347412, "num_tokens": 240243257.0, "step": 6330 }, { "epoch": 0.8053682737565195, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4338629245758057, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8665781021118164, "num_tokens": 240279844.0, "step": 6331 }, { "epoch": 0.8054954840351101, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6914474964141846, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8648814558982849, "num_tokens": 240316836.0, "step": 6332 }, { "epoch": 0.8056226943137006, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4379029273986816, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.845246434211731, "num_tokens": 240357794.0, "step": 6333 }, { "epoch": 0.805749904592291, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.704739570617676, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8423235416412354, "num_tokens": 240397929.0, "step": 6334 }, { "epoch": 0.8058771148708815, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.430452585220337, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8725073337554932, "num_tokens": 240434319.0, "step": 6335 }, { "epoch": 0.8060043251494721, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7120683193206787, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8607147932052612, "num_tokens": 240476755.0, "step": 6336 }, { "epoch": 0.8061315354280626, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.443448066711426, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8679386377334595, "num_tokens": 240513042.0, "step": 6337 }, { "epoch": 0.8062587457066531, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.71321702003479, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8643006086349487, "num_tokens": 240554805.0, "step": 6338 }, { "epoch": 0.8063859559852437, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4458529949188232, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8749562501907349, "num_tokens": 240593196.0, "step": 6339 }, { "epoch": 0.8065131662638341, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7110395431518555, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8646870851516724, "num_tokens": 240632640.0, "step": 6340 }, { "epoch": 0.8066403765424246, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4664809703826904, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8658480644226074, "num_tokens": 240672868.0, "step": 6341 }, { "epoch": 0.8067675868210151, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.707148551940918, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8465795516967773, "num_tokens": 240713499.0, "step": 6342 }, { "epoch": 0.8068947970996057, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.462744951248169, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8721616864204407, "num_tokens": 240759684.0, "step": 6343 }, { "epoch": 0.8070220073781962, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7032887935638428, "learning_rate": 1e-05, "loss": 0.501, "mean_token_accuracy": 0.8531733751296997, "num_tokens": 240799439.0, "step": 6344 }, { "epoch": 0.8071492176567867, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.471062421798706, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8558823466300964, "num_tokens": 240845569.0, "step": 6345 }, { "epoch": 0.8072764279353771, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6893458366394043, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8766467571258545, "num_tokens": 240879721.0, "step": 6346 }, { "epoch": 0.8074036382139677, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4632551670074463, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8643269538879395, "num_tokens": 240918433.0, "step": 6347 }, { "epoch": 0.8075308484925582, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.70686411857605, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8557443022727966, "num_tokens": 240949650.0, "step": 6348 }, { "epoch": 0.8076580587711487, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.464256525039673, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8689128756523132, "num_tokens": 240984803.0, "step": 6349 }, { "epoch": 0.8077852690497392, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6924681663513184, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8579161167144775, "num_tokens": 241023173.0, "step": 6350 }, { "epoch": 0.8079124793283298, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.452845811843872, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.865900993347168, "num_tokens": 241065235.0, "step": 6351 }, { "epoch": 0.8080396896069202, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.688988208770752, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8627258539199829, "num_tokens": 241106283.0, "step": 6352 }, { "epoch": 0.8081668998855107, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4647414684295654, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8506228923797607, "num_tokens": 241145544.0, "step": 6353 }, { "epoch": 0.8082941101641012, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.688796281814575, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8549363017082214, "num_tokens": 241183888.0, "step": 6354 }, { "epoch": 0.8084213204426918, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4564595222473145, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8770388960838318, "num_tokens": 241224650.0, "step": 6355 }, { "epoch": 0.8085485307212823, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.683861494064331, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8600970506668091, "num_tokens": 241263907.0, "step": 6356 }, { "epoch": 0.8086757409998728, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4685490131378174, "learning_rate": 1e-05, "loss": 0.5163, "mean_token_accuracy": 0.8405055999755859, "num_tokens": 241305009.0, "step": 6357 }, { "epoch": 0.8088029512784632, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6817798614501953, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8676966428756714, "num_tokens": 241344283.0, "step": 6358 }, { "epoch": 0.8089301615570538, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4657227993011475, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8634034395217896, "num_tokens": 241380243.0, "step": 6359 }, { "epoch": 0.8090573718356443, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.676079750061035, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8600481748580933, "num_tokens": 241414373.0, "step": 6360 }, { "epoch": 0.8091845821142348, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4574954509735107, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8669105768203735, "num_tokens": 241456393.0, "step": 6361 }, { "epoch": 0.8093117923928254, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6716485023498535, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8498086333274841, "num_tokens": 241496086.0, "step": 6362 }, { "epoch": 0.8094390026714159, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.45525860786438, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8658541440963745, "num_tokens": 241537814.0, "step": 6363 }, { "epoch": 0.8095662129500064, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.674287796020508, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8556068539619446, "num_tokens": 241576762.0, "step": 6364 }, { "epoch": 0.8096934232285968, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4544200897216797, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.873839795589447, "num_tokens": 241614667.0, "step": 6365 }, { "epoch": 0.8098206335071874, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6793575286865234, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8766921758651733, "num_tokens": 241647216.0, "step": 6366 }, { "epoch": 0.8099478437857779, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4537527561187744, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8692202568054199, "num_tokens": 241684278.0, "step": 6367 }, { "epoch": 0.8100750540643684, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6666688919067383, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8763267993927002, "num_tokens": 241723326.0, "step": 6368 }, { "epoch": 0.8102022643429589, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.452407121658325, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.86644446849823, "num_tokens": 241765251.0, "step": 6369 }, { "epoch": 0.8103294746215495, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.681462526321411, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8707073926925659, "num_tokens": 241801862.0, "step": 6370 }, { "epoch": 0.8104566849001399, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4524924755096436, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8615087270736694, "num_tokens": 241841448.0, "step": 6371 }, { "epoch": 0.8105838951787304, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6736903190612793, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8589931726455688, "num_tokens": 241883151.0, "step": 6372 }, { "epoch": 0.810711105457321, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.457735776901245, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8535315990447998, "num_tokens": 241924643.0, "step": 6373 }, { "epoch": 0.8108383157359115, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.686594009399414, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8567910194396973, "num_tokens": 241960361.0, "step": 6374 }, { "epoch": 0.810965526014502, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.450361728668213, "learning_rate": 1e-05, "loss": 0.51, "mean_token_accuracy": 0.8462563753128052, "num_tokens": 241999616.0, "step": 6375 }, { "epoch": 0.8110927362930925, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.68803334236145, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8572171926498413, "num_tokens": 242035824.0, "step": 6376 }, { "epoch": 0.811219946571683, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4621222019195557, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8660057783126831, "num_tokens": 242072680.0, "step": 6377 }, { "epoch": 0.8113471568502735, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6727113723754883, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8691341876983643, "num_tokens": 242111135.0, "step": 6378 }, { "epoch": 0.811474367128864, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.455214738845825, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8687176704406738, "num_tokens": 242148993.0, "step": 6379 }, { "epoch": 0.8116015774074545, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6735970973968506, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8579854965209961, "num_tokens": 242186230.0, "step": 6380 }, { "epoch": 0.8117287876860451, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.458077907562256, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8653087615966797, "num_tokens": 242222341.0, "step": 6381 }, { "epoch": 0.8118559979646356, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6842072010040283, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8532169461250305, "num_tokens": 242255408.0, "step": 6382 }, { "epoch": 0.811983208243226, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4693784713745117, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8402410745620728, "num_tokens": 242294664.0, "step": 6383 }, { "epoch": 0.8121104185218165, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6597678661346436, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8803294897079468, "num_tokens": 242328982.0, "step": 6384 }, { "epoch": 0.8122376288004071, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.46834135055542, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8508225679397583, "num_tokens": 242369542.0, "step": 6385 }, { "epoch": 0.8123648390789976, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.657660484313965, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8555547595024109, "num_tokens": 242406699.0, "step": 6386 }, { "epoch": 0.8124920493575881, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.469877243041992, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8542945384979248, "num_tokens": 242448057.0, "step": 6387 }, { "epoch": 0.8126192596361786, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.64520001411438, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8490592241287231, "num_tokens": 242489792.0, "step": 6388 }, { "epoch": 0.8127464699147691, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.454155683517456, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.856032133102417, "num_tokens": 242535458.0, "step": 6389 }, { "epoch": 0.8128736801933596, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.636457920074463, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.862252950668335, "num_tokens": 242577380.0, "step": 6390 }, { "epoch": 0.8130008904719501, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.460313558578491, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8579206466674805, "num_tokens": 242614353.0, "step": 6391 }, { "epoch": 0.8131281007505406, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.639348268508911, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8704582452774048, "num_tokens": 242648444.0, "step": 6392 }, { "epoch": 0.8132553110291312, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.454941749572754, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8590892553329468, "num_tokens": 242687658.0, "step": 6393 }, { "epoch": 0.8133825213077217, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.63778018951416, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8620714545249939, "num_tokens": 242726097.0, "step": 6394 }, { "epoch": 0.8135097315863121, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.448352098464966, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8555564880371094, "num_tokens": 242769229.0, "step": 6395 }, { "epoch": 0.8136369418649027, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.645867347717285, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.85893714427948, "num_tokens": 242805815.0, "step": 6396 }, { "epoch": 0.8137641521434932, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4708948135375977, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.847407877445221, "num_tokens": 242835929.0, "step": 6397 }, { "epoch": 0.8138913624220837, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6258811950683594, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8709140419960022, "num_tokens": 242875295.0, "step": 6398 }, { "epoch": 0.8140185727006742, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.45763897895813, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8592991828918457, "num_tokens": 242911825.0, "step": 6399 }, { "epoch": 0.8141457829792648, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6227731704711914, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8637648820877075, "num_tokens": 242950568.0, "step": 6400 }, { "epoch": 0.8142729932578552, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4795126914978027, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8678605556488037, "num_tokens": 242981223.0, "step": 6401 }, { "epoch": 0.8144002035364457, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.616504669189453, "learning_rate": 1e-05, "loss": 0.3948, "mean_token_accuracy": 0.8818107843399048, "num_tokens": 243016261.0, "step": 6402 }, { "epoch": 0.8145274138150362, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4605369567871094, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8478388786315918, "num_tokens": 243056213.0, "step": 6403 }, { "epoch": 0.8146546240936268, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.610182285308838, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8613113164901733, "num_tokens": 243097184.0, "step": 6404 }, { "epoch": 0.8147818343722173, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.455561876296997, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8606158494949341, "num_tokens": 243136995.0, "step": 6405 }, { "epoch": 0.8149090446508078, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6188793182373047, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8751032948493958, "num_tokens": 243173495.0, "step": 6406 }, { "epoch": 0.8150362549293982, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4580283164978027, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8596245646476746, "num_tokens": 243209198.0, "step": 6407 }, { "epoch": 0.8151634652079888, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.618621587753296, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8674128651618958, "num_tokens": 243249777.0, "step": 6408 }, { "epoch": 0.8152906754865793, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4493954181671143, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8510374426841736, "num_tokens": 243293094.0, "step": 6409 }, { "epoch": 0.8154178857651698, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6362247467041016, "learning_rate": 1e-05, "loss": 0.5252, "mean_token_accuracy": 0.845795214176178, "num_tokens": 243330881.0, "step": 6410 }, { "epoch": 0.8155450960437604, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4389734268188477, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8525031805038452, "num_tokens": 243371292.0, "step": 6411 }, { "epoch": 0.8156723063223509, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6609957218170166, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8475775718688965, "num_tokens": 243404925.0, "step": 6412 }, { "epoch": 0.8157995166009414, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4303503036499023, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8698216676712036, "num_tokens": 243442377.0, "step": 6413 }, { "epoch": 0.8159267268795318, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.676833152770996, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8676135540008545, "num_tokens": 243476857.0, "step": 6414 }, { "epoch": 0.8160539371581224, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4389994144439697, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8785849809646606, "num_tokens": 243506641.0, "step": 6415 }, { "epoch": 0.8161811474367129, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6881446838378906, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8569543957710266, "num_tokens": 243547555.0, "step": 6416 }, { "epoch": 0.8163083577153034, "ewc_loss": 0.051513671875, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.436147689819336, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8514728546142578, "num_tokens": 243588210.0, "step": 6417 }, { "epoch": 0.8164355679938939, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6946027278900146, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8587908148765564, "num_tokens": 243626913.0, "step": 6418 }, { "epoch": 0.8165627782724845, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.00015926361083984375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4316842555999756, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8726022243499756, "num_tokens": 243662645.0, "step": 6419 }, { "epoch": 0.8166899885510749, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7044355869293213, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.864661455154419, "num_tokens": 243704603.0, "step": 6420 }, { "epoch": 0.8168171988296654, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.446085214614868, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8563200831413269, "num_tokens": 243739571.0, "step": 6421 }, { "epoch": 0.8169444091082559, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.705270528793335, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8773584961891174, "num_tokens": 243782047.0, "step": 6422 }, { "epoch": 0.8170716193868465, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4448935985565186, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8448349237442017, "num_tokens": 243825159.0, "step": 6423 }, { "epoch": 0.817198829665437, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.718475341796875, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8604503273963928, "num_tokens": 243865627.0, "step": 6424 }, { "epoch": 0.8173260399440275, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4510326385498047, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8560985326766968, "num_tokens": 243905750.0, "step": 6425 }, { "epoch": 0.8174532502226179, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.732055902481079, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8532365560531616, "num_tokens": 243936176.0, "step": 6426 }, { "epoch": 0.8175804605012085, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4445061683654785, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8718612194061279, "num_tokens": 243979191.0, "step": 6427 }, { "epoch": 0.817707670779799, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.722249746322632, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8517322540283203, "num_tokens": 244025074.0, "step": 6428 }, { "epoch": 0.8178348810583895, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4699137210845947, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.865583062171936, "num_tokens": 244062033.0, "step": 6429 }, { "epoch": 0.8179620913369801, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7052252292633057, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.877692461013794, "num_tokens": 244100025.0, "step": 6430 }, { "epoch": 0.8180893016155706, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.473569869995117, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.8804618120193481, "num_tokens": 244136253.0, "step": 6431 }, { "epoch": 0.818216511894161, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.704951286315918, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8646048307418823, "num_tokens": 244174179.0, "step": 6432 }, { "epoch": 0.8183437221727515, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4752237796783447, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8744345307350159, "num_tokens": 244209096.0, "step": 6433 }, { "epoch": 0.8184709324513421, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6835877895355225, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8681634068489075, "num_tokens": 244248920.0, "step": 6434 }, { "epoch": 0.8185981427299326, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4727060794830322, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.861667275428772, "num_tokens": 244289338.0, "step": 6435 }, { "epoch": 0.8187253530085231, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6759371757507324, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8723863363265991, "num_tokens": 244324844.0, "step": 6436 }, { "epoch": 0.8188525632871136, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.478285312652588, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8461593985557556, "num_tokens": 244364815.0, "step": 6437 }, { "epoch": 0.8189797735657041, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.647986650466919, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8631635904312134, "num_tokens": 244407509.0, "step": 6438 }, { "epoch": 0.8191069838442946, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4712584018707275, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8755306601524353, "num_tokens": 244446663.0, "step": 6439 }, { "epoch": 0.8192341941228851, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6362545490264893, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8572823405265808, "num_tokens": 244492292.0, "step": 6440 }, { "epoch": 0.8193614044014756, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4695053100585938, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8620326519012451, "num_tokens": 244531390.0, "step": 6441 }, { "epoch": 0.8194886146800662, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.641554832458496, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8671392202377319, "num_tokens": 244567377.0, "step": 6442 }, { "epoch": 0.8196158249586567, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.460552453994751, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8780874013900757, "num_tokens": 244605690.0, "step": 6443 }, { "epoch": 0.8197430352372471, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.641493082046509, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8645837306976318, "num_tokens": 244643512.0, "step": 6444 }, { "epoch": 0.8198702455158376, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.455183982849121, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8506937026977539, "num_tokens": 244691048.0, "step": 6445 }, { "epoch": 0.8199974557944282, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6381943225860596, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8401888608932495, "num_tokens": 244734205.0, "step": 6446 }, { "epoch": 0.8201246660730187, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4749813079833984, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8602346181869507, "num_tokens": 244768302.0, "step": 6447 }, { "epoch": 0.8202518763516092, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6263153553009033, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8630748987197876, "num_tokens": 244808880.0, "step": 6448 }, { "epoch": 0.8203790866301998, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.461803436279297, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.872515082359314, "num_tokens": 244845507.0, "step": 6449 }, { "epoch": 0.8205062969087902, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.619509696960449, "learning_rate": 1e-05, "loss": 0.4991, "mean_token_accuracy": 0.8499948978424072, "num_tokens": 244888142.0, "step": 6450 }, { "epoch": 0.8206335071873807, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.452610969543457, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8785191774368286, "num_tokens": 244928612.0, "step": 6451 }, { "epoch": 0.8207607174659712, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.621159553527832, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.864317774772644, "num_tokens": 244969879.0, "step": 6452 }, { "epoch": 0.8208879277445618, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.461627960205078, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.861365556716919, "num_tokens": 245004597.0, "step": 6453 }, { "epoch": 0.8210151380231523, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6258411407470703, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8555379509925842, "num_tokens": 245050087.0, "step": 6454 }, { "epoch": 0.8211423483017428, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.45377516746521, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8549902439117432, "num_tokens": 245091626.0, "step": 6455 }, { "epoch": 0.8212695585803332, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.627315044403076, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8599230051040649, "num_tokens": 245126780.0, "step": 6456 }, { "epoch": 0.8213967688589238, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4490954875946045, "learning_rate": 1e-05, "loss": 0.5148, "mean_token_accuracy": 0.8444913625717163, "num_tokens": 245169226.0, "step": 6457 }, { "epoch": 0.8215239791375143, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.649541139602661, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.8495373725891113, "num_tokens": 245204983.0, "step": 6458 }, { "epoch": 0.8216511894161048, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.430551052093506, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8684933185577393, "num_tokens": 245249162.0, "step": 6459 }, { "epoch": 0.8217783996946953, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6641077995300293, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8476391434669495, "num_tokens": 245288883.0, "step": 6460 }, { "epoch": 0.8219056099732859, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.426100969314575, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8589184284210205, "num_tokens": 245328578.0, "step": 6461 }, { "epoch": 0.8220328202518764, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6840693950653076, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8573014140129089, "num_tokens": 245368086.0, "step": 6462 }, { "epoch": 0.8221600305304668, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4344944953918457, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8622184991836548, "num_tokens": 245405234.0, "step": 6463 }, { "epoch": 0.8222872408090574, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.704970359802246, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.869890034198761, "num_tokens": 245439759.0, "step": 6464 }, { "epoch": 0.8224144510876479, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.437685966491699, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8569084405899048, "num_tokens": 245474689.0, "step": 6465 }, { "epoch": 0.8225416613662384, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.7130725383758545, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8629311323165894, "num_tokens": 245509494.0, "step": 6466 }, { "epoch": 0.8226688716448289, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4405558109283447, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8648450374603271, "num_tokens": 245550798.0, "step": 6467 }, { "epoch": 0.8227960819234195, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.715749979019165, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8628612160682678, "num_tokens": 245588929.0, "step": 6468 }, { "epoch": 0.8229232922020099, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4447295665740967, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8647420406341553, "num_tokens": 245625952.0, "step": 6469 }, { "epoch": 0.8230505024806004, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.718430995941162, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8522401452064514, "num_tokens": 245668156.0, "step": 6470 }, { "epoch": 0.8231777127591909, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.465017318725586, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8630199432373047, "num_tokens": 245707117.0, "step": 6471 }, { "epoch": 0.8233049230377815, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7153561115264893, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8568793535232544, "num_tokens": 245747248.0, "step": 6472 }, { "epoch": 0.823432133316372, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4512383937835693, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8569567203521729, "num_tokens": 245795760.0, "step": 6473 }, { "epoch": 0.8235593435949625, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.729747772216797, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8663896322250366, "num_tokens": 245827778.0, "step": 6474 }, { "epoch": 0.8236865538735529, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4644548892974854, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8667798042297363, "num_tokens": 245866869.0, "step": 6475 }, { "epoch": 0.8238137641521435, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.714057207107544, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8752106428146362, "num_tokens": 245904686.0, "step": 6476 }, { "epoch": 0.823940974430734, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.481670618057251, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8593621850013733, "num_tokens": 245942169.0, "step": 6477 }, { "epoch": 0.8240681847093245, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7120680809020996, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8565552234649658, "num_tokens": 245979957.0, "step": 6478 }, { "epoch": 0.824195394987915, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.472888708114624, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8694807291030884, "num_tokens": 246019433.0, "step": 6479 }, { "epoch": 0.8243226052665056, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6966867446899414, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8606187105178833, "num_tokens": 246057595.0, "step": 6480 }, { "epoch": 0.824449815545096, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.463496446609497, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8731894493103027, "num_tokens": 246100141.0, "step": 6481 }, { "epoch": 0.8245770258236865, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6942710876464844, "learning_rate": 1e-05, "loss": 0.5279, "mean_token_accuracy": 0.8419256806373596, "num_tokens": 246141642.0, "step": 6482 }, { "epoch": 0.824704236102277, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.47709059715271, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8608008027076721, "num_tokens": 246177844.0, "step": 6483 }, { "epoch": 0.8248314463808676, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6835217475891113, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8583995699882507, "num_tokens": 246217470.0, "step": 6484 }, { "epoch": 0.8249586566594581, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4707181453704834, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.867617130279541, "num_tokens": 246256637.0, "step": 6485 }, { "epoch": 0.8250858669380486, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6868932247161865, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8702805042266846, "num_tokens": 246298894.0, "step": 6486 }, { "epoch": 0.8252130772166391, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.462425947189331, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8552750945091248, "num_tokens": 246340986.0, "step": 6487 }, { "epoch": 0.8253402874952296, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6807920932769775, "learning_rate": 1e-05, "loss": 0.5465, "mean_token_accuracy": 0.8342230916023254, "num_tokens": 246379201.0, "step": 6488 }, { "epoch": 0.8254674977738201, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4694952964782715, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8696997761726379, "num_tokens": 246411605.0, "step": 6489 }, { "epoch": 0.8255947080524106, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6673665046691895, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8635993003845215, "num_tokens": 246450908.0, "step": 6490 }, { "epoch": 0.8257219183310012, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4638569355010986, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.856674313545227, "num_tokens": 246493132.0, "step": 6491 }, { "epoch": 0.8258491286095917, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6720659732818604, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.850008487701416, "num_tokens": 246528532.0, "step": 6492 }, { "epoch": 0.8259763388881821, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.463613986968994, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8636062145233154, "num_tokens": 246567328.0, "step": 6493 }, { "epoch": 0.8261035491667726, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6690945625305176, "learning_rate": 1e-05, "loss": 0.5518, "mean_token_accuracy": 0.8386855125427246, "num_tokens": 246613798.0, "step": 6494 }, { "epoch": 0.8262307594453632, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.468073844909668, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8684625625610352, "num_tokens": 246649117.0, "step": 6495 }, { "epoch": 0.8263579697239537, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.654905080795288, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.865236759185791, "num_tokens": 246690304.0, "step": 6496 }, { "epoch": 0.8264851800025442, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.454970121383667, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8668644428253174, "num_tokens": 246728103.0, "step": 6497 }, { "epoch": 0.8266123902811348, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6653764247894287, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8715418577194214, "num_tokens": 246763902.0, "step": 6498 }, { "epoch": 0.8267396005597252, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4602181911468506, "learning_rate": 1e-05, "loss": 0.515, "mean_token_accuracy": 0.8468422889709473, "num_tokens": 246803899.0, "step": 6499 }, { "epoch": 0.8268668108383157, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6654417514801025, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8539549708366394, "num_tokens": 246840464.0, "step": 6500 }, { "epoch": 0.8269940211169062, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.00016021728515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4931271076202393, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8631606101989746, "num_tokens": 246886743.0, "step": 6501 }, { "epoch": 0.8271212313954968, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6374671459198, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8635639548301697, "num_tokens": 246925260.0, "step": 6502 }, { "epoch": 0.8272484416740873, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4782021045684814, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8544114828109741, "num_tokens": 246961523.0, "step": 6503 }, { "epoch": 0.8273756519526778, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6244473457336426, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8675472736358643, "num_tokens": 246996881.0, "step": 6504 }, { "epoch": 0.8275028622312682, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.476318836212158, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8555377721786499, "num_tokens": 247033492.0, "step": 6505 }, { "epoch": 0.8276300725098588, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.621835708618164, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8660458326339722, "num_tokens": 247065078.0, "step": 6506 }, { "epoch": 0.8277572827884493, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4702646732330322, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.850550651550293, "num_tokens": 247103876.0, "step": 6507 }, { "epoch": 0.8278844930670398, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6267001628875732, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8529406189918518, "num_tokens": 247142132.0, "step": 6508 }, { "epoch": 0.8280117033456303, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.456167697906494, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8618776798248291, "num_tokens": 247182141.0, "step": 6509 }, { "epoch": 0.8281389136242209, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6448380947113037, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.8445107936859131, "num_tokens": 247218240.0, "step": 6510 }, { "epoch": 0.8282661239028114, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.447406053543091, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8581919074058533, "num_tokens": 247257370.0, "step": 6511 }, { "epoch": 0.8283933341814018, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6575119495391846, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8627095818519592, "num_tokens": 247294205.0, "step": 6512 }, { "epoch": 0.8285205444599923, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.452130079269409, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8462213277816772, "num_tokens": 247329974.0, "step": 6513 }, { "epoch": 0.8286477547385829, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.669628858566284, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8577433228492737, "num_tokens": 247365279.0, "step": 6514 }, { "epoch": 0.8287749650171734, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.42290997505188, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8759359121322632, "num_tokens": 247405319.0, "step": 6515 }, { "epoch": 0.8289021752957639, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6759519577026367, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8750055432319641, "num_tokens": 247444399.0, "step": 6516 }, { "epoch": 0.8290293855743545, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4542922973632812, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8660479784011841, "num_tokens": 247475637.0, "step": 6517 }, { "epoch": 0.8291565958529449, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6789653301239014, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8667351007461548, "num_tokens": 247511009.0, "step": 6518 }, { "epoch": 0.8292838061315354, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.450225830078125, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.865888237953186, "num_tokens": 247549188.0, "step": 6519 }, { "epoch": 0.8294110164101259, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6792209148406982, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8570002913475037, "num_tokens": 247591914.0, "step": 6520 }, { "epoch": 0.8295382266887165, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4590272903442383, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8631932735443115, "num_tokens": 247626324.0, "step": 6521 }, { "epoch": 0.829665436967307, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.687238931655884, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8599218130111694, "num_tokens": 247662846.0, "step": 6522 }, { "epoch": 0.8297926472458975, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.45851731300354, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8660860061645508, "num_tokens": 247700831.0, "step": 6523 }, { "epoch": 0.8299198575244879, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6750428676605225, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8623759150505066, "num_tokens": 247740535.0, "step": 6524 }, { "epoch": 0.8300470678030785, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.464845657348633, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8591759204864502, "num_tokens": 247779490.0, "step": 6525 }, { "epoch": 0.830174278081669, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.675785541534424, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8528580665588379, "num_tokens": 247815867.0, "step": 6526 }, { "epoch": 0.8303014883602595, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4717841148376465, "learning_rate": 1e-05, "loss": 0.5636, "mean_token_accuracy": 0.832694411277771, "num_tokens": 247854871.0, "step": 6527 }, { "epoch": 0.83042869863885, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.66363787651062, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8528425693511963, "num_tokens": 247897711.0, "step": 6528 }, { "epoch": 0.8305559089174406, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4678618907928467, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8594876527786255, "num_tokens": 247933792.0, "step": 6529 }, { "epoch": 0.830683119196031, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6731033325195312, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8646537065505981, "num_tokens": 247964183.0, "step": 6530 }, { "epoch": 0.8308103294746215, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4522976875305176, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8520493507385254, "num_tokens": 248003064.0, "step": 6531 }, { "epoch": 0.830937539753212, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6719300746917725, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8675425052642822, "num_tokens": 248041025.0, "step": 6532 }, { "epoch": 0.8310647500318026, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.474215269088745, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8543594479560852, "num_tokens": 248079503.0, "step": 6533 }, { "epoch": 0.8311919603103931, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6609606742858887, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8601250648498535, "num_tokens": 248119541.0, "step": 6534 }, { "epoch": 0.8313191705889836, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4620444774627686, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8543589115142822, "num_tokens": 248163359.0, "step": 6535 }, { "epoch": 0.831446380867574, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6511332988739014, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8717672824859619, "num_tokens": 248204384.0, "step": 6536 }, { "epoch": 0.8315735911461646, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4609274864196777, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8688344359397888, "num_tokens": 248238812.0, "step": 6537 }, { "epoch": 0.8317008014247551, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.65639066696167, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8643481731414795, "num_tokens": 248277725.0, "step": 6538 }, { "epoch": 0.8318280117033456, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4746944904327393, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8555640578269958, "num_tokens": 248317591.0, "step": 6539 }, { "epoch": 0.8319552219819362, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6360249519348145, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8704102635383606, "num_tokens": 248356330.0, "step": 6540 }, { "epoch": 0.8320824322605267, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.463073968887329, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8642749786376953, "num_tokens": 248390854.0, "step": 6541 }, { "epoch": 0.8322096425391171, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6277081966400146, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8737343549728394, "num_tokens": 248431581.0, "step": 6542 }, { "epoch": 0.8323368528177076, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.450746774673462, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8702385425567627, "num_tokens": 248474009.0, "step": 6543 }, { "epoch": 0.8324640630962982, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6420304775238037, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8720608949661255, "num_tokens": 248512552.0, "step": 6544 }, { "epoch": 0.8325912733748887, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4570465087890625, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8670870661735535, "num_tokens": 248549729.0, "step": 6545 }, { "epoch": 0.8327184836534792, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6491596698760986, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8617897033691406, "num_tokens": 248589481.0, "step": 6546 }, { "epoch": 0.8328456939320698, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.461668014526367, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8735394477844238, "num_tokens": 248621871.0, "step": 6547 }, { "epoch": 0.8329729042106602, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.656508207321167, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8603439331054688, "num_tokens": 248659416.0, "step": 6548 }, { "epoch": 0.8331001144892507, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4421355724334717, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8611587882041931, "num_tokens": 248700052.0, "step": 6549 }, { "epoch": 0.8332273247678412, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.67988920211792, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8533397912979126, "num_tokens": 248731965.0, "step": 6550 }, { "epoch": 0.8333545350464318, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4445948600769043, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8569489121437073, "num_tokens": 248770030.0, "step": 6551 }, { "epoch": 0.8334817453250223, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6814346313476562, "learning_rate": 1e-05, "loss": 0.494, "mean_token_accuracy": 0.8525398969650269, "num_tokens": 248811990.0, "step": 6552 }, { "epoch": 0.8336089556036128, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4725539684295654, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8480455279350281, "num_tokens": 248842728.0, "step": 6553 }, { "epoch": 0.8337361658822032, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.666343927383423, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8456094264984131, "num_tokens": 248885084.0, "step": 6554 }, { "epoch": 0.8338633761607938, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4488394260406494, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.869877815246582, "num_tokens": 248923701.0, "step": 6555 }, { "epoch": 0.8339905864393843, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.668023109436035, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8736675977706909, "num_tokens": 248960796.0, "step": 6556 }, { "epoch": 0.8341177967179748, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.452493667602539, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8648236989974976, "num_tokens": 249000628.0, "step": 6557 }, { "epoch": 0.8342450069965653, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6875686645507812, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8663750886917114, "num_tokens": 249034409.0, "step": 6558 }, { "epoch": 0.8343722172751559, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.470491647720337, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8705922961235046, "num_tokens": 249071227.0, "step": 6559 }, { "epoch": 0.8344994275537464, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6715660095214844, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.865125298500061, "num_tokens": 249105859.0, "step": 6560 }, { "epoch": 0.8346266378323368, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4610092639923096, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.864780068397522, "num_tokens": 249148237.0, "step": 6561 }, { "epoch": 0.8347538481109273, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6807682514190674, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8530118465423584, "num_tokens": 249181745.0, "step": 6562 }, { "epoch": 0.8348810583895179, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4547736644744873, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.857340395450592, "num_tokens": 249221839.0, "step": 6563 }, { "epoch": 0.8350082686681084, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6731929779052734, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8565706014633179, "num_tokens": 249264338.0, "step": 6564 }, { "epoch": 0.8351354789466989, "ewc_loss": 0.0517578125, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4497437477111816, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8600153923034668, "num_tokens": 249304624.0, "step": 6565 }, { "epoch": 0.8352626892252895, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.681633472442627, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8625342845916748, "num_tokens": 249342996.0, "step": 6566 }, { "epoch": 0.8353898995038799, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.453610420227051, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8588981628417969, "num_tokens": 249384795.0, "step": 6567 }, { "epoch": 0.8355171097824704, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.7006375789642334, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8705556392669678, "num_tokens": 249418492.0, "step": 6568 }, { "epoch": 0.8356443200610609, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4574785232543945, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8638545274734497, "num_tokens": 249455714.0, "step": 6569 }, { "epoch": 0.8357715303396515, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.694674015045166, "learning_rate": 1e-05, "loss": 0.529, "mean_token_accuracy": 0.8457239866256714, "num_tokens": 249493162.0, "step": 6570 }, { "epoch": 0.835898740618242, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4529659748077393, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8522348403930664, "num_tokens": 249538154.0, "step": 6571 }, { "epoch": 0.8360259508968325, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6969797611236572, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8688843846321106, "num_tokens": 249576457.0, "step": 6572 }, { "epoch": 0.8361531611754229, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.470898151397705, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.859420657157898, "num_tokens": 249611408.0, "step": 6573 }, { "epoch": 0.8362803714540135, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.681795120239258, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8594327569007874, "num_tokens": 249651371.0, "step": 6574 }, { "epoch": 0.836407581732604, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.453622817993164, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8608644008636475, "num_tokens": 249697669.0, "step": 6575 }, { "epoch": 0.8365347920111945, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.677302598953247, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8540282845497131, "num_tokens": 249739640.0, "step": 6576 }, { "epoch": 0.836662002289785, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.46724534034729, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8522685766220093, "num_tokens": 249778430.0, "step": 6577 }, { "epoch": 0.8367892125683756, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.67311692237854, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8709636926651001, "num_tokens": 249817875.0, "step": 6578 }, { "epoch": 0.836916422846966, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001621246337890625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4498445987701416, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8822789192199707, "num_tokens": 249861987.0, "step": 6579 }, { "epoch": 0.8370436331255565, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.68046498298645, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.865843653678894, "num_tokens": 249903137.0, "step": 6580 }, { "epoch": 0.837170843404147, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.472705841064453, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8552666902542114, "num_tokens": 249937760.0, "step": 6581 }, { "epoch": 0.8372980536827376, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6662282943725586, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8576987385749817, "num_tokens": 249977280.0, "step": 6582 }, { "epoch": 0.8374252639613281, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4581048488616943, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8542625904083252, "num_tokens": 250022160.0, "step": 6583 }, { "epoch": 0.8375524742399186, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.672246217727661, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8662462830543518, "num_tokens": 250063328.0, "step": 6584 }, { "epoch": 0.837679684518509, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4626426696777344, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.871772050857544, "num_tokens": 250096994.0, "step": 6585 }, { "epoch": 0.8378068947970996, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.684540033340454, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8662807941436768, "num_tokens": 250126058.0, "step": 6586 }, { "epoch": 0.8379341050756901, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4558300971984863, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8486656546592712, "num_tokens": 250171965.0, "step": 6587 }, { "epoch": 0.8380613153542806, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6797702312469482, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8580917716026306, "num_tokens": 250209862.0, "step": 6588 }, { "epoch": 0.8381885256328712, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4644718170166016, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8607862591743469, "num_tokens": 250244895.0, "step": 6589 }, { "epoch": 0.8383157359114617, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6752238273620605, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8634724617004395, "num_tokens": 250283340.0, "step": 6590 }, { "epoch": 0.8384429461900521, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4680755138397217, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8602817058563232, "num_tokens": 250318671.0, "step": 6591 }, { "epoch": 0.8385701564686426, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.661127805709839, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8674730062484741, "num_tokens": 250357718.0, "step": 6592 }, { "epoch": 0.8386973667472332, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.495163917541504, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8616601228713989, "num_tokens": 250391583.0, "step": 6593 }, { "epoch": 0.8388245770258237, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6591482162475586, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8610561490058899, "num_tokens": 250420938.0, "step": 6594 }, { "epoch": 0.8389517873044142, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.465294361114502, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8725115656852722, "num_tokens": 250455608.0, "step": 6595 }, { "epoch": 0.8390789975830047, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6561801433563232, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8562920689582825, "num_tokens": 250489745.0, "step": 6596 }, { "epoch": 0.8392062078615952, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4708714485168457, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8608248233795166, "num_tokens": 250523711.0, "step": 6597 }, { "epoch": 0.8393334181401857, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6538634300231934, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8505614995956421, "num_tokens": 250562044.0, "step": 6598 }, { "epoch": 0.8394606284187762, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.44838547706604, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8585838675498962, "num_tokens": 250605930.0, "step": 6599 }, { "epoch": 0.8395878386973668, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6613049507141113, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8616330623626709, "num_tokens": 250640621.0, "step": 6600 }, { "epoch": 0.8397150489759573, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.456413507461548, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8640674948692322, "num_tokens": 250678056.0, "step": 6601 }, { "epoch": 0.8398422592545478, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.662130355834961, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8683540225028992, "num_tokens": 250719642.0, "step": 6602 }, { "epoch": 0.8399694695331382, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.469762086868286, "learning_rate": 1e-05, "loss": 0.5306, "mean_token_accuracy": 0.8400863409042358, "num_tokens": 250757643.0, "step": 6603 }, { "epoch": 0.8400966798117288, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.672323226928711, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8545202016830444, "num_tokens": 250795277.0, "step": 6604 }, { "epoch": 0.8402238900903193, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00016307830810546875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.448355197906494, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8593205809593201, "num_tokens": 250834161.0, "step": 6605 }, { "epoch": 0.8403511003689098, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6685214042663574, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8566156029701233, "num_tokens": 250872518.0, "step": 6606 }, { "epoch": 0.8404783106475003, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4528355598449707, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.866523265838623, "num_tokens": 250908516.0, "step": 6607 }, { "epoch": 0.8406055209260909, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.670081615447998, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8583266735076904, "num_tokens": 250949351.0, "step": 6608 }, { "epoch": 0.8407327312046813, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4556753635406494, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8628174066543579, "num_tokens": 250987145.0, "step": 6609 }, { "epoch": 0.8408599414832718, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6683950424194336, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8697218894958496, "num_tokens": 251024073.0, "step": 6610 }, { "epoch": 0.8409871517618623, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.939591646194458, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8528010845184326, "num_tokens": 251069537.0, "step": 6611 }, { "epoch": 0.8411143620404529, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5195584297180176, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8543356657028198, "num_tokens": 251103343.0, "step": 6612 }, { "epoch": 0.8412415723190434, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.621694326400757, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8553918600082397, "num_tokens": 251140437.0, "step": 6613 }, { "epoch": 0.8413687825976339, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.439119338989258, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8628259897232056, "num_tokens": 251178002.0, "step": 6614 }, { "epoch": 0.8414959928762245, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.651921033859253, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8723257184028625, "num_tokens": 251219694.0, "step": 6615 }, { "epoch": 0.8416232031548149, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4247548580169678, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8665037155151367, "num_tokens": 251255576.0, "step": 6616 }, { "epoch": 0.8417504134334054, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6844584941864014, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8566009998321533, "num_tokens": 251287245.0, "step": 6617 }, { "epoch": 0.8418776237119959, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.451486825942993, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8618879318237305, "num_tokens": 251320435.0, "step": 6618 }, { "epoch": 0.8420048339905865, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6822524070739746, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8699151277542114, "num_tokens": 251358893.0, "step": 6619 }, { "epoch": 0.842132044269177, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4339046478271484, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8718706965446472, "num_tokens": 251392805.0, "step": 6620 }, { "epoch": 0.8422592545477675, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.692201614379883, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8695958256721497, "num_tokens": 251436848.0, "step": 6621 }, { "epoch": 0.8423864648263579, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.439638614654541, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8710839748382568, "num_tokens": 251471590.0, "step": 6622 }, { "epoch": 0.8425136751049485, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7132394313812256, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8548920154571533, "num_tokens": 251509251.0, "step": 6623 }, { "epoch": 0.842640885383539, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.439451217651367, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8489624261856079, "num_tokens": 251549337.0, "step": 6624 }, { "epoch": 0.8427680956621295, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.735194206237793, "learning_rate": 1e-05, "loss": 0.5389, "mean_token_accuracy": 0.8406436443328857, "num_tokens": 251593953.0, "step": 6625 }, { "epoch": 0.84289530594072, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4493408203125, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.858957827091217, "num_tokens": 251631797.0, "step": 6626 }, { "epoch": 0.8430225162193106, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7470109462738037, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8479388952255249, "num_tokens": 251668378.0, "step": 6627 }, { "epoch": 0.843149726497901, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4491519927978516, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8492119312286377, "num_tokens": 251712248.0, "step": 6628 }, { "epoch": 0.8432769367764915, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.765258550643921, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8593060970306396, "num_tokens": 251748770.0, "step": 6629 }, { "epoch": 0.843404147055082, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4589874744415283, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8667093515396118, "num_tokens": 251787262.0, "step": 6630 }, { "epoch": 0.8435313573336726, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.782548666000366, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8479373455047607, "num_tokens": 251820584.0, "step": 6631 }, { "epoch": 0.8436585676122631, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4813928604125977, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8531514406204224, "num_tokens": 251861394.0, "step": 6632 }, { "epoch": 0.8437857778908536, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.75430965423584, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8607776165008545, "num_tokens": 251902144.0, "step": 6633 }, { "epoch": 0.843912988169444, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4775002002716064, "learning_rate": 1e-05, "loss": 0.3846, "mean_token_accuracy": 0.8831877708435059, "num_tokens": 251936162.0, "step": 6634 }, { "epoch": 0.8440401984480346, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.74863338470459, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8546895980834961, "num_tokens": 251977223.0, "step": 6635 }, { "epoch": 0.8441674087266251, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4997153282165527, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8655975461006165, "num_tokens": 252013413.0, "step": 6636 }, { "epoch": 0.8442946190052156, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7224414348602295, "learning_rate": 1e-05, "loss": 0.4954, "mean_token_accuracy": 0.8534746170043945, "num_tokens": 252051395.0, "step": 6637 }, { "epoch": 0.8444218292838062, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4936349391937256, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8552325963973999, "num_tokens": 252092982.0, "step": 6638 }, { "epoch": 0.8445490395623967, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.704442024230957, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8811968564987183, "num_tokens": 252128953.0, "step": 6639 }, { "epoch": 0.8446762498409871, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.484914541244507, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.850252628326416, "num_tokens": 252171438.0, "step": 6640 }, { "epoch": 0.8448034601195776, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.694901466369629, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8594726920127869, "num_tokens": 252210310.0, "step": 6641 }, { "epoch": 0.8449306703981682, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4960827827453613, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8752427101135254, "num_tokens": 252242690.0, "step": 6642 }, { "epoch": 0.8450578806767587, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6687676906585693, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8616142272949219, "num_tokens": 252281570.0, "step": 6643 }, { "epoch": 0.8451850909553492, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4968347549438477, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8477226495742798, "num_tokens": 252323478.0, "step": 6644 }, { "epoch": 0.8453123012339397, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.656688928604126, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8458799123764038, "num_tokens": 252362620.0, "step": 6645 }, { "epoch": 0.8454395115125302, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.489737033843994, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8713722229003906, "num_tokens": 252397549.0, "step": 6646 }, { "epoch": 0.8455667217911207, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.637315034866333, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8545500040054321, "num_tokens": 252433021.0, "step": 6647 }, { "epoch": 0.8456939320697112, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.492061138153076, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.860998272895813, "num_tokens": 252466455.0, "step": 6648 }, { "epoch": 0.8458211423483017, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6382980346679688, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8553472757339478, "num_tokens": 252503547.0, "step": 6649 }, { "epoch": 0.8459483526268923, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.493845224380493, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8552934527397156, "num_tokens": 252535621.0, "step": 6650 }, { "epoch": 0.8460755629054828, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6255009174346924, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8534262776374817, "num_tokens": 252573699.0, "step": 6651 }, { "epoch": 0.8462027731840732, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4649667739868164, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8688743114471436, "num_tokens": 252619816.0, "step": 6652 }, { "epoch": 0.8463299834626637, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6404664516448975, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8713576197624207, "num_tokens": 252652631.0, "step": 6653 }, { "epoch": 0.8464571937412543, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.468226909637451, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8667129278182983, "num_tokens": 252690534.0, "step": 6654 }, { "epoch": 0.8465844040198448, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6473309993743896, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8511185646057129, "num_tokens": 252723090.0, "step": 6655 }, { "epoch": 0.8467116142984353, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 5.108870983123779, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8754339218139648, "num_tokens": 252755879.0, "step": 6656 }, { "epoch": 0.8468388245770259, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 5.9604644775390625e-05, "grad_norm": 3.011439561843872, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8431683778762817, "num_tokens": 252799137.0, "step": 6657 }, { "epoch": 0.8469660348556163, "ewc_loss": 0.052001953125, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 3.552436828613281e-05, "grad_norm": 2.4998669624328613, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8597116470336914, "num_tokens": 252840871.0, "step": 6658 }, { "epoch": 0.8470932451342068, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000164031982421875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6385819911956787, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8559728860855103, "num_tokens": 252879708.0, "step": 6659 }, { "epoch": 0.8472204554127973, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6222805976867676, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8653193116188049, "num_tokens": 252914223.0, "step": 6660 }, { "epoch": 0.8473476656913879, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5792880058288574, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8614188432693481, "num_tokens": 252956666.0, "step": 6661 }, { "epoch": 0.8474748759699784, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5924971103668213, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8611552119255066, "num_tokens": 252997566.0, "step": 6662 }, { "epoch": 0.8476020862485689, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5500850677490234, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8687127232551575, "num_tokens": 253038545.0, "step": 6663 }, { "epoch": 0.8477292965271594, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7652125358581543, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8636059761047363, "num_tokens": 253078123.0, "step": 6664 }, { "epoch": 0.8478565068057499, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.477329730987549, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8734692931175232, "num_tokens": 253118273.0, "step": 6665 }, { "epoch": 0.8479837170843404, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6913204193115234, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8504897952079773, "num_tokens": 253152994.0, "step": 6666 }, { "epoch": 0.8481109273629309, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.44865083694458, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8742732405662537, "num_tokens": 253189736.0, "step": 6667 }, { "epoch": 0.8482381376415215, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.728391647338867, "learning_rate": 1e-05, "loss": 0.5057, "mean_token_accuracy": 0.850886344909668, "num_tokens": 253225872.0, "step": 6668 }, { "epoch": 0.848365347920112, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4370462894439697, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8755602836608887, "num_tokens": 253263902.0, "step": 6669 }, { "epoch": 0.8484925581987025, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.754538059234619, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8723702430725098, "num_tokens": 253297536.0, "step": 6670 }, { "epoch": 0.8486197684772929, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.445840358734131, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8726410865783691, "num_tokens": 253334624.0, "step": 6671 }, { "epoch": 0.8487469787558835, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7613818645477295, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8560341596603394, "num_tokens": 253379203.0, "step": 6672 }, { "epoch": 0.848874189034474, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4531023502349854, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8720558881759644, "num_tokens": 253417209.0, "step": 6673 }, { "epoch": 0.8490013993130645, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7791695594787598, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8489967584609985, "num_tokens": 253453909.0, "step": 6674 }, { "epoch": 0.849128609591655, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5080411434173584, "learning_rate": 1e-05, "loss": 0.5813, "mean_token_accuracy": 0.8288609981536865, "num_tokens": 253489448.0, "step": 6675 }, { "epoch": 0.8492558198702456, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.764415740966797, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8602219820022583, "num_tokens": 253517544.0, "step": 6676 }, { "epoch": 0.849383030148836, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.4665048122406006, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8685950040817261, "num_tokens": 253558829.0, "step": 6677 }, { "epoch": 0.8495102404274265, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.758070468902588, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8640734553337097, "num_tokens": 253593944.0, "step": 6678 }, { "epoch": 0.849637450706017, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.478996992111206, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8577369451522827, "num_tokens": 253633579.0, "step": 6679 }, { "epoch": 0.8497646609846076, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7351229190826416, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.857489824295044, "num_tokens": 253680247.0, "step": 6680 }, { "epoch": 0.8498918712631981, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5021274089813232, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8645627498626709, "num_tokens": 253714766.0, "step": 6681 }, { "epoch": 0.8500190815417886, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.738480806350708, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8542175889015198, "num_tokens": 253747643.0, "step": 6682 }, { "epoch": 0.850146291820379, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.488888740539551, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8407555222511292, "num_tokens": 253791252.0, "step": 6683 }, { "epoch": 0.8502735020989696, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7246017456054688, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8668668270111084, "num_tokens": 253822244.0, "step": 6684 }, { "epoch": 0.8504007123775601, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.494992256164551, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8547571301460266, "num_tokens": 253858680.0, "step": 6685 }, { "epoch": 0.8505279226561506, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6979458332061768, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8723584413528442, "num_tokens": 253896266.0, "step": 6686 }, { "epoch": 0.8506551329347412, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4965922832489014, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8623843193054199, "num_tokens": 253934059.0, "step": 6687 }, { "epoch": 0.8507823432133317, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6910624504089355, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8626912236213684, "num_tokens": 253967862.0, "step": 6688 }, { "epoch": 0.8509095534919221, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4884932041168213, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.854979395866394, "num_tokens": 254008277.0, "step": 6689 }, { "epoch": 0.8510367637705126, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6887643337249756, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.863654375076294, "num_tokens": 254039939.0, "step": 6690 }, { "epoch": 0.8511639740491032, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4752416610717773, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8671853542327881, "num_tokens": 254080497.0, "step": 6691 }, { "epoch": 0.8512911843276937, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6921768188476562, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8638961315155029, "num_tokens": 254111917.0, "step": 6692 }, { "epoch": 0.8514183946062842, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.476647138595581, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8711754083633423, "num_tokens": 254150245.0, "step": 6693 }, { "epoch": 0.8515456048848747, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.680633068084717, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8614748120307922, "num_tokens": 254187106.0, "step": 6694 }, { "epoch": 0.8516728151634652, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4858317375183105, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8531075119972229, "num_tokens": 254219877.0, "step": 6695 }, { "epoch": 0.8518000254420557, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.666271924972534, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8533322811126709, "num_tokens": 254264234.0, "step": 6696 }, { "epoch": 0.8519272357206462, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4833641052246094, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8668538331985474, "num_tokens": 254299762.0, "step": 6697 }, { "epoch": 0.8520544459992367, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.668424367904663, "learning_rate": 1e-05, "loss": 0.4998, "mean_token_accuracy": 0.8479421734809875, "num_tokens": 254336403.0, "step": 6698 }, { "epoch": 0.8521816562778273, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.470914602279663, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8662824630737305, "num_tokens": 254375971.0, "step": 6699 }, { "epoch": 0.8523088665564178, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6560122966766357, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8532570600509644, "num_tokens": 254418430.0, "step": 6700 }, { "epoch": 0.8524360768350082, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.599691390991211, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8622323274612427, "num_tokens": 254457275.0, "step": 6701 }, { "epoch": 0.8525632871135987, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5755131244659424, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8567171096801758, "num_tokens": 254499280.0, "step": 6702 }, { "epoch": 0.8526904973921893, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5194246768951416, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.856224000453949, "num_tokens": 254536571.0, "step": 6703 }, { "epoch": 0.8528177076707798, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.553774118423462, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8512500524520874, "num_tokens": 254576900.0, "step": 6704 }, { "epoch": 0.8529449179493703, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001659393310546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.514191150665283, "learning_rate": 1e-05, "loss": 0.5392, "mean_token_accuracy": 0.8379685878753662, "num_tokens": 254617791.0, "step": 6705 }, { "epoch": 0.8530721282279609, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5438177585601807, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8697592616081238, "num_tokens": 254658300.0, "step": 6706 }, { "epoch": 0.8531993385065513, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.510498046875, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8603761196136475, "num_tokens": 254698133.0, "step": 6707 }, { "epoch": 0.8533265487851418, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.539937734603882, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8664466738700867, "num_tokens": 254736099.0, "step": 6708 }, { "epoch": 0.8534537590637323, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5079121589660645, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8740042448043823, "num_tokens": 254773460.0, "step": 6709 }, { "epoch": 0.8535809693423229, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.53847336769104, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8666218519210815, "num_tokens": 254812722.0, "step": 6710 }, { "epoch": 0.8537081796209134, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4972803592681885, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8656169772148132, "num_tokens": 254851994.0, "step": 6711 }, { "epoch": 0.8538353898995039, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5604538917541504, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8541473150253296, "num_tokens": 254889111.0, "step": 6712 }, { "epoch": 0.8539626001780944, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4854626655578613, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8583831191062927, "num_tokens": 254931779.0, "step": 6713 }, { "epoch": 0.8540898104566849, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.559518814086914, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8653436303138733, "num_tokens": 254968709.0, "step": 6714 }, { "epoch": 0.8542170207352754, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4939584732055664, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.864795446395874, "num_tokens": 255000235.0, "step": 6715 }, { "epoch": 0.8543442310138659, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5508341789245605, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8570874333381653, "num_tokens": 255041157.0, "step": 6716 }, { "epoch": 0.8544714412924564, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.486224412918091, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8584180474281311, "num_tokens": 255077001.0, "step": 6717 }, { "epoch": 0.854598651571047, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.553921699523926, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8676151037216187, "num_tokens": 255113581.0, "step": 6718 }, { "epoch": 0.8547258618496375, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4713504314422607, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8749369382858276, "num_tokens": 255153183.0, "step": 6719 }, { "epoch": 0.8548530721282279, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5731828212738037, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8690820932388306, "num_tokens": 255192216.0, "step": 6720 }, { "epoch": 0.8549802824068184, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.461337089538574, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8671442270278931, "num_tokens": 255229969.0, "step": 6721 }, { "epoch": 0.855107492685409, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5891225337982178, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8520913124084473, "num_tokens": 255267232.0, "step": 6722 }, { "epoch": 0.8552347029639995, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.449542999267578, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8641818761825562, "num_tokens": 255304923.0, "step": 6723 }, { "epoch": 0.85536191324259, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6142258644104004, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8417156934738159, "num_tokens": 255342422.0, "step": 6724 }, { "epoch": 0.8554891235211806, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4313135147094727, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8691543340682983, "num_tokens": 255380218.0, "step": 6725 }, { "epoch": 0.855616333799771, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6473681926727295, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8637354969978333, "num_tokens": 255413964.0, "step": 6726 }, { "epoch": 0.8557435440783615, "ewc_loss": 0.05224609375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4164862632751465, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8690943121910095, "num_tokens": 255450421.0, "step": 6727 }, { "epoch": 0.855870754356952, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6830289363861084, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8628965616226196, "num_tokens": 255487239.0, "step": 6728 }, { "epoch": 0.8559979646355426, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.41701340675354, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.861678957939148, "num_tokens": 255529363.0, "step": 6729 }, { "epoch": 0.8561251749141331, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6943745613098145, "learning_rate": 1e-05, "loss": 0.3791, "mean_token_accuracy": 0.8856052160263062, "num_tokens": 255570790.0, "step": 6730 }, { "epoch": 0.8562523851927236, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4196016788482666, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8586322069168091, "num_tokens": 255608188.0, "step": 6731 }, { "epoch": 0.856379595471314, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7340619564056396, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8622913360595703, "num_tokens": 255640929.0, "step": 6732 }, { "epoch": 0.8565068057499046, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4148054122924805, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8659346103668213, "num_tokens": 255684579.0, "step": 6733 }, { "epoch": 0.8566340160284951, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7659847736358643, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8614081144332886, "num_tokens": 255721195.0, "step": 6734 }, { "epoch": 0.8567612263070856, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4282448291778564, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.862156867980957, "num_tokens": 255761697.0, "step": 6735 }, { "epoch": 0.8568884365856761, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7824060916900635, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8608371019363403, "num_tokens": 255800492.0, "step": 6736 }, { "epoch": 0.8570156468642667, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.447567939758301, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8636125326156616, "num_tokens": 255837074.0, "step": 6737 }, { "epoch": 0.8571428571428571, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.796175956726074, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.862147331237793, "num_tokens": 255876238.0, "step": 6738 }, { "epoch": 0.8572700674214476, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4881489276885986, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8473325371742249, "num_tokens": 255914622.0, "step": 6739 }, { "epoch": 0.8573972777000382, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7813026905059814, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8643656969070435, "num_tokens": 255954027.0, "step": 6740 }, { "epoch": 0.8575244879786287, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4924848079681396, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8544498085975647, "num_tokens": 255997416.0, "step": 6741 }, { "epoch": 0.8576516982572192, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7708280086517334, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8605931997299194, "num_tokens": 256030447.0, "step": 6742 }, { "epoch": 0.8577789085358097, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.496224880218506, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8655131459236145, "num_tokens": 256068533.0, "step": 6743 }, { "epoch": 0.8579061188144002, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7475686073303223, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8763513565063477, "num_tokens": 256104783.0, "step": 6744 }, { "epoch": 0.8580333290929907, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.512436866760254, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8656553030014038, "num_tokens": 256138272.0, "step": 6745 }, { "epoch": 0.8581605393715812, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7323367595672607, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8537033796310425, "num_tokens": 256171102.0, "step": 6746 }, { "epoch": 0.8582877496501717, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5013530254364014, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8632391691207886, "num_tokens": 256210593.0, "step": 6747 }, { "epoch": 0.8584149599287623, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7041170597076416, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8717069625854492, "num_tokens": 256248693.0, "step": 6748 }, { "epoch": 0.8585421702073528, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5033187866210938, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8679770231246948, "num_tokens": 256284875.0, "step": 6749 }, { "epoch": 0.8586693804859432, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.68750262260437, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8665367364883423, "num_tokens": 256319121.0, "step": 6750 }, { "epoch": 0.8587965907645337, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4971954822540283, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8716816902160645, "num_tokens": 256354112.0, "step": 6751 }, { "epoch": 0.8589238010431243, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6718204021453857, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8740776777267456, "num_tokens": 256387612.0, "step": 6752 }, { "epoch": 0.8590510113217148, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.501610040664673, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8614705801010132, "num_tokens": 256419720.0, "step": 6753 }, { "epoch": 0.8591782216003053, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6483330726623535, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.852588415145874, "num_tokens": 256456717.0, "step": 6754 }, { "epoch": 0.8593054318788959, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4931488037109375, "learning_rate": 1e-05, "loss": 0.5186, "mean_token_accuracy": 0.8431161642074585, "num_tokens": 256496480.0, "step": 6755 }, { "epoch": 0.8594326421574863, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6368179321289062, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8656677007675171, "num_tokens": 256531974.0, "step": 6756 }, { "epoch": 0.8595598524360768, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4915337562561035, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8564386367797852, "num_tokens": 256565983.0, "step": 6757 }, { "epoch": 0.8596870627146673, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.610658645629883, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8693735599517822, "num_tokens": 256607215.0, "step": 6758 }, { "epoch": 0.8598142729932579, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.477829694747925, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8747237920761108, "num_tokens": 256646213.0, "step": 6759 }, { "epoch": 0.8599414832718484, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6110808849334717, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8686627149581909, "num_tokens": 256687638.0, "step": 6760 }, { "epoch": 0.8600686935504389, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.497807264328003, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.859573483467102, "num_tokens": 256716852.0, "step": 6761 }, { "epoch": 0.8601959038290294, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.611523151397705, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8615584373474121, "num_tokens": 256748260.0, "step": 6762 }, { "epoch": 0.8603231141076199, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4859094619750977, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.8437864184379578, "num_tokens": 256786400.0, "step": 6763 }, { "epoch": 0.8604503243862104, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5958871841430664, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8721743822097778, "num_tokens": 256820186.0, "step": 6764 }, { "epoch": 0.8605775346648009, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.480750322341919, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8632463216781616, "num_tokens": 256855735.0, "step": 6765 }, { "epoch": 0.8607047449433914, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.602480411529541, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8712274432182312, "num_tokens": 256887809.0, "step": 6766 }, { "epoch": 0.860831955221982, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4780073165893555, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8658721446990967, "num_tokens": 256921801.0, "step": 6767 }, { "epoch": 0.8609591655005725, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5873818397521973, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8775280714035034, "num_tokens": 256960399.0, "step": 6768 }, { "epoch": 0.8610863757791629, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4724957942962646, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8655855655670166, "num_tokens": 256995094.0, "step": 6769 }, { "epoch": 0.8612135860577534, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.599470853805542, "learning_rate": 1e-05, "loss": 0.5395, "mean_token_accuracy": 0.842532753944397, "num_tokens": 257037792.0, "step": 6770 }, { "epoch": 0.861340796336344, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.472324848175049, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8659013509750366, "num_tokens": 257075376.0, "step": 6771 }, { "epoch": 0.8614680066149345, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.604177713394165, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8558946847915649, "num_tokens": 257115878.0, "step": 6772 }, { "epoch": 0.861595216893525, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4695465564727783, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8618626594543457, "num_tokens": 257150780.0, "step": 6773 }, { "epoch": 0.8617224271721156, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.597705364227295, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8655779361724854, "num_tokens": 257194704.0, "step": 6774 }, { "epoch": 0.861849637450706, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.461057186126709, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8504315614700317, "num_tokens": 257235337.0, "step": 6775 }, { "epoch": 0.8619768477292965, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6380016803741455, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8502895832061768, "num_tokens": 257268976.0, "step": 6776 }, { "epoch": 0.862104058007887, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.451277017593384, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8635815382003784, "num_tokens": 257304638.0, "step": 6777 }, { "epoch": 0.8622312682864776, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.643991708755493, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8546723127365112, "num_tokens": 257341773.0, "step": 6778 }, { "epoch": 0.8623584785650681, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.446012496948242, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8659875392913818, "num_tokens": 257375145.0, "step": 6779 }, { "epoch": 0.8624856888436586, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6562671661376953, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8684302568435669, "num_tokens": 257411070.0, "step": 6780 }, { "epoch": 0.862612899122249, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.440402030944824, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8512238264083862, "num_tokens": 257447283.0, "step": 6781 }, { "epoch": 0.8627401094008396, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6657097339630127, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8632829189300537, "num_tokens": 257487370.0, "step": 6782 }, { "epoch": 0.8628673196794301, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4445762634277344, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8492957353591919, "num_tokens": 257525028.0, "step": 6783 }, { "epoch": 0.8629945299580206, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.677604913711548, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8591930270195007, "num_tokens": 257563884.0, "step": 6784 }, { "epoch": 0.8631217402366111, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.443620204925537, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8666542172431946, "num_tokens": 257599431.0, "step": 6785 }, { "epoch": 0.8632489505152017, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.690763473510742, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8619610071182251, "num_tokens": 257636060.0, "step": 6786 }, { "epoch": 0.8633761607937921, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4392855167388916, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8699583411216736, "num_tokens": 257672374.0, "step": 6787 }, { "epoch": 0.8635033710723826, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7072031497955322, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8552958965301514, "num_tokens": 257710875.0, "step": 6788 }, { "epoch": 0.8636305813509731, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4393184185028076, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8504977226257324, "num_tokens": 257756697.0, "step": 6789 }, { "epoch": 0.8637577916295637, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7236688137054443, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8563367128372192, "num_tokens": 257795281.0, "step": 6790 }, { "epoch": 0.8638850019081542, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4441592693328857, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.873754620552063, "num_tokens": 257835743.0, "step": 6791 }, { "epoch": 0.8640122121867447, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.726750373840332, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8753623962402344, "num_tokens": 257878546.0, "step": 6792 }, { "epoch": 0.8641394224653351, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4500317573547363, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8669695854187012, "num_tokens": 257915480.0, "step": 6793 }, { "epoch": 0.8642666327439257, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7467849254608154, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8616223335266113, "num_tokens": 257951215.0, "step": 6794 }, { "epoch": 0.8643938430225162, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.463989019393921, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8574228882789612, "num_tokens": 257987623.0, "step": 6795 }, { "epoch": 0.8645210533011067, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00016689300537109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7352795600891113, "learning_rate": 1e-05, "loss": 0.5056, "mean_token_accuracy": 0.8497051000595093, "num_tokens": 258030078.0, "step": 6796 }, { "epoch": 0.8646482635796973, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4697601795196533, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8730307817459106, "num_tokens": 258064598.0, "step": 6797 }, { "epoch": 0.8647754738582878, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7381608486175537, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.86598140001297, "num_tokens": 258100121.0, "step": 6798 }, { "epoch": 0.8649026841368782, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4726169109344482, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8585948348045349, "num_tokens": 258138398.0, "step": 6799 }, { "epoch": 0.8650298944154687, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.72810697555542, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8618223667144775, "num_tokens": 258178730.0, "step": 6800 }, { "epoch": 0.8651571046940593, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4960899353027344, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8663744330406189, "num_tokens": 258211122.0, "step": 6801 }, { "epoch": 0.8652843149726498, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6973583698272705, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8685890436172485, "num_tokens": 258249753.0, "step": 6802 }, { "epoch": 0.8654115252512403, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.488905906677246, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8597419261932373, "num_tokens": 258287951.0, "step": 6803 }, { "epoch": 0.8655387355298308, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6792891025543213, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8534670472145081, "num_tokens": 258333216.0, "step": 6804 }, { "epoch": 0.8656659458084213, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4921798706054688, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8638879060745239, "num_tokens": 258370380.0, "step": 6805 }, { "epoch": 0.8657931560870118, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.665013551712036, "learning_rate": 1e-05, "loss": 0.5139, "mean_token_accuracy": 0.8505436778068542, "num_tokens": 258413730.0, "step": 6806 }, { "epoch": 0.8659203663656023, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4817728996276855, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8648179173469543, "num_tokens": 258451644.0, "step": 6807 }, { "epoch": 0.8660475766441929, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6498019695281982, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8760291337966919, "num_tokens": 258490084.0, "step": 6808 }, { "epoch": 0.8661747869227834, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4926583766937256, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.844653308391571, "num_tokens": 258530769.0, "step": 6809 }, { "epoch": 0.8663019972013739, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6542513370513916, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.854485273361206, "num_tokens": 258567578.0, "step": 6810 }, { "epoch": 0.8664292074799644, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4778449535369873, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8630673289299011, "num_tokens": 258605498.0, "step": 6811 }, { "epoch": 0.8665564177585549, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6268839836120605, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8799622058868408, "num_tokens": 258646931.0, "step": 6812 }, { "epoch": 0.8666836280371454, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4664430618286133, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8685917854309082, "num_tokens": 258683853.0, "step": 6813 }, { "epoch": 0.8668108383157359, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.632021188735962, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8634955883026123, "num_tokens": 258723706.0, "step": 6814 }, { "epoch": 0.8669380485943264, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.460674524307251, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8655523061752319, "num_tokens": 258762404.0, "step": 6815 }, { "epoch": 0.867065258872917, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6441054344177246, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8508463501930237, "num_tokens": 258803209.0, "step": 6816 }, { "epoch": 0.8671924691515075, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.465641975402832, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8613216876983643, "num_tokens": 258840001.0, "step": 6817 }, { "epoch": 0.8673196794300979, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.647219181060791, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8807051777839661, "num_tokens": 258871401.0, "step": 6818 }, { "epoch": 0.8674468897086884, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4792020320892334, "learning_rate": 1e-05, "loss": 0.5086, "mean_token_accuracy": 0.8477091789245605, "num_tokens": 258902898.0, "step": 6819 }, { "epoch": 0.867574099987279, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.630093574523926, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.866752564907074, "num_tokens": 258942136.0, "step": 6820 }, { "epoch": 0.8677013102658695, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.471039295196533, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.863292932510376, "num_tokens": 258976599.0, "step": 6821 }, { "epoch": 0.86782852054446, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6420328617095947, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8596823215484619, "num_tokens": 259010457.0, "step": 6822 }, { "epoch": 0.8679557308230506, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4629530906677246, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8619972467422485, "num_tokens": 259046368.0, "step": 6823 }, { "epoch": 0.868082941101641, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6323201656341553, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8697901964187622, "num_tokens": 259087240.0, "step": 6824 }, { "epoch": 0.8682101513802315, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.450181484222412, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8638335466384888, "num_tokens": 259128297.0, "step": 6825 }, { "epoch": 0.868337361658822, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.641779899597168, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8602994084358215, "num_tokens": 259170242.0, "step": 6826 }, { "epoch": 0.8684645719374126, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.456606864929199, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8612263798713684, "num_tokens": 259208580.0, "step": 6827 }, { "epoch": 0.8685917822160031, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6614482402801514, "learning_rate": 1e-05, "loss": 0.5309, "mean_token_accuracy": 0.8433194756507874, "num_tokens": 259249721.0, "step": 6828 }, { "epoch": 0.8687189924945936, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4343152046203613, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8690876364707947, "num_tokens": 259292056.0, "step": 6829 }, { "epoch": 0.868846202773184, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.686258554458618, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8710785508155823, "num_tokens": 259323390.0, "step": 6830 }, { "epoch": 0.8689734130517746, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.432175874710083, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.872818112373352, "num_tokens": 259360728.0, "step": 6831 }, { "epoch": 0.8691006233303651, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.70306134223938, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8606288433074951, "num_tokens": 259397619.0, "step": 6832 }, { "epoch": 0.8692278336089556, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.436289072036743, "learning_rate": 1e-05, "loss": 0.5017, "mean_token_accuracy": 0.8486266136169434, "num_tokens": 259440049.0, "step": 6833 }, { "epoch": 0.8693550438875461, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.724097967147827, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8679945468902588, "num_tokens": 259475657.0, "step": 6834 }, { "epoch": 0.8694822541661367, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.458385467529297, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8399925231933594, "num_tokens": 259511289.0, "step": 6835 }, { "epoch": 0.8696094644447271, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.719499349594116, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8754527568817139, "num_tokens": 259546704.0, "step": 6836 }, { "epoch": 0.8697366747233176, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4460289478302, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8741223812103271, "num_tokens": 259588490.0, "step": 6837 }, { "epoch": 0.8698638850019081, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7444586753845215, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8618091344833374, "num_tokens": 259619838.0, "step": 6838 }, { "epoch": 0.8699910952804987, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.456833600997925, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8536622524261475, "num_tokens": 259662232.0, "step": 6839 }, { "epoch": 0.8701183055590892, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.751296043395996, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.8397778272628784, "num_tokens": 259699824.0, "step": 6840 }, { "epoch": 0.8702455158376797, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4948174953460693, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8608481884002686, "num_tokens": 259728217.0, "step": 6841 }, { "epoch": 0.8703727261162701, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7210652828216553, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8768193125724792, "num_tokens": 259763522.0, "step": 6842 }, { "epoch": 0.8704999363948607, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.495499849319458, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8518940210342407, "num_tokens": 259797650.0, "step": 6843 }, { "epoch": 0.8706271466734512, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.697474241256714, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8522584438323975, "num_tokens": 259837265.0, "step": 6844 }, { "epoch": 0.8707543569520417, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.494126796722412, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8647669553756714, "num_tokens": 259873040.0, "step": 6845 }, { "epoch": 0.8708815672306323, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6897523403167725, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8470561504364014, "num_tokens": 259907627.0, "step": 6846 }, { "epoch": 0.8710087775092228, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4863569736480713, "learning_rate": 1e-05, "loss": 0.5265, "mean_token_accuracy": 0.8427786827087402, "num_tokens": 259948213.0, "step": 6847 }, { "epoch": 0.8711359877878132, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.673112630844116, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8563000559806824, "num_tokens": 259985239.0, "step": 6848 }, { "epoch": 0.8712631980664037, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4908344745635986, "learning_rate": 1e-05, "loss": 0.5553, "mean_token_accuracy": 0.833071768283844, "num_tokens": 260023046.0, "step": 6849 }, { "epoch": 0.8713904083449943, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6589324474334717, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8715184926986694, "num_tokens": 260060716.0, "step": 6850 }, { "epoch": 0.8715176186235848, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4836177825927734, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.880963146686554, "num_tokens": 260093939.0, "step": 6851 }, { "epoch": 0.8716448289021753, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6578898429870605, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.862524151802063, "num_tokens": 260129282.0, "step": 6852 }, { "epoch": 0.8717720391807658, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4545555114746094, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8769063949584961, "num_tokens": 260172167.0, "step": 6853 }, { "epoch": 0.8718992494593563, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6693902015686035, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8688340187072754, "num_tokens": 260212613.0, "step": 6854 }, { "epoch": 0.8720264597379468, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.465470314025879, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8690810203552246, "num_tokens": 260248774.0, "step": 6855 }, { "epoch": 0.8721536700165373, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6720645427703857, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8446296453475952, "num_tokens": 260291105.0, "step": 6856 }, { "epoch": 0.8722808802951278, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.453946113586426, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8695605397224426, "num_tokens": 260332316.0, "step": 6857 }, { "epoch": 0.8724080905737184, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6792070865631104, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.872683048248291, "num_tokens": 260365016.0, "step": 6858 }, { "epoch": 0.8725353008523089, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.46795392036438, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.8813179731369019, "num_tokens": 260396131.0, "step": 6859 }, { "epoch": 0.8726625111308994, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.676070213317871, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8480111956596375, "num_tokens": 260438053.0, "step": 6860 }, { "epoch": 0.8727897214094898, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4574055671691895, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8597455024719238, "num_tokens": 260479523.0, "step": 6861 }, { "epoch": 0.8729169316880804, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6749746799468994, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.868985116481781, "num_tokens": 260516623.0, "step": 6862 }, { "epoch": 0.8730441419666709, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4647860527038574, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8691589832305908, "num_tokens": 260552499.0, "step": 6863 }, { "epoch": 0.8731713522452614, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6673245429992676, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8645705580711365, "num_tokens": 260595636.0, "step": 6864 }, { "epoch": 0.873298562523852, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4515907764434814, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8684972524642944, "num_tokens": 260639787.0, "step": 6865 }, { "epoch": 0.8734257728024425, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.679995536804199, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.875707745552063, "num_tokens": 260674117.0, "step": 6866 }, { "epoch": 0.8735529830810329, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4621713161468506, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8637344241142273, "num_tokens": 260712757.0, "step": 6867 }, { "epoch": 0.8736801933596234, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.683931589126587, "learning_rate": 1e-05, "loss": 0.5203, "mean_token_accuracy": 0.8436931371688843, "num_tokens": 260750756.0, "step": 6868 }, { "epoch": 0.873807403638214, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.480273723602295, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8582047820091248, "num_tokens": 260782929.0, "step": 6869 }, { "epoch": 0.8739346139168045, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.668736457824707, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8626696467399597, "num_tokens": 260819762.0, "step": 6870 }, { "epoch": 0.874061824195395, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.457075595855713, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8554272055625916, "num_tokens": 260862543.0, "step": 6871 }, { "epoch": 0.8741890344739855, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6792047023773193, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.8498304486274719, "num_tokens": 260899835.0, "step": 6872 }, { "epoch": 0.874316244752576, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.464782476425171, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8494920134544373, "num_tokens": 260940593.0, "step": 6873 }, { "epoch": 0.8744434550311665, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.681610107421875, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8614206314086914, "num_tokens": 260977454.0, "step": 6874 }, { "epoch": 0.874570665309757, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4602835178375244, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8633185029029846, "num_tokens": 261019095.0, "step": 6875 }, { "epoch": 0.8746978755883476, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.686917304992676, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8530677556991577, "num_tokens": 261056428.0, "step": 6876 }, { "epoch": 0.8748250858669381, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4662270545959473, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8746110200881958, "num_tokens": 261090679.0, "step": 6877 }, { "epoch": 0.8749522961455286, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.680426836013794, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8589719533920288, "num_tokens": 261130640.0, "step": 6878 }, { "epoch": 0.875079506424119, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4679858684539795, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8752497434616089, "num_tokens": 261169064.0, "step": 6879 }, { "epoch": 0.8752067167027096, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6800029277801514, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8645386695861816, "num_tokens": 261205625.0, "step": 6880 }, { "epoch": 0.8753339269813001, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.450273036956787, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8704670667648315, "num_tokens": 261251859.0, "step": 6881 }, { "epoch": 0.8754611372598906, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6880109310150146, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8631500601768494, "num_tokens": 261289604.0, "step": 6882 }, { "epoch": 0.8755883475384811, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.451474189758301, "learning_rate": 1e-05, "loss": 0.3871, "mean_token_accuracy": 0.8854336738586426, "num_tokens": 261330053.0, "step": 6883 }, { "epoch": 0.8757155578170717, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6916720867156982, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8544016480445862, "num_tokens": 261371042.0, "step": 6884 }, { "epoch": 0.8758427680956621, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4754199981689453, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8548898696899414, "num_tokens": 261406911.0, "step": 6885 }, { "epoch": 0.8759699783742526, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6845221519470215, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8657510280609131, "num_tokens": 261443856.0, "step": 6886 }, { "epoch": 0.8760971886528431, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4646353721618652, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8572384119033813, "num_tokens": 261485987.0, "step": 6887 }, { "epoch": 0.8762243989314337, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6936051845550537, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.856429934501648, "num_tokens": 261519888.0, "step": 6888 }, { "epoch": 0.8763516092100242, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.459852457046509, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8631497621536255, "num_tokens": 261557435.0, "step": 6889 }, { "epoch": 0.8764788194886147, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.688730478286743, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8611688613891602, "num_tokens": 261598209.0, "step": 6890 }, { "epoch": 0.8766060297672051, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4696898460388184, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8498356342315674, "num_tokens": 261637972.0, "step": 6891 }, { "epoch": 0.8767332400457957, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.690524101257324, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8493620157241821, "num_tokens": 261676717.0, "step": 6892 }, { "epoch": 0.8768604503243862, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4924752712249756, "learning_rate": 1e-05, "loss": 0.5198, "mean_token_accuracy": 0.8440544009208679, "num_tokens": 261711434.0, "step": 6893 }, { "epoch": 0.8769876606029767, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.664031982421875, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8644355535507202, "num_tokens": 261749527.0, "step": 6894 }, { "epoch": 0.8771148708815673, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4659104347229004, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8715044260025024, "num_tokens": 261792031.0, "step": 6895 }, { "epoch": 0.8772420811601578, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.669070243835449, "learning_rate": 1e-05, "loss": 0.5258, "mean_token_accuracy": 0.8452783823013306, "num_tokens": 261831862.0, "step": 6896 }, { "epoch": 0.8773692914387482, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.468296766281128, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8675118684768677, "num_tokens": 261869949.0, "step": 6897 }, { "epoch": 0.8774965017173387, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.674250364303589, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.864623486995697, "num_tokens": 261904080.0, "step": 6898 }, { "epoch": 0.8776237119959293, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.491391897201538, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8500754237174988, "num_tokens": 261932809.0, "step": 6899 }, { "epoch": 0.8777509222745198, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6448822021484375, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8613423109054565, "num_tokens": 261974313.0, "step": 6900 }, { "epoch": 0.8778781325531103, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4651238918304443, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8761616349220276, "num_tokens": 262011505.0, "step": 6901 }, { "epoch": 0.8780053428317008, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6643781661987305, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8665198087692261, "num_tokens": 262042571.0, "step": 6902 }, { "epoch": 0.8781325531102913, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4569363594055176, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8541646003723145, "num_tokens": 262088038.0, "step": 6903 }, { "epoch": 0.8782597633888818, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.655158281326294, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8721223473548889, "num_tokens": 262126662.0, "step": 6904 }, { "epoch": 0.8783869736674723, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4586644172668457, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8650317788124084, "num_tokens": 262170525.0, "step": 6905 }, { "epoch": 0.8785141839460628, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.678243637084961, "learning_rate": 1e-05, "loss": 0.5225, "mean_token_accuracy": 0.8484099507331848, "num_tokens": 262210524.0, "step": 6906 }, { "epoch": 0.8786413942246534, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4679784774780273, "learning_rate": 1e-05, "loss": 0.5348, "mean_token_accuracy": 0.8439584970474243, "num_tokens": 262248636.0, "step": 6907 }, { "epoch": 0.8787686045032439, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6712729930877686, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8512163162231445, "num_tokens": 262291795.0, "step": 6908 }, { "epoch": 0.8788958147818343, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.456721305847168, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8661634922027588, "num_tokens": 262328484.0, "step": 6909 }, { "epoch": 0.8790230250604248, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.683725357055664, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8676846027374268, "num_tokens": 262362262.0, "step": 6910 }, { "epoch": 0.8791502353390154, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.461031436920166, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8541437387466431, "num_tokens": 262399082.0, "step": 6911 }, { "epoch": 0.8792774456176059, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7027854919433594, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8632231950759888, "num_tokens": 262434968.0, "step": 6912 }, { "epoch": 0.8794046558961964, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.453580141067505, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8663697242736816, "num_tokens": 262472989.0, "step": 6913 }, { "epoch": 0.879531866174787, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6930532455444336, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8628352880477905, "num_tokens": 262512526.0, "step": 6914 }, { "epoch": 0.8796590764533775, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4562947750091553, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8693232536315918, "num_tokens": 262548082.0, "step": 6915 }, { "epoch": 0.8797862867319679, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.766291856765747, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8460234999656677, "num_tokens": 262584786.0, "step": 6916 }, { "epoch": 0.8799134970105584, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4465854167938232, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8670707941055298, "num_tokens": 262625898.0, "step": 6917 }, { "epoch": 0.880040707289149, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.734644651412964, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8635501861572266, "num_tokens": 262664041.0, "step": 6918 }, { "epoch": 0.8801679175677395, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001678466796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4708330631256104, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8523043990135193, "num_tokens": 262702597.0, "step": 6919 }, { "epoch": 0.88029512784633, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.747074842453003, "learning_rate": 1e-05, "loss": 0.53, "mean_token_accuracy": 0.8437235951423645, "num_tokens": 262734951.0, "step": 6920 }, { "epoch": 0.8804223381249205, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4754838943481445, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8692209720611572, "num_tokens": 262771513.0, "step": 6921 }, { "epoch": 0.880549548403511, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7270069122314453, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8548927307128906, "num_tokens": 262813045.0, "step": 6922 }, { "epoch": 0.8806767586821015, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4873785972595215, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8720887303352356, "num_tokens": 262846663.0, "step": 6923 }, { "epoch": 0.880803968960692, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7236239910125732, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.863566517829895, "num_tokens": 262880056.0, "step": 6924 }, { "epoch": 0.8809311792392825, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.482121229171753, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8697659969329834, "num_tokens": 262914819.0, "step": 6925 }, { "epoch": 0.8810583895178731, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7133073806762695, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8584470152854919, "num_tokens": 262953167.0, "step": 6926 }, { "epoch": 0.8811855997964636, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.49351167678833, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8565582633018494, "num_tokens": 262991628.0, "step": 6927 }, { "epoch": 0.881312810075054, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.69769024848938, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8670279383659363, "num_tokens": 263029287.0, "step": 6928 }, { "epoch": 0.8814400203536445, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4923288822174072, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.862083911895752, "num_tokens": 263068982.0, "step": 6929 }, { "epoch": 0.8815672306322351, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6918768882751465, "learning_rate": 1e-05, "loss": 0.5535, "mean_token_accuracy": 0.8394150733947754, "num_tokens": 263111260.0, "step": 6930 }, { "epoch": 0.8816944409108256, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.478797674179077, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8646183013916016, "num_tokens": 263149310.0, "step": 6931 }, { "epoch": 0.8818216511894161, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6727898120880127, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8737927675247192, "num_tokens": 263188440.0, "step": 6932 }, { "epoch": 0.8819488614680067, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4764492511749268, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8528642058372498, "num_tokens": 263229079.0, "step": 6933 }, { "epoch": 0.8820760717465971, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.692992687225342, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8560717105865479, "num_tokens": 263260591.0, "step": 6934 }, { "epoch": 0.8822032820251876, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4728732109069824, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8507357239723206, "num_tokens": 263301256.0, "step": 6935 }, { "epoch": 0.8823304923037781, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6729228496551514, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8681554794311523, "num_tokens": 263337600.0, "step": 6936 }, { "epoch": 0.8824577025823687, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.464024782180786, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8579011559486389, "num_tokens": 263383124.0, "step": 6937 }, { "epoch": 0.8825849128609592, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6845321655273438, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8710159063339233, "num_tokens": 263416047.0, "step": 6938 }, { "epoch": 0.8827121231395497, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4564208984375, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8745942115783691, "num_tokens": 263458515.0, "step": 6939 }, { "epoch": 0.8828393334181401, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.68869948387146, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8537306189537048, "num_tokens": 263499833.0, "step": 6940 }, { "epoch": 0.8829665436967307, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.488187074661255, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8593714237213135, "num_tokens": 263531680.0, "step": 6941 }, { "epoch": 0.8830937539753212, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.684574604034424, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8492028713226318, "num_tokens": 263565889.0, "step": 6942 }, { "epoch": 0.8832209642539117, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4788055419921875, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8550268411636353, "num_tokens": 263604974.0, "step": 6943 }, { "epoch": 0.8833481745325023, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.665954828262329, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8624981045722961, "num_tokens": 263646960.0, "step": 6944 }, { "epoch": 0.8834753848110928, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.480043411254883, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8602840900421143, "num_tokens": 263686427.0, "step": 6945 }, { "epoch": 0.8836025950896832, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.663994312286377, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8725866079330444, "num_tokens": 263717520.0, "step": 6946 }, { "epoch": 0.8837298053682737, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4819536209106445, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8589975833892822, "num_tokens": 263756226.0, "step": 6947 }, { "epoch": 0.8838570156468643, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6509664058685303, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8730422258377075, "num_tokens": 263793961.0, "step": 6948 }, { "epoch": 0.8839842259254548, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4798455238342285, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8592180013656616, "num_tokens": 263827325.0, "step": 6949 }, { "epoch": 0.8841114362040453, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.657778024673462, "learning_rate": 1e-05, "loss": 0.525, "mean_token_accuracy": 0.8408240675926208, "num_tokens": 263862854.0, "step": 6950 }, { "epoch": 0.8842386464826358, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4645135402679443, "learning_rate": 1e-05, "loss": 0.5448, "mean_token_accuracy": 0.8365482091903687, "num_tokens": 263907035.0, "step": 6951 }, { "epoch": 0.8843658567612263, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6555020809173584, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8513610363006592, "num_tokens": 263949550.0, "step": 6952 }, { "epoch": 0.8844930670398168, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4630839824676514, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8620361089706421, "num_tokens": 263988966.0, "step": 6953 }, { "epoch": 0.8846202773184073, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.662970542907715, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.875550389289856, "num_tokens": 264023442.0, "step": 6954 }, { "epoch": 0.8847474875969978, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.444568634033203, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8714375495910645, "num_tokens": 264064185.0, "step": 6955 }, { "epoch": 0.8848746978755884, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6717307567596436, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8531174659729004, "num_tokens": 264107455.0, "step": 6956 }, { "epoch": 0.8850019081541789, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4597747325897217, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8551689386367798, "num_tokens": 264143090.0, "step": 6957 }, { "epoch": 0.8851291184327693, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6718931198120117, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8582487106323242, "num_tokens": 264180002.0, "step": 6958 }, { "epoch": 0.8852563287113598, "ewc_loss": 0.052490234375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4634037017822266, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8706341981887817, "num_tokens": 264215444.0, "step": 6959 }, { "epoch": 0.8853835389899504, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.671299457550049, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8619667291641235, "num_tokens": 264250758.0, "step": 6960 }, { "epoch": 0.8855107492685409, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.472837448120117, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8482183814048767, "num_tokens": 264290181.0, "step": 6961 }, { "epoch": 0.8856379595471314, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6780335903167725, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8629281520843506, "num_tokens": 264327480.0, "step": 6962 }, { "epoch": 0.885765169825722, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4524102210998535, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8632408976554871, "num_tokens": 264368762.0, "step": 6963 }, { "epoch": 0.8858923801043125, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6763203144073486, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8623244762420654, "num_tokens": 264410261.0, "step": 6964 }, { "epoch": 0.8860195903829029, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4552595615386963, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.867579460144043, "num_tokens": 264449111.0, "step": 6965 }, { "epoch": 0.8861468006614934, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.680040121078491, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8566844463348389, "num_tokens": 264487045.0, "step": 6966 }, { "epoch": 0.886274010940084, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4673829078674316, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8641889095306396, "num_tokens": 264524025.0, "step": 6967 }, { "epoch": 0.8864012212186745, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6864688396453857, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8544672727584839, "num_tokens": 264561361.0, "step": 6968 }, { "epoch": 0.886528431497265, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4678215980529785, "learning_rate": 1e-05, "loss": 0.54, "mean_token_accuracy": 0.8364854454994202, "num_tokens": 264603557.0, "step": 6969 }, { "epoch": 0.8866556417758555, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6847898960113525, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8743600249290466, "num_tokens": 264638539.0, "step": 6970 }, { "epoch": 0.886782852054446, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4602367877960205, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8635912537574768, "num_tokens": 264676592.0, "step": 6971 }, { "epoch": 0.8869100623330365, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6866202354431152, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.873022198677063, "num_tokens": 264708339.0, "step": 6972 }, { "epoch": 0.887037272611627, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.477046489715576, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.861314594745636, "num_tokens": 264741705.0, "step": 6973 }, { "epoch": 0.8871644828902175, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6897268295288086, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8599498271942139, "num_tokens": 264773095.0, "step": 6974 }, { "epoch": 0.8872916931688081, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4558863639831543, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8786944150924683, "num_tokens": 264813499.0, "step": 6975 }, { "epoch": 0.8874189034473986, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.7264063358306885, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8687089681625366, "num_tokens": 264851314.0, "step": 6976 }, { "epoch": 0.887546113725989, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4627833366394043, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8637164831161499, "num_tokens": 264886908.0, "step": 6977 }, { "epoch": 0.8876733240045795, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.714538812637329, "learning_rate": 1e-05, "loss": 0.5474, "mean_token_accuracy": 0.8360464572906494, "num_tokens": 264927516.0, "step": 6978 }, { "epoch": 0.8878005342831701, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.473778009414673, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8677815198898315, "num_tokens": 264966758.0, "step": 6979 }, { "epoch": 0.8879277445617606, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.711923360824585, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8527370691299438, "num_tokens": 265002267.0, "step": 6980 }, { "epoch": 0.8880549548403511, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.480086326599121, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8690413236618042, "num_tokens": 265034957.0, "step": 6981 }, { "epoch": 0.8881821651189417, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.701007604598999, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8628404140472412, "num_tokens": 265075587.0, "step": 6982 }, { "epoch": 0.8883093753975321, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4698336124420166, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8746223449707031, "num_tokens": 265111385.0, "step": 6983 }, { "epoch": 0.8884365856761226, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6870672702789307, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8721181154251099, "num_tokens": 265153674.0, "step": 6984 }, { "epoch": 0.8885637959547131, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4802305698394775, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8745059967041016, "num_tokens": 265190168.0, "step": 6985 }, { "epoch": 0.8886910062333037, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6742706298828125, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8733905553817749, "num_tokens": 265230145.0, "step": 6986 }, { "epoch": 0.8888182165118942, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4953293800354004, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8501236438751221, "num_tokens": 265263229.0, "step": 6987 }, { "epoch": 0.8889454267904847, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6598098278045654, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8539527654647827, "num_tokens": 265305426.0, "step": 6988 }, { "epoch": 0.8890726370690751, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.480652332305908, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8527026176452637, "num_tokens": 265345677.0, "step": 6989 }, { "epoch": 0.8891998473476657, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6494081020355225, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8716917634010315, "num_tokens": 265383987.0, "step": 6990 }, { "epoch": 0.8893270576262562, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4825918674468994, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8440786004066467, "num_tokens": 265423209.0, "step": 6991 }, { "epoch": 0.8894542679048467, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6455600261688232, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8519642353057861, "num_tokens": 265463304.0, "step": 6992 }, { "epoch": 0.8895814781834372, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00016880035400390625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4626336097717285, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8525068759918213, "num_tokens": 265506991.0, "step": 6993 }, { "epoch": 0.8897086884620278, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6640076637268066, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8634752631187439, "num_tokens": 265539050.0, "step": 6994 }, { "epoch": 0.8898358987406182, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.454521894454956, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8601610660552979, "num_tokens": 265583593.0, "step": 6995 }, { "epoch": 0.8899631090192087, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6684319972991943, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8431472778320312, "num_tokens": 265623873.0, "step": 6996 }, { "epoch": 0.8900903192977992, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4483232498168945, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8571395874023438, "num_tokens": 265666191.0, "step": 6997 }, { "epoch": 0.8902175295763898, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.678784132003784, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8586532473564148, "num_tokens": 265703929.0, "step": 6998 }, { "epoch": 0.8903447398549803, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4613752365112305, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8643453121185303, "num_tokens": 265737246.0, "step": 6999 }, { "epoch": 0.8904719501335708, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6898117065429688, "learning_rate": 1e-05, "loss": 0.5119, "mean_token_accuracy": 0.8526133894920349, "num_tokens": 265774549.0, "step": 7000 }, { "epoch": 0.8905991604121613, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.469743490219116, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.8561145067214966, "num_tokens": 265806972.0, "step": 7001 }, { "epoch": 0.8907263706907518, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6808273792266846, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8531211614608765, "num_tokens": 265844505.0, "step": 7002 }, { "epoch": 0.8908535809693423, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4550211429595947, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8629456758499146, "num_tokens": 265885350.0, "step": 7003 }, { "epoch": 0.8909807912479328, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.676314115524292, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8660131096839905, "num_tokens": 265925345.0, "step": 7004 }, { "epoch": 0.8911080015265234, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.456648588180542, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8665913939476013, "num_tokens": 265961329.0, "step": 7005 }, { "epoch": 0.8912352118051139, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.7010087966918945, "learning_rate": 1e-05, "loss": 0.5545, "mean_token_accuracy": 0.8351008892059326, "num_tokens": 266000836.0, "step": 7006 }, { "epoch": 0.8913624220837043, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.469592571258545, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8562504053115845, "num_tokens": 266037178.0, "step": 7007 }, { "epoch": 0.8914896323622948, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.686903715133667, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8697971701622009, "num_tokens": 266077125.0, "step": 7008 }, { "epoch": 0.8916168426408854, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.463169813156128, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8459287881851196, "num_tokens": 266117313.0, "step": 7009 }, { "epoch": 0.8917440529194759, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6914680004119873, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8617170453071594, "num_tokens": 266157432.0, "step": 7010 }, { "epoch": 0.8918712631980664, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4626474380493164, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8610574007034302, "num_tokens": 266197170.0, "step": 7011 }, { "epoch": 0.891998473476657, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6976263523101807, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.853605329990387, "num_tokens": 266230383.0, "step": 7012 }, { "epoch": 0.8921256837552475, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4661612510681152, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.862567126750946, "num_tokens": 266268685.0, "step": 7013 }, { "epoch": 0.8922528940338379, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6904377937316895, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8584076762199402, "num_tokens": 266308875.0, "step": 7014 }, { "epoch": 0.8923801043124284, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4841442108154297, "learning_rate": 1e-05, "loss": 0.5457, "mean_token_accuracy": 0.8351256847381592, "num_tokens": 266346540.0, "step": 7015 }, { "epoch": 0.892507314591019, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.689793825149536, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8623189330101013, "num_tokens": 266383359.0, "step": 7016 }, { "epoch": 0.8926345248696095, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.482179880142212, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.855829119682312, "num_tokens": 266420235.0, "step": 7017 }, { "epoch": 0.8927617351482, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6914405822753906, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8588813543319702, "num_tokens": 266455097.0, "step": 7018 }, { "epoch": 0.8928889454267905, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.477794647216797, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8649864196777344, "num_tokens": 266488471.0, "step": 7019 }, { "epoch": 0.893016155705381, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.682039260864258, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8559457063674927, "num_tokens": 266528107.0, "step": 7020 }, { "epoch": 0.8931433659839715, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.469684362411499, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8686087131500244, "num_tokens": 266567859.0, "step": 7021 }, { "epoch": 0.893270576262562, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6692168712615967, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8683974146842957, "num_tokens": 266609722.0, "step": 7022 }, { "epoch": 0.8933977865411525, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4567575454711914, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8741761445999146, "num_tokens": 266650546.0, "step": 7023 }, { "epoch": 0.8935249968197431, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.69246768951416, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8585190176963806, "num_tokens": 266686508.0, "step": 7024 }, { "epoch": 0.8936522070983336, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.458496332168579, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8763315081596375, "num_tokens": 266725528.0, "step": 7025 }, { "epoch": 0.893779417376924, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7134017944335938, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8763583898544312, "num_tokens": 266754818.0, "step": 7026 }, { "epoch": 0.8939066276555145, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.467766046524048, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.859705924987793, "num_tokens": 266790723.0, "step": 7027 }, { "epoch": 0.8940338379341051, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.706423759460449, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8644248247146606, "num_tokens": 266824889.0, "step": 7028 }, { "epoch": 0.8941610482126956, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4581871032714844, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8607720732688904, "num_tokens": 266866129.0, "step": 7029 }, { "epoch": 0.8942882584912861, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.703453779220581, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8593728542327881, "num_tokens": 266910005.0, "step": 7030 }, { "epoch": 0.8944154687698767, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4644558429718018, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8689906001091003, "num_tokens": 266947366.0, "step": 7031 }, { "epoch": 0.8945426790484671, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7136447429656982, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8577650785446167, "num_tokens": 266982402.0, "step": 7032 }, { "epoch": 0.8946698893270576, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.493595600128174, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8512191772460938, "num_tokens": 267015397.0, "step": 7033 }, { "epoch": 0.8947970996056481, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.697024345397949, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.875383734703064, "num_tokens": 267050204.0, "step": 7034 }, { "epoch": 0.8949243098842387, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.473123073577881, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8766580820083618, "num_tokens": 267085127.0, "step": 7035 }, { "epoch": 0.8950515201628292, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6954994201660156, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8516067266464233, "num_tokens": 267122210.0, "step": 7036 }, { "epoch": 0.8951787304414197, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.484100103378296, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8598612546920776, "num_tokens": 267161124.0, "step": 7037 }, { "epoch": 0.8953059407200101, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6892459392547607, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8716585636138916, "num_tokens": 267195339.0, "step": 7038 }, { "epoch": 0.8954331509986007, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.484700918197632, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8564284443855286, "num_tokens": 267228712.0, "step": 7039 }, { "epoch": 0.8955603612771912, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6931111812591553, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.851793646812439, "num_tokens": 267262796.0, "step": 7040 }, { "epoch": 0.8956875715557817, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4882116317749023, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8547037839889526, "num_tokens": 267304101.0, "step": 7041 }, { "epoch": 0.8958147818343722, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.665321111679077, "learning_rate": 1e-05, "loss": 0.5211, "mean_token_accuracy": 0.8462560176849365, "num_tokens": 267348943.0, "step": 7042 }, { "epoch": 0.8959419921129628, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.470977544784546, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8675728440284729, "num_tokens": 267390436.0, "step": 7043 }, { "epoch": 0.8960692023915532, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.665010690689087, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8735134601593018, "num_tokens": 267428069.0, "step": 7044 }, { "epoch": 0.8961964126701437, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.482128858566284, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8477298021316528, "num_tokens": 267468029.0, "step": 7045 }, { "epoch": 0.8963236229487342, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6640119552612305, "learning_rate": 1e-05, "loss": 0.4905, "mean_token_accuracy": 0.8551445007324219, "num_tokens": 267504013.0, "step": 7046 }, { "epoch": 0.8964508332273248, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4831178188323975, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8699824810028076, "num_tokens": 267537696.0, "step": 7047 }, { "epoch": 0.8965780435059153, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6585195064544678, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8690779805183411, "num_tokens": 267568971.0, "step": 7048 }, { "epoch": 0.8967052537845058, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.473828077316284, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8580318689346313, "num_tokens": 267607829.0, "step": 7049 }, { "epoch": 0.8968324640630962, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6547555923461914, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8650739192962646, "num_tokens": 267649338.0, "step": 7050 }, { "epoch": 0.8969596743416868, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4784679412841797, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8683586120605469, "num_tokens": 267685053.0, "step": 7051 }, { "epoch": 0.8970868846202773, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6681199073791504, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8682419657707214, "num_tokens": 267722291.0, "step": 7052 }, { "epoch": 0.8972140948988678, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.469794511795044, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.863280177116394, "num_tokens": 267760533.0, "step": 7053 }, { "epoch": 0.8973413051774584, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6634342670440674, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8646067380905151, "num_tokens": 267796655.0, "step": 7054 }, { "epoch": 0.8974685154560489, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4496335983276367, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.865280032157898, "num_tokens": 267835950.0, "step": 7055 }, { "epoch": 0.8975957257346393, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.672590732574463, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8658039569854736, "num_tokens": 267871796.0, "step": 7056 }, { "epoch": 0.8977229360132298, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4614741802215576, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8657225370407104, "num_tokens": 267910562.0, "step": 7057 }, { "epoch": 0.8978501462918204, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6728687286376953, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8646694421768188, "num_tokens": 267948962.0, "step": 7058 }, { "epoch": 0.8979773565704109, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4587936401367188, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8542846441268921, "num_tokens": 267993158.0, "step": 7059 }, { "epoch": 0.8981045668490014, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.685816764831543, "learning_rate": 1e-05, "loss": 0.4971, "mean_token_accuracy": 0.8530104756355286, "num_tokens": 268033366.0, "step": 7060 }, { "epoch": 0.898231777127592, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4529592990875244, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8631669282913208, "num_tokens": 268074125.0, "step": 7061 }, { "epoch": 0.8983589874061825, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.692351818084717, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8583213090896606, "num_tokens": 268112132.0, "step": 7062 }, { "epoch": 0.8984861976847729, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.481656074523926, "learning_rate": 1e-05, "loss": 0.5303, "mean_token_accuracy": 0.8395177125930786, "num_tokens": 268148161.0, "step": 7063 }, { "epoch": 0.8986134079633634, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6704137325286865, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.885675847530365, "num_tokens": 268184977.0, "step": 7064 }, { "epoch": 0.898740618241954, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.478090524673462, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8625022172927856, "num_tokens": 268219355.0, "step": 7065 }, { "epoch": 0.8988678285205445, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6580541133880615, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8662139177322388, "num_tokens": 268260533.0, "step": 7066 }, { "epoch": 0.898995038799135, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4900245666503906, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.857576310634613, "num_tokens": 268294415.0, "step": 7067 }, { "epoch": 0.8991222490777255, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.64758563041687, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8692201972007751, "num_tokens": 268331038.0, "step": 7068 }, { "epoch": 0.899249459356316, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.48861026763916, "learning_rate": 1e-05, "loss": 0.5087, "mean_token_accuracy": 0.8454980850219727, "num_tokens": 268369438.0, "step": 7069 }, { "epoch": 0.8993766696349065, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.648171901702881, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8640215396881104, "num_tokens": 268403808.0, "step": 7070 }, { "epoch": 0.899503879913497, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4723753929138184, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8750834465026855, "num_tokens": 268442575.0, "step": 7071 }, { "epoch": 0.8996310901920875, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6519064903259277, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8606356978416443, "num_tokens": 268475387.0, "step": 7072 }, { "epoch": 0.8997583004706781, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4687509536743164, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.861476719379425, "num_tokens": 268513514.0, "step": 7073 }, { "epoch": 0.8998855107492686, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6475634574890137, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.866497278213501, "num_tokens": 268553690.0, "step": 7074 }, { "epoch": 0.900012721027859, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4541854858398438, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8703371286392212, "num_tokens": 268593285.0, "step": 7075 }, { "epoch": 0.9001399313064495, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6650142669677734, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8637877702713013, "num_tokens": 268631929.0, "step": 7076 }, { "epoch": 0.9002671415850401, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.469672918319702, "learning_rate": 1e-05, "loss": 0.5308, "mean_token_accuracy": 0.8342748880386353, "num_tokens": 268669882.0, "step": 7077 }, { "epoch": 0.9003943518636306, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6598012447357178, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8475983142852783, "num_tokens": 268712709.0, "step": 7078 }, { "epoch": 0.9005215621422211, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.00017070770263671875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.453183174133301, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8591703176498413, "num_tokens": 268751514.0, "step": 7079 }, { "epoch": 0.9006487724208116, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.663054943084717, "learning_rate": 1e-05, "loss": 0.5073, "mean_token_accuracy": 0.8481398820877075, "num_tokens": 268796841.0, "step": 7080 }, { "epoch": 0.9007759826994021, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4408328533172607, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8716960549354553, "num_tokens": 268841536.0, "step": 7081 }, { "epoch": 0.9009031929779926, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6726467609405518, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.861008882522583, "num_tokens": 268884636.0, "step": 7082 }, { "epoch": 0.9010304032565831, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.457064390182495, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8579679131507874, "num_tokens": 268924114.0, "step": 7083 }, { "epoch": 0.9011576135351737, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6730573177337646, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8691791892051697, "num_tokens": 268966476.0, "step": 7084 }, { "epoch": 0.9012848238137642, "ewc_loss": 0.052734375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.458590507507324, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8721129298210144, "num_tokens": 268998224.0, "step": 7085 }, { "epoch": 0.9014120340923547, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.694018602371216, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8555728197097778, "num_tokens": 269043552.0, "step": 7086 }, { "epoch": 0.9015392443709451, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4526453018188477, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8488357067108154, "num_tokens": 269085180.0, "step": 7087 }, { "epoch": 0.9016664546495357, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.699991226196289, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.850806713104248, "num_tokens": 269121694.0, "step": 7088 }, { "epoch": 0.9017936649281262, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4546165466308594, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8451322317123413, "num_tokens": 269164896.0, "step": 7089 }, { "epoch": 0.9019208752067167, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.701805830001831, "learning_rate": 1e-05, "loss": 0.3896, "mean_token_accuracy": 0.8829320073127747, "num_tokens": 269201095.0, "step": 7090 }, { "epoch": 0.9020480854853072, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4530787467956543, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8651063442230225, "num_tokens": 269243055.0, "step": 7091 }, { "epoch": 0.9021752957638978, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.715851068496704, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8699367642402649, "num_tokens": 269279902.0, "step": 7092 }, { "epoch": 0.9023025060424882, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4618678092956543, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8768525719642639, "num_tokens": 269315773.0, "step": 7093 }, { "epoch": 0.9024297163210787, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7222816944122314, "learning_rate": 1e-05, "loss": 0.5268, "mean_token_accuracy": 0.8451356291770935, "num_tokens": 269353127.0, "step": 7094 }, { "epoch": 0.9025569265996692, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4756381511688232, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8631725311279297, "num_tokens": 269387023.0, "step": 7095 }, { "epoch": 0.9026841368782598, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.733266830444336, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8429228067398071, "num_tokens": 269419983.0, "step": 7096 }, { "epoch": 0.9028113471568503, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4609766006469727, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8680245876312256, "num_tokens": 269460580.0, "step": 7097 }, { "epoch": 0.9029385574354408, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7160866260528564, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8774126768112183, "num_tokens": 269497711.0, "step": 7098 }, { "epoch": 0.9030657677140312, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.46508526802063, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8628338575363159, "num_tokens": 269538111.0, "step": 7099 }, { "epoch": 0.9031929779926218, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.727203607559204, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.881482720375061, "num_tokens": 269570695.0, "step": 7100 }, { "epoch": 0.9033201882712123, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 14.20290756225586, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8641780614852905, "num_tokens": 269605849.0, "step": 7101 }, { "epoch": 0.9034473985498028, "ewc_loss": 0.11181640625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 9.489059448242188e-05, "grad_norm": 4.030272006988525, "learning_rate": 1e-05, "loss": 0.5272, "mean_token_accuracy": 0.8583826422691345, "num_tokens": 269646445.0, "step": 7102 }, { "epoch": 0.9035746088283934, "ewc_loss": 0.05078125, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 3.337860107421875e-05, "grad_norm": 2.042250633239746, "learning_rate": 1e-05, "loss": 0.5092, "mean_token_accuracy": 0.8470670580863953, "num_tokens": 269686806.0, "step": 7103 }, { "epoch": 0.9037018191069839, "ewc_loss": 0.0966796875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 7.963180541992188e-05, "grad_norm": 3.882431983947754, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8527712225914001, "num_tokens": 269727311.0, "step": 7104 }, { "epoch": 0.9038290293855743, "ewc_loss": 0.10400390625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 8.678436279296875e-05, "grad_norm": 3.4750149250030518, "learning_rate": 1e-05, "loss": 0.5833, "mean_token_accuracy": 0.8413087129592896, "num_tokens": 269765066.0, "step": 7105 }, { "epoch": 0.9039562396641648, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 6.246566772460938e-05, "grad_norm": 3.0284717082977295, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8547961711883545, "num_tokens": 269807065.0, "step": 7106 }, { "epoch": 0.9040834499427554, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 3.109027862548828, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.877174973487854, "num_tokens": 269839874.0, "step": 7107 }, { "epoch": 0.9042106602213459, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 6.29425048828125e-05, "grad_norm": 3.010143280029297, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8714317083358765, "num_tokens": 269877647.0, "step": 7108 }, { "epoch": 0.9043378704999364, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 5.745887756347656e-05, "grad_norm": 2.905907154083252, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.862665057182312, "num_tokens": 269912403.0, "step": 7109 }, { "epoch": 0.9044650807785269, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000171661376953125, "ewc_loss_parallel": 5.507469177246094e-05, "grad_norm": 2.8695755004882812, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8690698146820068, "num_tokens": 269946086.0, "step": 7110 }, { "epoch": 0.9045922910571175, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 5.1975250244140625e-05, "grad_norm": 2.779601573944092, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8730554580688477, "num_tokens": 269985820.0, "step": 7111 }, { "epoch": 0.9047195013357079, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 2.788426399230957, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8709142208099365, "num_tokens": 270021377.0, "step": 7112 }, { "epoch": 0.9048467116142984, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.6978085041046143, "learning_rate": 1e-05, "loss": 0.5288, "mean_token_accuracy": 0.8484251499176025, "num_tokens": 270056928.0, "step": 7113 }, { "epoch": 0.9049739218928889, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7005534172058105, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8730787038803101, "num_tokens": 270093720.0, "step": 7114 }, { "epoch": 0.9051011321714795, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.621574878692627, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8542739748954773, "num_tokens": 270140151.0, "step": 7115 }, { "epoch": 0.90522834245007, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6646673679351807, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8655257225036621, "num_tokens": 270173970.0, "step": 7116 }, { "epoch": 0.9053555527286605, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.591297149658203, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8638671636581421, "num_tokens": 270207014.0, "step": 7117 }, { "epoch": 0.905482763007251, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.640307903289795, "learning_rate": 1e-05, "loss": 0.5176, "mean_token_accuracy": 0.8480227589607239, "num_tokens": 270246635.0, "step": 7118 }, { "epoch": 0.9056099732858415, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.55267071723938, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8588262796401978, "num_tokens": 270289902.0, "step": 7119 }, { "epoch": 0.905737183564432, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6313016414642334, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8475340008735657, "num_tokens": 270321173.0, "step": 7120 }, { "epoch": 0.9058643938430225, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5155093669891357, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8719518184661865, "num_tokens": 270360281.0, "step": 7121 }, { "epoch": 0.9059916041216131, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.614964723587036, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.869957685470581, "num_tokens": 270396609.0, "step": 7122 }, { "epoch": 0.9061188144002036, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.512739896774292, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8496644496917725, "num_tokens": 270436365.0, "step": 7123 }, { "epoch": 0.906246024678794, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6142380237579346, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8670715093612671, "num_tokens": 270471314.0, "step": 7124 }, { "epoch": 0.9063732349573845, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4906563758850098, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8529945611953735, "num_tokens": 270510169.0, "step": 7125 }, { "epoch": 0.9065004452359751, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6071691513061523, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8687957525253296, "num_tokens": 270549787.0, "step": 7126 }, { "epoch": 0.9066276555145656, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4920244216918945, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8702128529548645, "num_tokens": 270584059.0, "step": 7127 }, { "epoch": 0.9067548657931561, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6086089611053467, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8684539794921875, "num_tokens": 270618059.0, "step": 7128 }, { "epoch": 0.9068820760717466, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.482872247695923, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8618227243423462, "num_tokens": 270656968.0, "step": 7129 }, { "epoch": 0.9070092863503371, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6262056827545166, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8611904978752136, "num_tokens": 270692088.0, "step": 7130 }, { "epoch": 0.9071364966289276, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.471764326095581, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8682990074157715, "num_tokens": 270729480.0, "step": 7131 }, { "epoch": 0.9072637069075181, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6223597526550293, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8707789182662964, "num_tokens": 270769062.0, "step": 7132 }, { "epoch": 0.9073909171861086, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4654455184936523, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.857312798500061, "num_tokens": 270808142.0, "step": 7133 }, { "epoch": 0.9075181274646992, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.639909029006958, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.858487606048584, "num_tokens": 270842232.0, "step": 7134 }, { "epoch": 0.9076453377432897, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.463839530944824, "learning_rate": 1e-05, "loss": 0.5047, "mean_token_accuracy": 0.8512067198753357, "num_tokens": 270881588.0, "step": 7135 }, { "epoch": 0.9077725480218801, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6483213901519775, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8571305274963379, "num_tokens": 270921122.0, "step": 7136 }, { "epoch": 0.9078997583004706, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4515974521636963, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8567532300949097, "num_tokens": 270959037.0, "step": 7137 }, { "epoch": 0.9080269685790612, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6592092514038086, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8648971319198608, "num_tokens": 271002731.0, "step": 7138 }, { "epoch": 0.9081541788576517, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.448777437210083, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8646453022956848, "num_tokens": 271038487.0, "step": 7139 }, { "epoch": 0.9082813891362422, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.700860023498535, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.859128475189209, "num_tokens": 271068811.0, "step": 7140 }, { "epoch": 0.9084085994148328, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.439105749130249, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8751242160797119, "num_tokens": 271111019.0, "step": 7141 }, { "epoch": 0.9085358096934232, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.703218936920166, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.859000027179718, "num_tokens": 271149317.0, "step": 7142 }, { "epoch": 0.9086630199720137, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.443892002105713, "learning_rate": 1e-05, "loss": 0.3956, "mean_token_accuracy": 0.8803732395172119, "num_tokens": 271190664.0, "step": 7143 }, { "epoch": 0.9087902302506042, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7365877628326416, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8674012422561646, "num_tokens": 271225071.0, "step": 7144 }, { "epoch": 0.9089174405291948, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.441028594970703, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8631085157394409, "num_tokens": 271264228.0, "step": 7145 }, { "epoch": 0.9090446508077853, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7545909881591797, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8629664778709412, "num_tokens": 271300888.0, "step": 7146 }, { "epoch": 0.9091718610863758, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.449795722961426, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8621165752410889, "num_tokens": 271339629.0, "step": 7147 }, { "epoch": 0.9092990713649662, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7709476947784424, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8644111156463623, "num_tokens": 271379116.0, "step": 7148 }, { "epoch": 0.9094262816435568, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.464768409729004, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8515621423721313, "num_tokens": 271420829.0, "step": 7149 }, { "epoch": 0.9095534919221473, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7765018939971924, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.859472930431366, "num_tokens": 271457934.0, "step": 7150 }, { "epoch": 0.9096807022007378, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4876155853271484, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8625330328941345, "num_tokens": 271497441.0, "step": 7151 }, { "epoch": 0.9098079124793284, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.762129306793213, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8556864857673645, "num_tokens": 271532358.0, "step": 7152 }, { "epoch": 0.9099351227579189, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4902307987213135, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8673903942108154, "num_tokens": 271570103.0, "step": 7153 }, { "epoch": 0.9100623330365093, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.751570463180542, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8668763637542725, "num_tokens": 271604542.0, "step": 7154 }, { "epoch": 0.9101895433150998, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4942119121551514, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8519070744514465, "num_tokens": 271646901.0, "step": 7155 }, { "epoch": 0.9103167535936904, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7451059818267822, "learning_rate": 1e-05, "loss": 0.5184, "mean_token_accuracy": 0.8434500098228455, "num_tokens": 271684922.0, "step": 7156 }, { "epoch": 0.9104439638722809, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4938108921051025, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8795208930969238, "num_tokens": 271721206.0, "step": 7157 }, { "epoch": 0.9105711741508714, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.724646806716919, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.874558687210083, "num_tokens": 271757978.0, "step": 7158 }, { "epoch": 0.9106983844294619, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5326435565948486, "learning_rate": 1e-05, "loss": 0.5508, "mean_token_accuracy": 0.8376814723014832, "num_tokens": 271792208.0, "step": 7159 }, { "epoch": 0.9108255947080524, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7022793292999268, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8563473224639893, "num_tokens": 271828853.0, "step": 7160 }, { "epoch": 0.9109528049866429, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.520190954208374, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8478764295578003, "num_tokens": 271868596.0, "step": 7161 }, { "epoch": 0.9110800152652334, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6583988666534424, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8679834008216858, "num_tokens": 271911292.0, "step": 7162 }, { "epoch": 0.9112072255438239, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.508791208267212, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8698479533195496, "num_tokens": 271954512.0, "step": 7163 }, { "epoch": 0.9113344358224145, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6608238220214844, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8569951057434082, "num_tokens": 271994229.0, "step": 7164 }, { "epoch": 0.911461646101005, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.496492385864258, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8680860996246338, "num_tokens": 272031712.0, "step": 7165 }, { "epoch": 0.9115888563795955, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.643812894821167, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8581575155258179, "num_tokens": 272070269.0, "step": 7166 }, { "epoch": 0.9117160666581859, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.483426570892334, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8619556427001953, "num_tokens": 272115407.0, "step": 7167 }, { "epoch": 0.9118432769367765, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.637497901916504, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8618111610412598, "num_tokens": 272161463.0, "step": 7168 }, { "epoch": 0.911970487215367, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5072176456451416, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8679133057594299, "num_tokens": 272196712.0, "step": 7169 }, { "epoch": 0.9120976974939575, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6314938068389893, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8572160601615906, "num_tokens": 272235617.0, "step": 7170 }, { "epoch": 0.912224907772548, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5067038536071777, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8667806386947632, "num_tokens": 272266464.0, "step": 7171 }, { "epoch": 0.9123521180511386, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6292314529418945, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8487915992736816, "num_tokens": 272301069.0, "step": 7172 }, { "epoch": 0.912479328329729, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5066118240356445, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8459692001342773, "num_tokens": 272333119.0, "step": 7173 }, { "epoch": 0.9126065386083195, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.636096239089966, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.849628210067749, "num_tokens": 272365560.0, "step": 7174 }, { "epoch": 0.9127337488869101, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.472449541091919, "learning_rate": 1e-05, "loss": 0.3961, "mean_token_accuracy": 0.8805546164512634, "num_tokens": 272401290.0, "step": 7175 }, { "epoch": 0.9128609591655006, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6226556301116943, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8671205043792725, "num_tokens": 272442925.0, "step": 7176 }, { "epoch": 0.9129881694440911, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.477879285812378, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8558799028396606, "num_tokens": 272477192.0, "step": 7177 }, { "epoch": 0.9131153797226816, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6607072353363037, "learning_rate": 1e-05, "loss": 0.5143, "mean_token_accuracy": 0.8455618619918823, "num_tokens": 272506987.0, "step": 7178 }, { "epoch": 0.9132425900012721, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4553451538085938, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8682258129119873, "num_tokens": 272544399.0, "step": 7179 }, { "epoch": 0.9133698002798626, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6648268699645996, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.861426591873169, "num_tokens": 272577639.0, "step": 7180 }, { "epoch": 0.9134970105584531, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4513590335845947, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8673946261405945, "num_tokens": 272615358.0, "step": 7181 }, { "epoch": 0.9136242208370436, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6790103912353516, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8593098521232605, "num_tokens": 272652927.0, "step": 7182 }, { "epoch": 0.9137514311156342, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.466254472732544, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8567745685577393, "num_tokens": 272682509.0, "step": 7183 }, { "epoch": 0.9138786413942247, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6848297119140625, "learning_rate": 1e-05, "loss": 0.5142, "mean_token_accuracy": 0.8472599983215332, "num_tokens": 272723295.0, "step": 7184 }, { "epoch": 0.9140058516728151, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.441533088684082, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8478147983551025, "num_tokens": 272764985.0, "step": 7185 }, { "epoch": 0.9141330619514056, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.694709539413452, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8644570112228394, "num_tokens": 272807926.0, "step": 7186 }, { "epoch": 0.9142602722299962, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4615955352783203, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8666201233863831, "num_tokens": 272844249.0, "step": 7187 }, { "epoch": 0.9143874825085867, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7095038890838623, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8601812720298767, "num_tokens": 272878029.0, "step": 7188 }, { "epoch": 0.9145146927871772, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4730284214019775, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8490371704101562, "num_tokens": 272913185.0, "step": 7189 }, { "epoch": 0.9146419030657678, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.704068899154663, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8666336536407471, "num_tokens": 272946835.0, "step": 7190 }, { "epoch": 0.9147691133443582, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4651637077331543, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.855938196182251, "num_tokens": 272983467.0, "step": 7191 }, { "epoch": 0.9148963236229487, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.703281879425049, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8637479543685913, "num_tokens": 273022464.0, "step": 7192 }, { "epoch": 0.9150235339015392, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4711318016052246, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8452332615852356, "num_tokens": 273059746.0, "step": 7193 }, { "epoch": 0.9151507441801298, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6863629817962646, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8716903924942017, "num_tokens": 273100815.0, "step": 7194 }, { "epoch": 0.9152779544587203, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4768450260162354, "learning_rate": 1e-05, "loss": 0.4925, "mean_token_accuracy": 0.8511253595352173, "num_tokens": 273139429.0, "step": 7195 }, { "epoch": 0.9154051647373108, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.692417860031128, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8679816722869873, "num_tokens": 273174159.0, "step": 7196 }, { "epoch": 0.9155323750159012, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4681575298309326, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8707278370857239, "num_tokens": 273211434.0, "step": 7197 }, { "epoch": 0.9156595852944918, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7114322185516357, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8544195294380188, "num_tokens": 273243463.0, "step": 7198 }, { "epoch": 0.9157867955730823, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.464966297149658, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8662046194076538, "num_tokens": 273283515.0, "step": 7199 }, { "epoch": 0.9159140058516728, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.9096198081970215, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8554617166519165, "num_tokens": 273322927.0, "step": 7200 }, { "epoch": 0.9160412161302633, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.463064670562744, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8699356317520142, "num_tokens": 273359157.0, "step": 7201 }, { "epoch": 0.9161684264088539, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8082852363586426, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8666951656341553, "num_tokens": 273394385.0, "step": 7202 }, { "epoch": 0.9162956366874443, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.481271982192993, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8544966578483582, "num_tokens": 273430286.0, "step": 7203 }, { "epoch": 0.9164228469660348, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.813406467437744, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8705495595932007, "num_tokens": 273467821.0, "step": 7204 }, { "epoch": 0.9165500572446253, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.508897542953491, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8560593128204346, "num_tokens": 273507659.0, "step": 7205 }, { "epoch": 0.9166772675232159, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7895185947418213, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8487501740455627, "num_tokens": 273546445.0, "step": 7206 }, { "epoch": 0.9168044778018064, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.536714792251587, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8567327857017517, "num_tokens": 273580354.0, "step": 7207 }, { "epoch": 0.9169316880803969, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.746831178665161, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8594712615013123, "num_tokens": 273619194.0, "step": 7208 }, { "epoch": 0.9170588983589874, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5272302627563477, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8682489991188049, "num_tokens": 273662888.0, "step": 7209 }, { "epoch": 0.9171861086375779, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7259957790374756, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8665977716445923, "num_tokens": 273697823.0, "step": 7210 }, { "epoch": 0.9173133189161684, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.527487277984619, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8500698804855347, "num_tokens": 273737541.0, "step": 7211 }, { "epoch": 0.9174405291947589, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6893832683563232, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8550909757614136, "num_tokens": 273777565.0, "step": 7212 }, { "epoch": 0.9175677394733495, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.527146816253662, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8502627611160278, "num_tokens": 273814269.0, "step": 7213 }, { "epoch": 0.91769494975194, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6641576290130615, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.862741231918335, "num_tokens": 273855828.0, "step": 7214 }, { "epoch": 0.9178221600305305, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.529696464538574, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8604862689971924, "num_tokens": 273887514.0, "step": 7215 }, { "epoch": 0.9179493703091209, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6505866050720215, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.852367639541626, "num_tokens": 273923134.0, "step": 7216 }, { "epoch": 0.9180765805877115, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.502699613571167, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.866753339767456, "num_tokens": 273962089.0, "step": 7217 }, { "epoch": 0.918203790866302, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6384336948394775, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8632282018661499, "num_tokens": 273997150.0, "step": 7218 }, { "epoch": 0.9183310011448925, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.490051746368408, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8783633708953857, "num_tokens": 274032144.0, "step": 7219 }, { "epoch": 0.918458211423483, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6397554874420166, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8761754631996155, "num_tokens": 274063344.0, "step": 7220 }, { "epoch": 0.9185854217020736, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.482062816619873, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8681605458259583, "num_tokens": 274102242.0, "step": 7221 }, { "epoch": 0.918712631980664, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.627735137939453, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8631315231323242, "num_tokens": 274146560.0, "step": 7222 }, { "epoch": 0.9188398422592545, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.495224952697754, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.850183367729187, "num_tokens": 274180793.0, "step": 7223 }, { "epoch": 0.918967052537845, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6196229457855225, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8651227951049805, "num_tokens": 274221207.0, "step": 7224 }, { "epoch": 0.9190942628164356, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4725775718688965, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8553447127342224, "num_tokens": 274265413.0, "step": 7225 }, { "epoch": 0.9192214730950261, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6362664699554443, "learning_rate": 1e-05, "loss": 0.5391, "mean_token_accuracy": 0.8375787734985352, "num_tokens": 274305875.0, "step": 7226 }, { "epoch": 0.9193486833736166, "ewc_loss": 0.053466796875, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.471102237701416, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8573793768882751, "num_tokens": 274344223.0, "step": 7227 }, { "epoch": 0.919475893652207, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.636186361312866, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8596170544624329, "num_tokens": 274382671.0, "step": 7228 }, { "epoch": 0.9196031039307976, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4790070056915283, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8613300323486328, "num_tokens": 274413475.0, "step": 7229 }, { "epoch": 0.9197303142093881, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.7058298587799072, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8792739510536194, "num_tokens": 274453714.0, "step": 7230 }, { "epoch": 0.9198575244879786, "ewc_loss": 0.052978515625, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4694530963897705, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8550904393196106, "num_tokens": 274490388.0, "step": 7231 }, { "epoch": 0.9199847347665692, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001735687255859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6595311164855957, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8656259775161743, "num_tokens": 274529939.0, "step": 7232 }, { "epoch": 0.9201119450451597, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.451010227203369, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8591248393058777, "num_tokens": 274570967.0, "step": 7233 }, { "epoch": 0.9202391553237501, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.696493148803711, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8647791147232056, "num_tokens": 274601977.0, "step": 7234 }, { "epoch": 0.9203663656023406, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4504334926605225, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8686908483505249, "num_tokens": 274638509.0, "step": 7235 }, { "epoch": 0.9204935758809312, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.702786445617676, "learning_rate": 1e-05, "loss": 0.3894, "mean_token_accuracy": 0.8831648826599121, "num_tokens": 274674325.0, "step": 7236 }, { "epoch": 0.9206207861595217, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.455286979675293, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.863495409488678, "num_tokens": 274709686.0, "step": 7237 }, { "epoch": 0.9207479964381122, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7211883068084717, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8570547103881836, "num_tokens": 274748917.0, "step": 7238 }, { "epoch": 0.9208752067167028, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4463143348693848, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8509813547134399, "num_tokens": 274794119.0, "step": 7239 }, { "epoch": 0.9210024169952932, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7440569400787354, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8519169092178345, "num_tokens": 274827907.0, "step": 7240 }, { "epoch": 0.9211296272738837, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.46439790725708, "learning_rate": 1e-05, "loss": 0.3895, "mean_token_accuracy": 0.882569432258606, "num_tokens": 274865821.0, "step": 7241 }, { "epoch": 0.9212568375524742, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.73751163482666, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8653774261474609, "num_tokens": 274903550.0, "step": 7242 }, { "epoch": 0.9213840478310648, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4686543941497803, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8647844791412354, "num_tokens": 274944212.0, "step": 7243 }, { "epoch": 0.9215112581096553, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.743422269821167, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8631726503372192, "num_tokens": 274980175.0, "step": 7244 }, { "epoch": 0.9216384683882458, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4853224754333496, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8562241792678833, "num_tokens": 275020018.0, "step": 7245 }, { "epoch": 0.9217656786668362, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.720587730407715, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8573328256607056, "num_tokens": 275063538.0, "step": 7246 }, { "epoch": 0.9218928889454268, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4905688762664795, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8589874505996704, "num_tokens": 275099884.0, "step": 7247 }, { "epoch": 0.9220200992240173, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.833726644515991, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8567973375320435, "num_tokens": 275140292.0, "step": 7248 }, { "epoch": 0.9221473095026078, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.483119487762451, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8641663193702698, "num_tokens": 275178480.0, "step": 7249 }, { "epoch": 0.9222745197811983, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.754220485687256, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8569298982620239, "num_tokens": 275215855.0, "step": 7250 }, { "epoch": 0.9224017300597889, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4934916496276855, "learning_rate": 1e-05, "loss": 0.5128, "mean_token_accuracy": 0.8470859527587891, "num_tokens": 275254895.0, "step": 7251 }, { "epoch": 0.9225289403383793, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.746661901473999, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8616849184036255, "num_tokens": 275292352.0, "step": 7252 }, { "epoch": 0.9226561506169698, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5008344650268555, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8558638095855713, "num_tokens": 275335867.0, "step": 7253 }, { "epoch": 0.9227833608955603, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7233548164367676, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8553869724273682, "num_tokens": 275376551.0, "step": 7254 }, { "epoch": 0.9229105711741509, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5172066688537598, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.85468590259552, "num_tokens": 275412568.0, "step": 7255 }, { "epoch": 0.9230377814527414, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6944496631622314, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8619469404220581, "num_tokens": 275450534.0, "step": 7256 }, { "epoch": 0.9231649917313319, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.51167631149292, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8568856716156006, "num_tokens": 275493165.0, "step": 7257 }, { "epoch": 0.9232922020099223, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.669332981109619, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8633467555046082, "num_tokens": 275533935.0, "step": 7258 }, { "epoch": 0.9234194122885129, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.517366647720337, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8566027879714966, "num_tokens": 275570937.0, "step": 7259 }, { "epoch": 0.9235466225671034, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6598057746887207, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8684226870536804, "num_tokens": 275606036.0, "step": 7260 }, { "epoch": 0.9236738328456939, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4983537197113037, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8653595447540283, "num_tokens": 275645849.0, "step": 7261 }, { "epoch": 0.9238010431242845, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6384379863739014, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8746844530105591, "num_tokens": 275682192.0, "step": 7262 }, { "epoch": 0.923928253402875, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4975476264953613, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8535258173942566, "num_tokens": 275722074.0, "step": 7263 }, { "epoch": 0.9240554636814655, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.633463144302368, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8563574552536011, "num_tokens": 275764966.0, "step": 7264 }, { "epoch": 0.9241826739600559, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4882891178131104, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8742071390151978, "num_tokens": 275805085.0, "step": 7265 }, { "epoch": 0.9243098842386465, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6444804668426514, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8659976124763489, "num_tokens": 275837533.0, "step": 7266 }, { "epoch": 0.924437094517237, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.476876735687256, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8597093820571899, "num_tokens": 275877168.0, "step": 7267 }, { "epoch": 0.9245643047958275, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6753292083740234, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8529394865036011, "num_tokens": 275917861.0, "step": 7268 }, { "epoch": 0.924691515074418, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.496300220489502, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8496951460838318, "num_tokens": 275952433.0, "step": 7269 }, { "epoch": 0.9248187253530086, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6447386741638184, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8623547554016113, "num_tokens": 275993124.0, "step": 7270 }, { "epoch": 0.924945935631599, "ewc_loss": 0.05322265625, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.485863447189331, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8707832098007202, "num_tokens": 276032967.0, "step": 7271 }, { "epoch": 0.9250731459101895, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6408443450927734, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8654674887657166, "num_tokens": 276068840.0, "step": 7272 }, { "epoch": 0.92520035618878, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.472271680831909, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8577393889427185, "num_tokens": 276109196.0, "step": 7273 }, { "epoch": 0.9253275664673706, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6379973888397217, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8692415356636047, "num_tokens": 276151544.0, "step": 7274 }, { "epoch": 0.9254547767459611, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4894375801086426, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8586318492889404, "num_tokens": 276192849.0, "step": 7275 }, { "epoch": 0.9255819870245516, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.638882637023926, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8748078346252441, "num_tokens": 276226543.0, "step": 7276 }, { "epoch": 0.925709197303142, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017452239990234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4641244411468506, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8679183721542358, "num_tokens": 276267199.0, "step": 7277 }, { "epoch": 0.9258364075817326, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.639965534210205, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8573158979415894, "num_tokens": 276306786.0, "step": 7278 }, { "epoch": 0.9259636178603231, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4819324016571045, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8612263202667236, "num_tokens": 276343479.0, "step": 7279 }, { "epoch": 0.9260908281389136, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6613664627075195, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8591981530189514, "num_tokens": 276378067.0, "step": 7280 }, { "epoch": 0.9262180384175042, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4764909744262695, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8612935543060303, "num_tokens": 276407610.0, "step": 7281 }, { "epoch": 0.9263452486960947, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.665663957595825, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.854016900062561, "num_tokens": 276443664.0, "step": 7282 }, { "epoch": 0.9264724589746851, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.456099271774292, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8659719228744507, "num_tokens": 276480446.0, "step": 7283 }, { "epoch": 0.9265996692532756, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.671574831008911, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8684881925582886, "num_tokens": 276515012.0, "step": 7284 }, { "epoch": 0.9267268795318662, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.454667806625366, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8818906545639038, "num_tokens": 276555804.0, "step": 7285 }, { "epoch": 0.9268540898104567, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.691728115081787, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8489424586296082, "num_tokens": 276596290.0, "step": 7286 }, { "epoch": 0.9269813000890472, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4399526119232178, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8525108098983765, "num_tokens": 276641767.0, "step": 7287 }, { "epoch": 0.9271085103676378, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.706713914871216, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8761712312698364, "num_tokens": 276678375.0, "step": 7288 }, { "epoch": 0.9272357206462282, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4644927978515625, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8614369034767151, "num_tokens": 276715835.0, "step": 7289 }, { "epoch": 0.9273629309248187, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7186639308929443, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8586930632591248, "num_tokens": 276750323.0, "step": 7290 }, { "epoch": 0.9274901412034092, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.46164870262146, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8625540733337402, "num_tokens": 276786867.0, "step": 7291 }, { "epoch": 0.9276173514819998, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.721919536590576, "learning_rate": 1e-05, "loss": 0.5414, "mean_token_accuracy": 0.8401682376861572, "num_tokens": 276831263.0, "step": 7292 }, { "epoch": 0.9277445617605903, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.463083505630493, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8667038679122925, "num_tokens": 276867573.0, "step": 7293 }, { "epoch": 0.9278717720391808, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7154741287231445, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8825068473815918, "num_tokens": 276900414.0, "step": 7294 }, { "epoch": 0.9279989823177712, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.47088623046875, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8556420803070068, "num_tokens": 276938336.0, "step": 7295 }, { "epoch": 0.9281261925963618, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.706388473510742, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8705413341522217, "num_tokens": 276981277.0, "step": 7296 }, { "epoch": 0.9282534028749523, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.474571704864502, "learning_rate": 1e-05, "loss": 0.511, "mean_token_accuracy": 0.8556815385818481, "num_tokens": 277023346.0, "step": 7297 }, { "epoch": 0.9283806131535428, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7317798137664795, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8633835315704346, "num_tokens": 277054821.0, "step": 7298 }, { "epoch": 0.9285078234321333, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.476602792739868, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8646210432052612, "num_tokens": 277092124.0, "step": 7299 }, { "epoch": 0.9286350337107239, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.715090274810791, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8613454699516296, "num_tokens": 277130119.0, "step": 7300 }, { "epoch": 0.9287622439893143, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4771575927734375, "learning_rate": 1e-05, "loss": 0.4025, "mean_token_accuracy": 0.8797441124916077, "num_tokens": 277167777.0, "step": 7301 }, { "epoch": 0.9288894542679048, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.702974319458008, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8700682520866394, "num_tokens": 277207273.0, "step": 7302 }, { "epoch": 0.9290166645464953, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4706389904022217, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8705654144287109, "num_tokens": 277248996.0, "step": 7303 }, { "epoch": 0.9291438748250859, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.72029709815979, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8473378419876099, "num_tokens": 277286212.0, "step": 7304 }, { "epoch": 0.9292710851036764, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4825329780578613, "learning_rate": 1e-05, "loss": 0.5169, "mean_token_accuracy": 0.8443410396575928, "num_tokens": 277330482.0, "step": 7305 }, { "epoch": 0.9293982953822669, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7114813327789307, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8690869808197021, "num_tokens": 277364965.0, "step": 7306 }, { "epoch": 0.9295255056608573, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4964964389801025, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.86179518699646, "num_tokens": 277392227.0, "step": 7307 }, { "epoch": 0.9296527159394479, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6790363788604736, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.871618390083313, "num_tokens": 277435658.0, "step": 7308 }, { "epoch": 0.9297799262180384, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4885168075561523, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8607823848724365, "num_tokens": 277472513.0, "step": 7309 }, { "epoch": 0.9299071364966289, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6675779819488525, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8493809700012207, "num_tokens": 277517740.0, "step": 7310 }, { "epoch": 0.9300343467752195, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5045394897460938, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.8770710229873657, "num_tokens": 277547304.0, "step": 7311 }, { "epoch": 0.93016155705381, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.672015428543091, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8499913215637207, "num_tokens": 277581380.0, "step": 7312 }, { "epoch": 0.9302887673324005, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4811768531799316, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8701245784759521, "num_tokens": 277621615.0, "step": 7313 }, { "epoch": 0.9304159776109909, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6816554069519043, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8513094186782837, "num_tokens": 277655157.0, "step": 7314 }, { "epoch": 0.9305431878895815, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4982597827911377, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8480643033981323, "num_tokens": 277688634.0, "step": 7315 }, { "epoch": 0.930670398168172, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6570539474487305, "learning_rate": 1e-05, "loss": 0.5174, "mean_token_accuracy": 0.8500158786773682, "num_tokens": 277736123.0, "step": 7316 }, { "epoch": 0.9307976084467625, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4752349853515625, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8614595532417297, "num_tokens": 277776197.0, "step": 7317 }, { "epoch": 0.930924818725353, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6474530696868896, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8737409114837646, "num_tokens": 277820477.0, "step": 7318 }, { "epoch": 0.9310520290039436, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4745616912841797, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8637115955352783, "num_tokens": 277864660.0, "step": 7319 }, { "epoch": 0.931179239282534, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.661837577819824, "learning_rate": 1e-05, "loss": 0.5061, "mean_token_accuracy": 0.847568154335022, "num_tokens": 277905702.0, "step": 7320 }, { "epoch": 0.9313064495611245, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.463383197784424, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8669766783714294, "num_tokens": 277946033.0, "step": 7321 }, { "epoch": 0.931433659839715, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.67708683013916, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8621162176132202, "num_tokens": 277981556.0, "step": 7322 }, { "epoch": 0.9315608701183056, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4586620330810547, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8753730654716492, "num_tokens": 278021455.0, "step": 7323 }, { "epoch": 0.9316880803968961, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.682859420776367, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8690874576568604, "num_tokens": 278057840.0, "step": 7324 }, { "epoch": 0.9318152906754866, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.471369743347168, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8566388487815857, "num_tokens": 278093186.0, "step": 7325 }, { "epoch": 0.931942500954077, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6949462890625, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8615102171897888, "num_tokens": 278127235.0, "step": 7326 }, { "epoch": 0.9320697112326676, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.460850715637207, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8637644052505493, "num_tokens": 278165236.0, "step": 7327 }, { "epoch": 0.9321969215112581, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.70112681388855, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8503917455673218, "num_tokens": 278205153.0, "step": 7328 }, { "epoch": 0.9323241317898486, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4533989429473877, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8748961687088013, "num_tokens": 278240808.0, "step": 7329 }, { "epoch": 0.9324513420684392, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.704926013946533, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8548042178153992, "num_tokens": 278281364.0, "step": 7330 }, { "epoch": 0.9325785523470297, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.467911720275879, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.867447555065155, "num_tokens": 278316021.0, "step": 7331 }, { "epoch": 0.9327057626256201, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7113399505615234, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.856898844242096, "num_tokens": 278357744.0, "step": 7332 }, { "epoch": 0.9328329729042106, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4829330444335938, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8619017601013184, "num_tokens": 278390826.0, "step": 7333 }, { "epoch": 0.9329601831828012, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7152626514434814, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8462464809417725, "num_tokens": 278424395.0, "step": 7334 }, { "epoch": 0.9330873934613917, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4637651443481445, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8752219676971436, "num_tokens": 278462180.0, "step": 7335 }, { "epoch": 0.9332146037399822, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.724241018295288, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8571944236755371, "num_tokens": 278499349.0, "step": 7336 }, { "epoch": 0.9333418140185727, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4957010746002197, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8508002758026123, "num_tokens": 278530782.0, "step": 7337 }, { "epoch": 0.9334690242971632, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.715607166290283, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8569344878196716, "num_tokens": 278563262.0, "step": 7338 }, { "epoch": 0.9335962345757537, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5038726329803467, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8628180027008057, "num_tokens": 278598882.0, "step": 7339 }, { "epoch": 0.9337234448543442, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7047908306121826, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.863236665725708, "num_tokens": 278628535.0, "step": 7340 }, { "epoch": 0.9338506551329347, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.484484910964966, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8579658269882202, "num_tokens": 278667732.0, "step": 7341 }, { "epoch": 0.9339778654115253, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6858420372009277, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.864804744720459, "num_tokens": 278706925.0, "step": 7342 }, { "epoch": 0.9341050756901158, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4884462356567383, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.874539852142334, "num_tokens": 278740491.0, "step": 7343 }, { "epoch": 0.9342322859687062, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001773834228515625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6893999576568604, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8512752056121826, "num_tokens": 278776739.0, "step": 7344 }, { "epoch": 0.9343594962472968, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4900312423706055, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8722524642944336, "num_tokens": 278813019.0, "step": 7345 }, { "epoch": 0.9344867065258873, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.671830892562866, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8616634011268616, "num_tokens": 278853807.0, "step": 7346 }, { "epoch": 0.9346139168044778, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.482816696166992, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8795560598373413, "num_tokens": 278891643.0, "step": 7347 }, { "epoch": 0.9347411270830683, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6946628093719482, "learning_rate": 1e-05, "loss": 0.5688, "mean_token_accuracy": 0.8425242304801941, "num_tokens": 278929915.0, "step": 7348 }, { "epoch": 0.9348683373616589, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4951350688934326, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8451430201530457, "num_tokens": 278964400.0, "step": 7349 }, { "epoch": 0.9349955476402493, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6726877689361572, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8637543320655823, "num_tokens": 279003322.0, "step": 7350 }, { "epoch": 0.9351227579188398, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4838175773620605, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8508055210113525, "num_tokens": 279045630.0, "step": 7351 }, { "epoch": 0.9352499681974303, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.667843818664551, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8619601726531982, "num_tokens": 279085330.0, "step": 7352 }, { "epoch": 0.9353771784760209, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4715371131896973, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8683807849884033, "num_tokens": 279125499.0, "step": 7353 }, { "epoch": 0.9355043887546114, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6871306896209717, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8654619455337524, "num_tokens": 279154772.0, "step": 7354 }, { "epoch": 0.9356315990332019, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4721133708953857, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.858717679977417, "num_tokens": 279196947.0, "step": 7355 }, { "epoch": 0.9357588093117923, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6789674758911133, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8780922293663025, "num_tokens": 279236872.0, "step": 7356 }, { "epoch": 0.9358860195903829, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4635543823242188, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8707452416419983, "num_tokens": 279275456.0, "step": 7357 }, { "epoch": 0.9360132298689734, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6999576091766357, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8588639497756958, "num_tokens": 279308406.0, "step": 7358 }, { "epoch": 0.9361404401475639, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4629268646240234, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8602041006088257, "num_tokens": 279349522.0, "step": 7359 }, { "epoch": 0.9362676504261545, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.692176580429077, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8740293383598328, "num_tokens": 279388242.0, "step": 7360 }, { "epoch": 0.936394860704745, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4817934036254883, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.875084400177002, "num_tokens": 279421239.0, "step": 7361 }, { "epoch": 0.9365220709833355, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.692824125289917, "learning_rate": 1e-05, "loss": 0.5402, "mean_token_accuracy": 0.8388951420783997, "num_tokens": 279464823.0, "step": 7362 }, { "epoch": 0.9366492812619259, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4783380031585693, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8501896262168884, "num_tokens": 279507645.0, "step": 7363 }, { "epoch": 0.9367764915405165, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6961374282836914, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8595002889633179, "num_tokens": 279539905.0, "step": 7364 }, { "epoch": 0.936903701819107, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.468229293823242, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8712048530578613, "num_tokens": 279577929.0, "step": 7365 }, { "epoch": 0.9370309120976975, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.696239709854126, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8653638362884521, "num_tokens": 279618421.0, "step": 7366 }, { "epoch": 0.937158122376288, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4856724739074707, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8682802319526672, "num_tokens": 279656563.0, "step": 7367 }, { "epoch": 0.9372853326548786, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.695434093475342, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8539974689483643, "num_tokens": 279693159.0, "step": 7368 }, { "epoch": 0.937412542933469, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4745616912841797, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8760433197021484, "num_tokens": 279731492.0, "step": 7369 }, { "epoch": 0.9375397532120595, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.684504270553589, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8669113516807556, "num_tokens": 279774906.0, "step": 7370 }, { "epoch": 0.93766696349065, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.491468906402588, "learning_rate": 1e-05, "loss": 0.5295, "mean_token_accuracy": 0.841199517250061, "num_tokens": 279809797.0, "step": 7371 }, { "epoch": 0.9377941737692406, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6726956367492676, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8679144382476807, "num_tokens": 279850107.0, "step": 7372 }, { "epoch": 0.9379213840478311, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.48062801361084, "learning_rate": 1e-05, "loss": 0.3843, "mean_token_accuracy": 0.8856745958328247, "num_tokens": 279885115.0, "step": 7373 }, { "epoch": 0.9380485943264216, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6724321842193604, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8701285719871521, "num_tokens": 279921343.0, "step": 7374 }, { "epoch": 0.938175804605012, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4867615699768066, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8736281394958496, "num_tokens": 279953059.0, "step": 7375 }, { "epoch": 0.9383030148836026, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.667586326599121, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8695468902587891, "num_tokens": 279987993.0, "step": 7376 }, { "epoch": 0.9384302251621931, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4854159355163574, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8547671437263489, "num_tokens": 280024620.0, "step": 7377 }, { "epoch": 0.9385574354407836, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.653900146484375, "learning_rate": 1e-05, "loss": 0.5242, "mean_token_accuracy": 0.8446899652481079, "num_tokens": 280069011.0, "step": 7378 }, { "epoch": 0.9386846457193742, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4709644317626953, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8828608989715576, "num_tokens": 280105877.0, "step": 7379 }, { "epoch": 0.9388118559979647, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.652935028076172, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8738338351249695, "num_tokens": 280143155.0, "step": 7380 }, { "epoch": 0.9389390662765551, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4711709022521973, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8713721632957458, "num_tokens": 280181490.0, "step": 7381 }, { "epoch": 0.9390662765551456, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.669034719467163, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8515911102294922, "num_tokens": 280221959.0, "step": 7382 }, { "epoch": 0.9391934868337362, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017642974853515625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4574108123779297, "learning_rate": 1e-05, "loss": 0.3936, "mean_token_accuracy": 0.8824279308319092, "num_tokens": 280265224.0, "step": 7383 }, { "epoch": 0.9393206971123267, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6792960166931152, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8661972284317017, "num_tokens": 280302039.0, "step": 7384 }, { "epoch": 0.9394479073909172, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.475672721862793, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.864824116230011, "num_tokens": 280338559.0, "step": 7385 }, { "epoch": 0.9395751176695077, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6765167713165283, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8739444017410278, "num_tokens": 280376731.0, "step": 7386 }, { "epoch": 0.9397023279480982, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4778738021850586, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8604600429534912, "num_tokens": 280414201.0, "step": 7387 }, { "epoch": 0.9398295382266887, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.670100450515747, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8646599650382996, "num_tokens": 280456490.0, "step": 7388 }, { "epoch": 0.9399567485052792, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.464866876602173, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8612477779388428, "num_tokens": 280499593.0, "step": 7389 }, { "epoch": 0.9400839587838697, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6836771965026855, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8559961318969727, "num_tokens": 280535722.0, "step": 7390 }, { "epoch": 0.9402111690624603, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4779865741729736, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8678075671195984, "num_tokens": 280569666.0, "step": 7391 }, { "epoch": 0.9403383793410508, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6684935092926025, "learning_rate": 1e-05, "loss": 0.3726, "mean_token_accuracy": 0.8905411958694458, "num_tokens": 280607359.0, "step": 7392 }, { "epoch": 0.9404655896196412, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.483715534210205, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8607603311538696, "num_tokens": 280641054.0, "step": 7393 }, { "epoch": 0.9405927998982317, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.664482355117798, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8524173498153687, "num_tokens": 280683798.0, "step": 7394 }, { "epoch": 0.9407200101768223, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.465010404586792, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8529443740844727, "num_tokens": 280729407.0, "step": 7395 }, { "epoch": 0.9408472204554128, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6829674243927, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8651341199874878, "num_tokens": 280764327.0, "step": 7396 }, { "epoch": 0.9409744307340033, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.472015380859375, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8595766425132751, "num_tokens": 280802056.0, "step": 7397 }, { "epoch": 0.9411016410125939, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6778478622436523, "learning_rate": 1e-05, "loss": 0.3775, "mean_token_accuracy": 0.8852733373641968, "num_tokens": 280837269.0, "step": 7398 }, { "epoch": 0.9412288512911843, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4855451583862305, "learning_rate": 1e-05, "loss": 0.5257, "mean_token_accuracy": 0.8419918417930603, "num_tokens": 280873914.0, "step": 7399 }, { "epoch": 0.9413560615697748, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6758034229278564, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8648366928100586, "num_tokens": 280911436.0, "step": 7400 }, { "epoch": 0.9414832718483653, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.462722063064575, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8730812072753906, "num_tokens": 280952395.0, "step": 7401 }, { "epoch": 0.9416104821269559, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.684535264968872, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8475000858306885, "num_tokens": 280990455.0, "step": 7402 }, { "epoch": 0.9417376924055464, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.494938850402832, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8645370602607727, "num_tokens": 281020584.0, "step": 7403 }, { "epoch": 0.9418649026841369, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6625585556030273, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8749043345451355, "num_tokens": 281059907.0, "step": 7404 }, { "epoch": 0.9419921129627273, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4878721237182617, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8666731119155884, "num_tokens": 281097425.0, "step": 7405 }, { "epoch": 0.9421193232413179, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6458466053009033, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8613591194152832, "num_tokens": 281146696.0, "step": 7406 }, { "epoch": 0.9422465335199084, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.488248109817505, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8500696420669556, "num_tokens": 281186166.0, "step": 7407 }, { "epoch": 0.9423737437984989, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6485066413879395, "learning_rate": 1e-05, "loss": 0.5648, "mean_token_accuracy": 0.8318200707435608, "num_tokens": 281230372.0, "step": 7408 }, { "epoch": 0.9425009540770894, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5144553184509277, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.864614725112915, "num_tokens": 281256035.0, "step": 7409 }, { "epoch": 0.94262816435568, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6218295097351074, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8769392371177673, "num_tokens": 281290792.0, "step": 7410 }, { "epoch": 0.9427553746342705, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4980661869049072, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8685490489006042, "num_tokens": 281329949.0, "step": 7411 }, { "epoch": 0.9428825849128609, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.615920066833496, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8712031841278076, "num_tokens": 281366383.0, "step": 7412 }, { "epoch": 0.9430097951914514, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4823763370513916, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8534568548202515, "num_tokens": 281408537.0, "step": 7413 }, { "epoch": 0.943137005470042, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6139461994171143, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.857016384601593, "num_tokens": 281447745.0, "step": 7414 }, { "epoch": 0.9432642157486325, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.477137565612793, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8688997030258179, "num_tokens": 281484430.0, "step": 7415 }, { "epoch": 0.943391426027223, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.632669687271118, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8816448450088501, "num_tokens": 281519720.0, "step": 7416 }, { "epoch": 0.9435186363058136, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4838714599609375, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8490866422653198, "num_tokens": 281556039.0, "step": 7417 }, { "epoch": 0.943645846584404, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6356217861175537, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8551521301269531, "num_tokens": 281594781.0, "step": 7418 }, { "epoch": 0.9437730568629945, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.472486734390259, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8511857986450195, "num_tokens": 281632743.0, "step": 7419 }, { "epoch": 0.943900267141585, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6494715213775635, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8636561632156372, "num_tokens": 281669706.0, "step": 7420 }, { "epoch": 0.9440274774201756, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4628334045410156, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8679943084716797, "num_tokens": 281706108.0, "step": 7421 }, { "epoch": 0.9441546876987661, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6656253337860107, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8573094010353088, "num_tokens": 281743437.0, "step": 7422 }, { "epoch": 0.9442818979773566, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.454352378845215, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8763555288314819, "num_tokens": 281779652.0, "step": 7423 }, { "epoch": 0.944409108255947, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6776950359344482, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8678661584854126, "num_tokens": 281815841.0, "step": 7424 }, { "epoch": 0.9445363185345376, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.454785108566284, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.863260805606842, "num_tokens": 281853991.0, "step": 7425 }, { "epoch": 0.9446635288131281, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.723797559738159, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8578519225120544, "num_tokens": 281883264.0, "step": 7426 }, { "epoch": 0.9447907390917186, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.436129570007324, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8779997229576111, "num_tokens": 281920266.0, "step": 7427 }, { "epoch": 0.9449179493703092, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.72998309135437, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8637121319770813, "num_tokens": 281959949.0, "step": 7428 }, { "epoch": 0.9450451596488997, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.450206756591797, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8690268993377686, "num_tokens": 281992255.0, "step": 7429 }, { "epoch": 0.9451723699274901, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.755035161972046, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8635736703872681, "num_tokens": 282028000.0, "step": 7430 }, { "epoch": 0.9452995802060806, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4611575603485107, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8648630380630493, "num_tokens": 282064332.0, "step": 7431 }, { "epoch": 0.9454267904846712, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.766223192214966, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8519827127456665, "num_tokens": 282102185.0, "step": 7432 }, { "epoch": 0.9455540007632617, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00017833709716796875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.492502212524414, "learning_rate": 1e-05, "loss": 0.5646, "mean_token_accuracy": 0.8270163536071777, "num_tokens": 282138708.0, "step": 7433 }, { "epoch": 0.9456812110418522, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7627599239349365, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8743115067481995, "num_tokens": 282168223.0, "step": 7434 }, { "epoch": 0.9458084213204427, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.485079288482666, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8739787936210632, "num_tokens": 282205226.0, "step": 7435 }, { "epoch": 0.9459356315990332, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7570927143096924, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8506002426147461, "num_tokens": 282239131.0, "step": 7436 }, { "epoch": 0.9460628418776237, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.493972063064575, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8644991517066956, "num_tokens": 282276175.0, "step": 7437 }, { "epoch": 0.9461900521562142, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7553279399871826, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8543891906738281, "num_tokens": 282308598.0, "step": 7438 }, { "epoch": 0.9463172624348047, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4953718185424805, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8812236189842224, "num_tokens": 282342193.0, "step": 7439 }, { "epoch": 0.9464444727133953, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7295572757720947, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8643089532852173, "num_tokens": 282380531.0, "step": 7440 }, { "epoch": 0.9465716829919858, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5262629985809326, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8593107461929321, "num_tokens": 282410713.0, "step": 7441 }, { "epoch": 0.9466988932705762, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7007172107696533, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8664152026176453, "num_tokens": 282451881.0, "step": 7442 }, { "epoch": 0.9468261035491667, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.529289484024048, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8474951982498169, "num_tokens": 282486031.0, "step": 7443 }, { "epoch": 0.9469533138277573, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.663189649581909, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.87520432472229, "num_tokens": 282523753.0, "step": 7444 }, { "epoch": 0.9470805241063478, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5150043964385986, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8574360013008118, "num_tokens": 282563322.0, "step": 7445 }, { "epoch": 0.9472077343849383, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6486475467681885, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8602508902549744, "num_tokens": 282602574.0, "step": 7446 }, { "epoch": 0.9473349446635289, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5150046348571777, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8691905736923218, "num_tokens": 282636608.0, "step": 7447 }, { "epoch": 0.9474621549421193, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6290407180786133, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8785219788551331, "num_tokens": 282675320.0, "step": 7448 }, { "epoch": 0.9475893652207098, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.51533842086792, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.858440637588501, "num_tokens": 282712537.0, "step": 7449 }, { "epoch": 0.9477165754993003, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6283504962921143, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8674967288970947, "num_tokens": 282744620.0, "step": 7450 }, { "epoch": 0.9478437857778909, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4957075119018555, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8711508512496948, "num_tokens": 282786480.0, "step": 7451 }, { "epoch": 0.9479709960564814, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6302404403686523, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8716288805007935, "num_tokens": 282819143.0, "step": 7452 }, { "epoch": 0.9480982063350719, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.481519937515259, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8653075695037842, "num_tokens": 282859658.0, "step": 7453 }, { "epoch": 0.9482254166136623, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.643031120300293, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8619989156723022, "num_tokens": 282896190.0, "step": 7454 }, { "epoch": 0.9483526268922529, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.46964168548584, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8541628122329712, "num_tokens": 282937458.0, "step": 7455 }, { "epoch": 0.9484798371708434, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.657773017883301, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8641147613525391, "num_tokens": 282970669.0, "step": 7456 }, { "epoch": 0.9486070474494339, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.459296703338623, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8600070476531982, "num_tokens": 283012372.0, "step": 7457 }, { "epoch": 0.9487342577280244, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6855435371398926, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8513622283935547, "num_tokens": 283046165.0, "step": 7458 }, { "epoch": 0.948861468006615, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4543263912200928, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8613786101341248, "num_tokens": 283086539.0, "step": 7459 }, { "epoch": 0.9489886782852054, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7010862827301025, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8593613505363464, "num_tokens": 283123643.0, "step": 7460 }, { "epoch": 0.9491158885637959, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.447031021118164, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8805776238441467, "num_tokens": 283162354.0, "step": 7461 }, { "epoch": 0.9492430988423864, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7231035232543945, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8703055381774902, "num_tokens": 283196814.0, "step": 7462 }, { "epoch": 0.949370309120977, "ewc_loss": 0.0537109375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.449427843093872, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8566664457321167, "num_tokens": 283238051.0, "step": 7463 }, { "epoch": 0.9494975193995675, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.739439010620117, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8539102673530579, "num_tokens": 283278025.0, "step": 7464 }, { "epoch": 0.949624729678158, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4674177169799805, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8686832189559937, "num_tokens": 283311279.0, "step": 7465 }, { "epoch": 0.9497519399567486, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.741248607635498, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8558358550071716, "num_tokens": 283351350.0, "step": 7466 }, { "epoch": 0.949879150235339, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4672017097473145, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8682747483253479, "num_tokens": 283387374.0, "step": 7467 }, { "epoch": 0.9500063605139295, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.741713285446167, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8569660186767578, "num_tokens": 283431305.0, "step": 7468 }, { "epoch": 0.95013357079252, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.494878053665161, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8634315729141235, "num_tokens": 283464941.0, "step": 7469 }, { "epoch": 0.9502607810711106, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7270162105560303, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8454137444496155, "num_tokens": 283507254.0, "step": 7470 }, { "epoch": 0.9503879913497011, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4938552379608154, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8510751128196716, "num_tokens": 283545117.0, "step": 7471 }, { "epoch": 0.9505152016282916, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7260515689849854, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8520669937133789, "num_tokens": 283577786.0, "step": 7472 }, { "epoch": 0.950642411906882, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.492537498474121, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8690555691719055, "num_tokens": 283613486.0, "step": 7473 }, { "epoch": 0.9507696221854726, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.703562021255493, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8630048036575317, "num_tokens": 283654480.0, "step": 7474 }, { "epoch": 0.9508968324640631, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000179290771484375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.513537883758545, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8571103811264038, "num_tokens": 283686089.0, "step": 7475 }, { "epoch": 0.9510240427426536, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6812238693237305, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8633265495300293, "num_tokens": 283726483.0, "step": 7476 }, { "epoch": 0.9511512530212441, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5191988945007324, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.867072582244873, "num_tokens": 283756170.0, "step": 7477 }, { "epoch": 0.9512784632998347, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6484148502349854, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8646735548973083, "num_tokens": 283794950.0, "step": 7478 }, { "epoch": 0.9514056735784251, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.510624647140503, "learning_rate": 1e-05, "loss": 0.5385, "mean_token_accuracy": 0.8405571579933167, "num_tokens": 283835596.0, "step": 7479 }, { "epoch": 0.9515328838570156, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.629234552383423, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8543534278869629, "num_tokens": 283875380.0, "step": 7480 }, { "epoch": 0.9516600941356061, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.00018024444580078125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.514292001724243, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8588827848434448, "num_tokens": 283910026.0, "step": 7481 }, { "epoch": 0.9517873044141967, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.614785671234131, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.866973876953125, "num_tokens": 283950533.0, "step": 7482 }, { "epoch": 0.9519145146927872, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.524862051010132, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8639546632766724, "num_tokens": 283980939.0, "step": 7483 }, { "epoch": 0.9520417249713777, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.601022481918335, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8643607497215271, "num_tokens": 284018923.0, "step": 7484 }, { "epoch": 0.9521689352499682, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.496354579925537, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.8803245425224304, "num_tokens": 284059752.0, "step": 7485 }, { "epoch": 0.9522961455285587, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5895721912384033, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8698832392692566, "num_tokens": 284103783.0, "step": 7486 }, { "epoch": 0.9524233558071492, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.5078275203704834, "learning_rate": 1e-05, "loss": 0.5357, "mean_token_accuracy": 0.8425400257110596, "num_tokens": 284144595.0, "step": 7487 }, { "epoch": 0.9525505660857397, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.593360662460327, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.858441174030304, "num_tokens": 284182373.0, "step": 7488 }, { "epoch": 0.9526777763643303, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.50359845161438, "learning_rate": 1e-05, "loss": 0.5274, "mean_token_accuracy": 0.840674102306366, "num_tokens": 284222537.0, "step": 7489 }, { "epoch": 0.9528049866429208, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.588134527206421, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.857122540473938, "num_tokens": 284264780.0, "step": 7490 }, { "epoch": 0.9529321969215112, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4973509311676025, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8760356903076172, "num_tokens": 284298663.0, "step": 7491 }, { "epoch": 0.9530594072001017, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5938150882720947, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8669863343238831, "num_tokens": 284330516.0, "step": 7492 }, { "epoch": 0.9531866174786923, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4958696365356445, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.8508114814758301, "num_tokens": 284370414.0, "step": 7493 }, { "epoch": 0.9533138277572828, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5821516513824463, "learning_rate": 1e-05, "loss": 0.3874, "mean_token_accuracy": 0.884300947189331, "num_tokens": 284409079.0, "step": 7494 }, { "epoch": 0.9534410380358733, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4787042140960693, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8712784051895142, "num_tokens": 284451937.0, "step": 7495 }, { "epoch": 0.9535682483144639, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6002540588378906, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8580912351608276, "num_tokens": 284490174.0, "step": 7496 }, { "epoch": 0.9536954585930543, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4811148643493652, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8708257675170898, "num_tokens": 284530578.0, "step": 7497 }, { "epoch": 0.9538226688716448, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.635590076446533, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8501207828521729, "num_tokens": 284562412.0, "step": 7498 }, { "epoch": 0.9539498791502353, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.462432384490967, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8629090785980225, "num_tokens": 284601846.0, "step": 7499 }, { "epoch": 0.9540770894288259, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6327474117279053, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8572522401809692, "num_tokens": 284640918.0, "step": 7500 }, { "epoch": 0.9542042997074164, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4649055004119873, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8717312812805176, "num_tokens": 284672268.0, "step": 7501 }, { "epoch": 0.9543315099860069, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.651224374771118, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8694508671760559, "num_tokens": 284708638.0, "step": 7502 }, { "epoch": 0.9544587202645973, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4710752964019775, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8416383266448975, "num_tokens": 284747670.0, "step": 7503 }, { "epoch": 0.9545859305431879, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6578116416931152, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8565041422843933, "num_tokens": 284786561.0, "step": 7504 }, { "epoch": 0.9547131408217784, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.450322151184082, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8668084144592285, "num_tokens": 284828631.0, "step": 7505 }, { "epoch": 0.9548403511003689, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6726748943328857, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.86335688829422, "num_tokens": 284865782.0, "step": 7506 }, { "epoch": 0.9549675613789594, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4430017471313477, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.869206428527832, "num_tokens": 284904000.0, "step": 7507 }, { "epoch": 0.95509477165755, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6862518787384033, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8669112920761108, "num_tokens": 284946432.0, "step": 7508 }, { "epoch": 0.9552219819361404, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.449582099914551, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8645862340927124, "num_tokens": 284985534.0, "step": 7509 }, { "epoch": 0.9553491922147309, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7025723457336426, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8755375742912292, "num_tokens": 285023710.0, "step": 7510 }, { "epoch": 0.9554764024933214, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4473464488983154, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8510684967041016, "num_tokens": 285064182.0, "step": 7511 }, { "epoch": 0.955603612771912, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.727353096008301, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8671671152114868, "num_tokens": 285100896.0, "step": 7512 }, { "epoch": 0.9557308230505025, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.450650930404663, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8632473349571228, "num_tokens": 285139347.0, "step": 7513 }, { "epoch": 0.955858033329093, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7350101470947266, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8620054721832275, "num_tokens": 285183047.0, "step": 7514 }, { "epoch": 0.9559852436076836, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4487030506134033, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.862323522567749, "num_tokens": 285225407.0, "step": 7515 }, { "epoch": 0.956112453886274, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.764296531677246, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8684571981430054, "num_tokens": 285258465.0, "step": 7516 }, { "epoch": 0.9562396641648645, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.464529037475586, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8640989065170288, "num_tokens": 285297221.0, "step": 7517 }, { "epoch": 0.956366874443455, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.754770278930664, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8680852651596069, "num_tokens": 285338145.0, "step": 7518 }, { "epoch": 0.9564940847220456, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.466442823410034, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8655928373336792, "num_tokens": 285381647.0, "step": 7519 }, { "epoch": 0.9566212950006361, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7667744159698486, "learning_rate": 1e-05, "loss": 0.52, "mean_token_accuracy": 0.8436129689216614, "num_tokens": 285424468.0, "step": 7520 }, { "epoch": 0.9567485052792266, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4843037128448486, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8679606914520264, "num_tokens": 285470415.0, "step": 7521 }, { "epoch": 0.956875715557817, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.758263111114502, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8515236973762512, "num_tokens": 285513299.0, "step": 7522 }, { "epoch": 0.9570029258364076, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5367233753204346, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8482374548912048, "num_tokens": 285555855.0, "step": 7523 }, { "epoch": 0.9571301361149981, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.713059425354004, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.867585301399231, "num_tokens": 285589670.0, "step": 7524 }, { "epoch": 0.9572573463935886, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5135092735290527, "learning_rate": 1e-05, "loss": 0.5029, "mean_token_accuracy": 0.8512259125709534, "num_tokens": 285627257.0, "step": 7525 }, { "epoch": 0.9573845566721791, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6775290966033936, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.867457926273346, "num_tokens": 285670100.0, "step": 7526 }, { "epoch": 0.9575117669507697, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.514036178588867, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8670731782913208, "num_tokens": 285706902.0, "step": 7527 }, { "epoch": 0.9576389772293601, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.665998697280884, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8591924905776978, "num_tokens": 285746952.0, "step": 7528 }, { "epoch": 0.9577661875079506, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.515288829803467, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.876562237739563, "num_tokens": 285781532.0, "step": 7529 }, { "epoch": 0.9578933977865411, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6481189727783203, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8570821285247803, "num_tokens": 285818725.0, "step": 7530 }, { "epoch": 0.9580206080651317, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5148308277130127, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8543365597724915, "num_tokens": 285855502.0, "step": 7531 }, { "epoch": 0.9581478183437222, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6414647102355957, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8603424429893494, "num_tokens": 285893815.0, "step": 7532 }, { "epoch": 0.9582750286223127, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5073981285095215, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8596402406692505, "num_tokens": 285928709.0, "step": 7533 }, { "epoch": 0.9584022389009031, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.630948305130005, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.8790827393531799, "num_tokens": 285959358.0, "step": 7534 }, { "epoch": 0.9585294491794937, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.504380464553833, "learning_rate": 1e-05, "loss": 0.5151, "mean_token_accuracy": 0.8455202579498291, "num_tokens": 285997643.0, "step": 7535 }, { "epoch": 0.9586566594580842, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.621708631515503, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8665245771408081, "num_tokens": 286032866.0, "step": 7536 }, { "epoch": 0.9587838697366747, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.507061243057251, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8630112409591675, "num_tokens": 286066585.0, "step": 7537 }, { "epoch": 0.9589110800152653, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6160526275634766, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8481494188308716, "num_tokens": 286106151.0, "step": 7538 }, { "epoch": 0.9590382902938558, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4880735874176025, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8683465719223022, "num_tokens": 286142409.0, "step": 7539 }, { "epoch": 0.9591655005724462, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6249029636383057, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8522084355354309, "num_tokens": 286179408.0, "step": 7540 }, { "epoch": 0.9592927108510367, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.491590976715088, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8620842695236206, "num_tokens": 286212767.0, "step": 7541 }, { "epoch": 0.9594199211296273, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6235294342041016, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8613041639328003, "num_tokens": 286250176.0, "step": 7542 }, { "epoch": 0.9595471314082178, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.481204032897949, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8599808812141418, "num_tokens": 286292060.0, "step": 7543 }, { "epoch": 0.9596743416868083, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6250524520874023, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8689707517623901, "num_tokens": 286333047.0, "step": 7544 }, { "epoch": 0.9598015519653988, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.467139959335327, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8658989667892456, "num_tokens": 286373465.0, "step": 7545 }, { "epoch": 0.9599287622439893, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.642618417739868, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8523476123809814, "num_tokens": 286414431.0, "step": 7546 }, { "epoch": 0.9600559725225798, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.48138165473938, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8606394529342651, "num_tokens": 286449431.0, "step": 7547 }, { "epoch": 0.9601831828011703, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6465091705322266, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8651761412620544, "num_tokens": 286486563.0, "step": 7548 }, { "epoch": 0.9603103930797608, "ewc_loss": 0.053955078125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4648873805999756, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.854958176612854, "num_tokens": 286528691.0, "step": 7549 }, { "epoch": 0.9604376033583514, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6513431072235107, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8591628074645996, "num_tokens": 286568726.0, "step": 7550 }, { "epoch": 0.9605648136369419, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4674201011657715, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8714259266853333, "num_tokens": 286608159.0, "step": 7551 }, { "epoch": 0.9606920239155323, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6751585006713867, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.858333170413971, "num_tokens": 286645494.0, "step": 7552 }, { "epoch": 0.9608192341941229, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4535329341888428, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8719104528427124, "num_tokens": 286679819.0, "step": 7553 }, { "epoch": 0.9609464444727134, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.698421001434326, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8481616377830505, "num_tokens": 286714116.0, "step": 7554 }, { "epoch": 0.9610736547513039, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.455796718597412, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8757560849189758, "num_tokens": 286746905.0, "step": 7555 }, { "epoch": 0.9612008650298944, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6883864402770996, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.883817195892334, "num_tokens": 286784724.0, "step": 7556 }, { "epoch": 0.961328075308485, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4505202770233154, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8679399490356445, "num_tokens": 286825672.0, "step": 7557 }, { "epoch": 0.9614552855870754, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7291061878204346, "learning_rate": 1e-05, "loss": 0.5156, "mean_token_accuracy": 0.8508524894714355, "num_tokens": 286862072.0, "step": 7558 }, { "epoch": 0.9615824958656659, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4488210678100586, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8636059165000916, "num_tokens": 286903111.0, "step": 7559 }, { "epoch": 0.9617097061442564, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.731631278991699, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8674629926681519, "num_tokens": 286947957.0, "step": 7560 }, { "epoch": 0.961836916422847, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.468480348587036, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8633767366409302, "num_tokens": 286983254.0, "step": 7561 }, { "epoch": 0.9619641267014375, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.739699125289917, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8841840028762817, "num_tokens": 287015294.0, "step": 7562 }, { "epoch": 0.962091336980028, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.468991279602051, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8673576712608337, "num_tokens": 287053439.0, "step": 7563 }, { "epoch": 0.9622185472586186, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7483232021331787, "learning_rate": 1e-05, "loss": 0.5384, "mean_token_accuracy": 0.840002179145813, "num_tokens": 287096970.0, "step": 7564 }, { "epoch": 0.962345757537209, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001811981201171875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.478701591491699, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8555009961128235, "num_tokens": 287139179.0, "step": 7565 }, { "epoch": 0.9624729678157995, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7456398010253906, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8568768501281738, "num_tokens": 287177375.0, "step": 7566 }, { "epoch": 0.96260017809439, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.496971607208252, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8655220866203308, "num_tokens": 287211296.0, "step": 7567 }, { "epoch": 0.9627273883729806, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.729606866836548, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.86927330493927, "num_tokens": 287246913.0, "step": 7568 }, { "epoch": 0.9628545986515711, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4997308254241943, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.859586238861084, "num_tokens": 287283639.0, "step": 7569 }, { "epoch": 0.9629818089301616, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.714604616165161, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8702921867370605, "num_tokens": 287322694.0, "step": 7570 }, { "epoch": 0.963109019208752, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.513390302658081, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8539503812789917, "num_tokens": 287361103.0, "step": 7571 }, { "epoch": 0.9632362294873426, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7050893306732178, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8478831648826599, "num_tokens": 287400666.0, "step": 7572 }, { "epoch": 0.9633634397659331, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4960134029388428, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8785972595214844, "num_tokens": 287434729.0, "step": 7573 }, { "epoch": 0.9634906500445236, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.685685396194458, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8739875555038452, "num_tokens": 287472194.0, "step": 7574 }, { "epoch": 0.9636178603231141, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.488924026489258, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8573129773139954, "num_tokens": 287510139.0, "step": 7575 }, { "epoch": 0.9637450706017047, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.690519332885742, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8517539501190186, "num_tokens": 287549529.0, "step": 7576 }, { "epoch": 0.9638722808802951, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4758753776550293, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8467219471931458, "num_tokens": 287596525.0, "step": 7577 }, { "epoch": 0.9639994911588856, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.687991142272949, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8791491985321045, "num_tokens": 287632284.0, "step": 7578 }, { "epoch": 0.9641267014374761, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4694576263427734, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8672453165054321, "num_tokens": 287671867.0, "step": 7579 }, { "epoch": 0.9642539117160667, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.7065181732177734, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8536983132362366, "num_tokens": 287706715.0, "step": 7580 }, { "epoch": 0.9643811219946572, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4693214893341064, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8647652864456177, "num_tokens": 287746269.0, "step": 7581 }, { "epoch": 0.9645083322732477, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7091615200042725, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8537908792495728, "num_tokens": 287786194.0, "step": 7582 }, { "epoch": 0.9646355425518381, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.489565849304199, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8551480174064636, "num_tokens": 287820873.0, "step": 7583 }, { "epoch": 0.9647627528304287, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.725679397583008, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8627996444702148, "num_tokens": 287847356.0, "step": 7584 }, { "epoch": 0.9648899631090192, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.493227958679199, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8574591875076294, "num_tokens": 287879611.0, "step": 7585 }, { "epoch": 0.9650171733876097, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6982970237731934, "learning_rate": 1e-05, "loss": 0.5168, "mean_token_accuracy": 0.8489129543304443, "num_tokens": 287920865.0, "step": 7586 }, { "epoch": 0.9651443836662003, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4769275188446045, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8562233448028564, "num_tokens": 287963997.0, "step": 7587 }, { "epoch": 0.9652715939447908, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7015600204467773, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8554785847663879, "num_tokens": 287997283.0, "step": 7588 }, { "epoch": 0.9653988042233812, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4955101013183594, "learning_rate": 1e-05, "loss": 0.5382, "mean_token_accuracy": 0.8376519083976746, "num_tokens": 288037876.0, "step": 7589 }, { "epoch": 0.9655260145019717, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6950230598449707, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.8467854857444763, "num_tokens": 288078835.0, "step": 7590 }, { "epoch": 0.9656532247805623, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4918644428253174, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8559343218803406, "num_tokens": 288118342.0, "step": 7591 }, { "epoch": 0.9657804350591528, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.680758476257324, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8656069040298462, "num_tokens": 288159629.0, "step": 7592 }, { "epoch": 0.9659076453377433, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4936697483062744, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8560553789138794, "num_tokens": 288199887.0, "step": 7593 }, { "epoch": 0.9660348556163338, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.685985565185547, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8527483344078064, "num_tokens": 288233958.0, "step": 7594 }, { "epoch": 0.9661620658949243, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4897305965423584, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8586973547935486, "num_tokens": 288274735.0, "step": 7595 }, { "epoch": 0.9662892761735148, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.67966365814209, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8661941885948181, "num_tokens": 288309699.0, "step": 7596 }, { "epoch": 0.9664164864521053, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.491647243499756, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8616300225257874, "num_tokens": 288344397.0, "step": 7597 }, { "epoch": 0.9665436967306958, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6649279594421387, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8673838376998901, "num_tokens": 288385612.0, "step": 7598 }, { "epoch": 0.9666709070092864, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.499889373779297, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8482938408851624, "num_tokens": 288420570.0, "step": 7599 }, { "epoch": 0.9667981172878769, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.662186861038208, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.849604070186615, "num_tokens": 288456917.0, "step": 7600 }, { "epoch": 0.9669253275664673, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.484682559967041, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8468281030654907, "num_tokens": 288498228.0, "step": 7601 }, { "epoch": 0.9670525378450578, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.666001319885254, "learning_rate": 1e-05, "loss": 0.5297, "mean_token_accuracy": 0.8420672416687012, "num_tokens": 288534082.0, "step": 7602 }, { "epoch": 0.9671797481236484, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4777257442474365, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8573250770568848, "num_tokens": 288576142.0, "step": 7603 }, { "epoch": 0.9673069584022389, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6699607372283936, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8551927804946899, "num_tokens": 288609707.0, "step": 7604 }, { "epoch": 0.9674341686808294, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4899892807006836, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8467370867729187, "num_tokens": 288648228.0, "step": 7605 }, { "epoch": 0.96756137895942, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.656683921813965, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8675678968429565, "num_tokens": 288689845.0, "step": 7606 }, { "epoch": 0.9676885892380104, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4740498065948486, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.87486732006073, "num_tokens": 288729571.0, "step": 7607 }, { "epoch": 0.9678157995166009, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.669989585876465, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8757810592651367, "num_tokens": 288764070.0, "step": 7608 }, { "epoch": 0.9679430097951914, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.475081205368042, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8639671802520752, "num_tokens": 288805557.0, "step": 7609 }, { "epoch": 0.968070220073782, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6703286170959473, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.869185209274292, "num_tokens": 288840567.0, "step": 7610 }, { "epoch": 0.9681974303523725, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.463959217071533, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8600850701332092, "num_tokens": 288881952.0, "step": 7611 }, { "epoch": 0.968324640630963, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.690998077392578, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8604251146316528, "num_tokens": 288919208.0, "step": 7612 }, { "epoch": 0.9684518509095535, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.461014986038208, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8608642816543579, "num_tokens": 288961622.0, "step": 7613 }, { "epoch": 0.968579061188144, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.693337917327881, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8703088760375977, "num_tokens": 288999799.0, "step": 7614 }, { "epoch": 0.9687062714667345, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.453836441040039, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8720771074295044, "num_tokens": 289040434.0, "step": 7615 }, { "epoch": 0.968833481745325, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.73730206489563, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8603795766830444, "num_tokens": 289072847.0, "step": 7616 }, { "epoch": 0.9689606920239155, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4528098106384277, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.8475543260574341, "num_tokens": 289114838.0, "step": 7617 }, { "epoch": 0.9690879023025061, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7494170665740967, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8562630414962769, "num_tokens": 289151937.0, "step": 7618 }, { "epoch": 0.9692151125810966, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.465773820877075, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.862419843673706, "num_tokens": 289188630.0, "step": 7619 }, { "epoch": 0.969342322859687, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.74660062789917, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8649671077728271, "num_tokens": 289229412.0, "step": 7620 }, { "epoch": 0.9694695331382776, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4653143882751465, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8718655109405518, "num_tokens": 289265195.0, "step": 7621 }, { "epoch": 0.9695967434168681, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.75449538230896, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8525255918502808, "num_tokens": 289306203.0, "step": 7622 }, { "epoch": 0.9697239536954586, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.484743595123291, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8572189211845398, "num_tokens": 289345196.0, "step": 7623 }, { "epoch": 0.9698511639740491, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.752413272857666, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8714271783828735, "num_tokens": 289377221.0, "step": 7624 }, { "epoch": 0.9699783742526397, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4946675300598145, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8563501238822937, "num_tokens": 289413196.0, "step": 7625 }, { "epoch": 0.9701055845312301, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.728830575942993, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.855699896812439, "num_tokens": 289452774.0, "step": 7626 }, { "epoch": 0.9702327948098206, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4983794689178467, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8619100451469421, "num_tokens": 289489806.0, "step": 7627 }, { "epoch": 0.9703600050884111, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7186667919158936, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8619834780693054, "num_tokens": 289528362.0, "step": 7628 }, { "epoch": 0.9704872153670017, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.506901264190674, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8626279830932617, "num_tokens": 289562725.0, "step": 7629 }, { "epoch": 0.9706144256455922, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.693192958831787, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8565386533737183, "num_tokens": 289605968.0, "step": 7630 }, { "epoch": 0.9707416359241827, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.50915789604187, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8542460203170776, "num_tokens": 289639065.0, "step": 7631 }, { "epoch": 0.9708688462027731, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.682952404022217, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8409268260002136, "num_tokens": 289679616.0, "step": 7632 }, { "epoch": 0.9709960564813637, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.511798858642578, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8623855710029602, "num_tokens": 289713189.0, "step": 7633 }, { "epoch": 0.9711232667599542, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6569886207580566, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8661020398139954, "num_tokens": 289751406.0, "step": 7634 }, { "epoch": 0.9712504770385447, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4904565811157227, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8615497350692749, "num_tokens": 289795735.0, "step": 7635 }, { "epoch": 0.9713776873171353, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.670698404312134, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8485462665557861, "num_tokens": 289833574.0, "step": 7636 }, { "epoch": 0.9715048975957258, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.507291555404663, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8721271753311157, "num_tokens": 289871922.0, "step": 7637 }, { "epoch": 0.9716321078743162, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.648627281188965, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8551025390625, "num_tokens": 289913517.0, "step": 7638 }, { "epoch": 0.9717593181529067, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4924123287200928, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8601994514465332, "num_tokens": 289957046.0, "step": 7639 }, { "epoch": 0.9718865284314973, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.637786626815796, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8621325492858887, "num_tokens": 289997983.0, "step": 7640 }, { "epoch": 0.9720137387100878, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.483273506164551, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8632100224494934, "num_tokens": 290035923.0, "step": 7641 }, { "epoch": 0.9721409489886783, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6429524421691895, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8700624704360962, "num_tokens": 290075152.0, "step": 7642 }, { "epoch": 0.9722681592672688, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.494452714920044, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8753389120101929, "num_tokens": 290108360.0, "step": 7643 }, { "epoch": 0.9723953695458593, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.631591320037842, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8708916902542114, "num_tokens": 290148698.0, "step": 7644 }, { "epoch": 0.9725225798244498, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.486478567123413, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8692934513092041, "num_tokens": 290185847.0, "step": 7645 }, { "epoch": 0.9726497901030403, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.624581813812256, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8619488477706909, "num_tokens": 290228763.0, "step": 7646 }, { "epoch": 0.9727770003816308, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5120043754577637, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8580621480941772, "num_tokens": 290262040.0, "step": 7647 }, { "epoch": 0.9729042106602214, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6172053813934326, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8676329851150513, "num_tokens": 290302205.0, "step": 7648 }, { "epoch": 0.9730314209388119, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.493056535720825, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8674854040145874, "num_tokens": 290340132.0, "step": 7649 }, { "epoch": 0.9731586312174023, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.610358476638794, "learning_rate": 1e-05, "loss": 0.3791, "mean_token_accuracy": 0.8873137831687927, "num_tokens": 290375808.0, "step": 7650 }, { "epoch": 0.9732858414959928, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.487588405609131, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8581449389457703, "num_tokens": 290415538.0, "step": 7651 }, { "epoch": 0.9734130517745834, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6252329349517822, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8696261644363403, "num_tokens": 290453363.0, "step": 7652 }, { "epoch": 0.9735402620531739, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.509190797805786, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8549940586090088, "num_tokens": 290483869.0, "step": 7653 }, { "epoch": 0.9736674723317644, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.606201171875, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8697367310523987, "num_tokens": 290521772.0, "step": 7654 }, { "epoch": 0.973794682610355, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4906692504882812, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8468201756477356, "num_tokens": 290563226.0, "step": 7655 }, { "epoch": 0.9739218928889454, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6043689250946045, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8612401485443115, "num_tokens": 290602209.0, "step": 7656 }, { "epoch": 0.9740491031675359, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4789340496063232, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8525964021682739, "num_tokens": 290645554.0, "step": 7657 }, { "epoch": 0.9741763134461264, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6121714115142822, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8705807328224182, "num_tokens": 290683982.0, "step": 7658 }, { "epoch": 0.974303523724717, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4996392726898193, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8534183502197266, "num_tokens": 290717720.0, "step": 7659 }, { "epoch": 0.9744307340033075, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6042652130126953, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8591042757034302, "num_tokens": 290759711.0, "step": 7660 }, { "epoch": 0.974557944281898, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4904444217681885, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8641034364700317, "num_tokens": 290795767.0, "step": 7661 }, { "epoch": 0.9746851545604885, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5985403060913086, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.868964433670044, "num_tokens": 290839148.0, "step": 7662 }, { "epoch": 0.974812364839079, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4786744117736816, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8655277490615845, "num_tokens": 290878325.0, "step": 7663 }, { "epoch": 0.9749395751176695, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00018215179443359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.620436906814575, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8582109212875366, "num_tokens": 290913053.0, "step": 7664 }, { "epoch": 0.97506678539626, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4690840244293213, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8684195280075073, "num_tokens": 290954984.0, "step": 7665 }, { "epoch": 0.9751939956748505, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.636275291442871, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8539400100708008, "num_tokens": 290996052.0, "step": 7666 }, { "epoch": 0.9753212059534411, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.465183973312378, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.8499119877815247, "num_tokens": 291034021.0, "step": 7667 }, { "epoch": 0.9754484162320316, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6566102504730225, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8733438849449158, "num_tokens": 291064216.0, "step": 7668 }, { "epoch": 0.975575626510622, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.452113389968872, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8696193695068359, "num_tokens": 291100119.0, "step": 7669 }, { "epoch": 0.9757028367892125, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6740827560424805, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8607669472694397, "num_tokens": 291136709.0, "step": 7670 }, { "epoch": 0.9758300470678031, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4694247245788574, "learning_rate": 1e-05, "loss": 0.516, "mean_token_accuracy": 0.8452777862548828, "num_tokens": 291175038.0, "step": 7671 }, { "epoch": 0.9759572573463936, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6897029876708984, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8645281791687012, "num_tokens": 291206950.0, "step": 7672 }, { "epoch": 0.9760844676249841, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.454141139984131, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.854473352432251, "num_tokens": 291247847.0, "step": 7673 }, { "epoch": 0.9762116779035747, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.702688455581665, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8752220869064331, "num_tokens": 291282262.0, "step": 7674 }, { "epoch": 0.9763388881821651, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.460836887359619, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.864037036895752, "num_tokens": 291315090.0, "step": 7675 }, { "epoch": 0.9764660984607556, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7042346000671387, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8698633909225464, "num_tokens": 291356380.0, "step": 7676 }, { "epoch": 0.9765933087393461, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4428398609161377, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8626141548156738, "num_tokens": 291401133.0, "step": 7677 }, { "epoch": 0.9767205190179367, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.730557918548584, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8650118708610535, "num_tokens": 291439999.0, "step": 7678 }, { "epoch": 0.9768477292965272, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.461594581604004, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8668882846832275, "num_tokens": 291476784.0, "step": 7679 }, { "epoch": 0.9769749395751177, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.737302780151367, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8730556964874268, "num_tokens": 291517677.0, "step": 7680 }, { "epoch": 0.9771021498537081, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4691224098205566, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8669053316116333, "num_tokens": 291551838.0, "step": 7681 }, { "epoch": 0.9772293601322987, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.741884708404541, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8716400861740112, "num_tokens": 291591192.0, "step": 7682 }, { "epoch": 0.9773565704108892, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4810266494750977, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8625966906547546, "num_tokens": 291626885.0, "step": 7683 }, { "epoch": 0.9774837806894797, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7393441200256348, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8648325204849243, "num_tokens": 291666421.0, "step": 7684 }, { "epoch": 0.9776109909680702, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4826555252075195, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8604056239128113, "num_tokens": 291704973.0, "step": 7685 }, { "epoch": 0.9777382012466608, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7397117614746094, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.857286810874939, "num_tokens": 291750279.0, "step": 7686 }, { "epoch": 0.9778654115252512, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4977028369903564, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8600841164588928, "num_tokens": 291784390.0, "step": 7687 }, { "epoch": 0.9779926218038417, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7220654487609863, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8702412247657776, "num_tokens": 291826385.0, "step": 7688 }, { "epoch": 0.9781198320824323, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5071187019348145, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8542750477790833, "num_tokens": 291858129.0, "step": 7689 }, { "epoch": 0.9782470423610228, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7101387977600098, "learning_rate": 1e-05, "loss": 0.5372, "mean_token_accuracy": 0.8487231731414795, "num_tokens": 291902763.0, "step": 7690 }, { "epoch": 0.9783742526396133, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4840848445892334, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8751805424690247, "num_tokens": 291940686.0, "step": 7691 }, { "epoch": 0.9785014629182038, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7007412910461426, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8632607460021973, "num_tokens": 291976849.0, "step": 7692 }, { "epoch": 0.9786286731967943, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4856979846954346, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8772976398468018, "num_tokens": 292014105.0, "step": 7693 }, { "epoch": 0.9787558834753848, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6956982612609863, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.865411639213562, "num_tokens": 292055083.0, "step": 7694 }, { "epoch": 0.9788830937539753, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4889626502990723, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8655496835708618, "num_tokens": 292089713.0, "step": 7695 }, { "epoch": 0.9790103040325658, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6935408115386963, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8517268896102905, "num_tokens": 292128139.0, "step": 7696 }, { "epoch": 0.9791375143111564, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4974446296691895, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8596072793006897, "num_tokens": 292164674.0, "step": 7697 }, { "epoch": 0.9792647245897469, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.683248996734619, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8652325868606567, "num_tokens": 292197175.0, "step": 7698 }, { "epoch": 0.9793919348683373, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.49875807762146, "learning_rate": 1e-05, "loss": 0.5003, "mean_token_accuracy": 0.8492069840431213, "num_tokens": 292237051.0, "step": 7699 }, { "epoch": 0.9795191451469278, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.671785831451416, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8724011778831482, "num_tokens": 292272817.0, "step": 7700 }, { "epoch": 0.9796463554255184, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.488274574279785, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8620663285255432, "num_tokens": 292310239.0, "step": 7701 }, { "epoch": 0.9797735657041089, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6599674224853516, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8613826632499695, "num_tokens": 292351090.0, "step": 7702 }, { "epoch": 0.9799007759826994, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4934608936309814, "learning_rate": 1e-05, "loss": 0.5238, "mean_token_accuracy": 0.843317985534668, "num_tokens": 292390500.0, "step": 7703 }, { "epoch": 0.98002798626129, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.672288417816162, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8549079895019531, "num_tokens": 292423483.0, "step": 7704 }, { "epoch": 0.9801551965398804, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4788758754730225, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8764691352844238, "num_tokens": 292460791.0, "step": 7705 }, { "epoch": 0.9802824068184709, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.653230905532837, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8782081604003906, "num_tokens": 292502214.0, "step": 7706 }, { "epoch": 0.9804096170970614, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4914002418518066, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.848804771900177, "num_tokens": 292541845.0, "step": 7707 }, { "epoch": 0.980536827375652, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6636099815368652, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8743237853050232, "num_tokens": 292573932.0, "step": 7708 }, { "epoch": 0.9806640376542425, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.471498489379883, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8697500228881836, "num_tokens": 292615932.0, "step": 7709 }, { "epoch": 0.980791247932833, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.694089651107788, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.8518287539482117, "num_tokens": 292648680.0, "step": 7710 }, { "epoch": 0.9809184582114235, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.467557430267334, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8755515217781067, "num_tokens": 292687987.0, "step": 7711 }, { "epoch": 0.981045668490014, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.687041997909546, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8597634434700012, "num_tokens": 292732243.0, "step": 7712 }, { "epoch": 0.9811728787686045, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.475153684616089, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8690849542617798, "num_tokens": 292764869.0, "step": 7713 }, { "epoch": 0.981300089047195, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6853904724121094, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8717696666717529, "num_tokens": 292806401.0, "step": 7714 }, { "epoch": 0.9814272993257855, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4724280834198, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8706803321838379, "num_tokens": 292843165.0, "step": 7715 }, { "epoch": 0.9815545096043761, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6920530796051025, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8623216152191162, "num_tokens": 292885092.0, "step": 7716 }, { "epoch": 0.9816817198829666, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4795384407043457, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8538058996200562, "num_tokens": 292920595.0, "step": 7717 }, { "epoch": 0.981808930161557, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.713772773742676, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.854360818862915, "num_tokens": 292954079.0, "step": 7718 }, { "epoch": 0.9819361404401475, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4835379123687744, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8683643341064453, "num_tokens": 292985554.0, "step": 7719 }, { "epoch": 0.9820633507187381, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7026619911193848, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8637796640396118, "num_tokens": 293021832.0, "step": 7720 }, { "epoch": 0.9821905609973286, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4855387210845947, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8615056276321411, "num_tokens": 293054220.0, "step": 7721 }, { "epoch": 0.9823177712759191, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.690297842025757, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8529639840126038, "num_tokens": 293099348.0, "step": 7722 }, { "epoch": 0.9824449815545097, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4826245307922363, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8612395524978638, "num_tokens": 293134801.0, "step": 7723 }, { "epoch": 0.9825721918331001, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.70198392868042, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8403555154800415, "num_tokens": 293172023.0, "step": 7724 }, { "epoch": 0.9826994021116906, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4715380668640137, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.86332106590271, "num_tokens": 293210019.0, "step": 7725 }, { "epoch": 0.9828266123902811, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.696439504623413, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8628915548324585, "num_tokens": 293249614.0, "step": 7726 }, { "epoch": 0.9829538226688717, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.471907615661621, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8728106021881104, "num_tokens": 293289860.0, "step": 7727 }, { "epoch": 0.9830810329474622, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7033562660217285, "learning_rate": 1e-05, "loss": 0.5133, "mean_token_accuracy": 0.8502930402755737, "num_tokens": 293330337.0, "step": 7728 }, { "epoch": 0.9832082432260527, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4705677032470703, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8639962673187256, "num_tokens": 293370062.0, "step": 7729 }, { "epoch": 0.9833354535046431, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7126708030700684, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8608478903770447, "num_tokens": 293409544.0, "step": 7730 }, { "epoch": 0.9834626637832337, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.478513240814209, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8495283126831055, "num_tokens": 293453924.0, "step": 7731 }, { "epoch": 0.9835898740618242, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7080116271972656, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8672839403152466, "num_tokens": 293493759.0, "step": 7732 }, { "epoch": 0.9837170843404147, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.490985631942749, "learning_rate": 1e-05, "loss": 0.528, "mean_token_accuracy": 0.8411169052124023, "num_tokens": 293530155.0, "step": 7733 }, { "epoch": 0.9838442946190052, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6956627368927, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8605153560638428, "num_tokens": 293573083.0, "step": 7734 }, { "epoch": 0.9839715048975958, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4864909648895264, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8707878589630127, "num_tokens": 293608312.0, "step": 7735 }, { "epoch": 0.9840987151761862, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.695732355117798, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8608916997909546, "num_tokens": 293645036.0, "step": 7736 }, { "epoch": 0.9842259254547767, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5078306198120117, "learning_rate": 1e-05, "loss": 0.5496, "mean_token_accuracy": 0.8383255004882812, "num_tokens": 293679063.0, "step": 7737 }, { "epoch": 0.9843531357333672, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6682093143463135, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8655309081077576, "num_tokens": 293720239.0, "step": 7738 }, { "epoch": 0.9844803460119578, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4908645153045654, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8782756328582764, "num_tokens": 293757154.0, "step": 7739 }, { "epoch": 0.9846075562905483, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6683385372161865, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8515933752059937, "num_tokens": 293797181.0, "step": 7740 }, { "epoch": 0.9847347665691388, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4911279678344727, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8604882955551147, "num_tokens": 293838130.0, "step": 7741 }, { "epoch": 0.9848619768477292, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6701905727386475, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8587222099304199, "num_tokens": 293877341.0, "step": 7742 }, { "epoch": 0.9849891871263198, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.475922107696533, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8673856258392334, "num_tokens": 293913764.0, "step": 7743 }, { "epoch": 0.9851163974049103, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.667808771133423, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8563585877418518, "num_tokens": 293954142.0, "step": 7744 }, { "epoch": 0.9852436076835008, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4713211059570312, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8557019233703613, "num_tokens": 293999016.0, "step": 7745 }, { "epoch": 0.9853708179620914, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.685328483581543, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8635613918304443, "num_tokens": 294033545.0, "step": 7746 }, { "epoch": 0.9854980282406819, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4672186374664307, "learning_rate": 1e-05, "loss": 0.4098, "mean_token_accuracy": 0.8769856095314026, "num_tokens": 294076763.0, "step": 7747 }, { "epoch": 0.9856252385192723, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018405914306640625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6773014068603516, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8841163516044617, "num_tokens": 294114661.0, "step": 7748 }, { "epoch": 0.9857524487978628, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.482388496398926, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8579752445220947, "num_tokens": 294148819.0, "step": 7749 }, { "epoch": 0.9858796590764534, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.678317070007324, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8666154146194458, "num_tokens": 294190820.0, "step": 7750 }, { "epoch": 0.9860068693550439, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4664559364318848, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8624445796012878, "num_tokens": 294230814.0, "step": 7751 }, { "epoch": 0.9861340796336344, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6791090965270996, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8608306646347046, "num_tokens": 294271811.0, "step": 7752 }, { "epoch": 0.986261289912225, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.46443772315979, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8603658676147461, "num_tokens": 294313901.0, "step": 7753 }, { "epoch": 0.9863885001908154, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.685225486755371, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8523297309875488, "num_tokens": 294353275.0, "step": 7754 }, { "epoch": 0.9865157104694059, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4749715328216553, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8806003332138062, "num_tokens": 294390569.0, "step": 7755 }, { "epoch": 0.9866429207479964, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6783132553100586, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8510981798171997, "num_tokens": 294433050.0, "step": 7756 }, { "epoch": 0.986770131026587, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4677743911743164, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.871369481086731, "num_tokens": 294471198.0, "step": 7757 }, { "epoch": 0.9868973413051775, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6876907348632812, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.866814136505127, "num_tokens": 294511855.0, "step": 7758 }, { "epoch": 0.987024551583768, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4753332138061523, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8675620555877686, "num_tokens": 294547056.0, "step": 7759 }, { "epoch": 0.9871517618623584, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7022318840026855, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8590134382247925, "num_tokens": 294584516.0, "step": 7760 }, { "epoch": 0.987278972140949, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.468043088912964, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8613702654838562, "num_tokens": 294622253.0, "step": 7761 }, { "epoch": 0.9874061824195395, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.684548854827881, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8709514141082764, "num_tokens": 294664842.0, "step": 7762 }, { "epoch": 0.98753339269813, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.481032371520996, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8563944697380066, "num_tokens": 294707135.0, "step": 7763 }, { "epoch": 0.9876606029767205, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6831703186035156, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8490162491798401, "num_tokens": 294751356.0, "step": 7764 }, { "epoch": 0.9877878132553111, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4822418689727783, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8558192253112793, "num_tokens": 294789444.0, "step": 7765 }, { "epoch": 0.9879150235339016, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.692812442779541, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8534280061721802, "num_tokens": 294823247.0, "step": 7766 }, { "epoch": 0.988042233812492, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4730706214904785, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8510633707046509, "num_tokens": 294862932.0, "step": 7767 }, { "epoch": 0.9881694440910825, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6875908374786377, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8621102571487427, "num_tokens": 294902686.0, "step": 7768 }, { "epoch": 0.9882966543696731, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.46866774559021, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8696351647377014, "num_tokens": 294942028.0, "step": 7769 }, { "epoch": 0.9884238646482636, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.703373908996582, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8621089458465576, "num_tokens": 294977699.0, "step": 7770 }, { "epoch": 0.9885510749268541, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4796199798583984, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8613351583480835, "num_tokens": 295014441.0, "step": 7771 }, { "epoch": 0.9886782852054447, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.691761016845703, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8712213039398193, "num_tokens": 295053569.0, "step": 7772 }, { "epoch": 0.9888054954840351, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4938254356384277, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8601182699203491, "num_tokens": 295083607.0, "step": 7773 }, { "epoch": 0.9889327057626256, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.701991319656372, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8514755964279175, "num_tokens": 295118074.0, "step": 7774 }, { "epoch": 0.9890599160412161, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4780261516571045, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8644586205482483, "num_tokens": 295157926.0, "step": 7775 }, { "epoch": 0.9891871263198067, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6847245693206787, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8609301447868347, "num_tokens": 295201297.0, "step": 7776 }, { "epoch": 0.9893143365983972, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.484553575515747, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8568562865257263, "num_tokens": 295238178.0, "step": 7777 }, { "epoch": 0.9894415468769877, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.685297966003418, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8626866936683655, "num_tokens": 295273843.0, "step": 7778 }, { "epoch": 0.9895687571555781, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.484426259994507, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8662831783294678, "num_tokens": 295308703.0, "step": 7779 }, { "epoch": 0.9896959674341687, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6782753467559814, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8608644008636475, "num_tokens": 295347710.0, "step": 7780 }, { "epoch": 0.9898231777127592, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4811015129089355, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8705617189407349, "num_tokens": 295382255.0, "step": 7781 }, { "epoch": 0.9899503879913497, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6793925762176514, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8670086860656738, "num_tokens": 295418185.0, "step": 7782 }, { "epoch": 0.9900775982699402, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.49591064453125, "learning_rate": 1e-05, "loss": 0.5243, "mean_token_accuracy": 0.843096911907196, "num_tokens": 295455124.0, "step": 7783 }, { "epoch": 0.9902048085485308, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6635212898254395, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8500479459762573, "num_tokens": 295498058.0, "step": 7784 }, { "epoch": 0.9903320188271212, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.485337734222412, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8764268159866333, "num_tokens": 295534623.0, "step": 7785 }, { "epoch": 0.9904592291057117, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6525228023529053, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8537321090698242, "num_tokens": 295580064.0, "step": 7786 }, { "epoch": 0.9905864393843022, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4873368740081787, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8500086069107056, "num_tokens": 295620720.0, "step": 7787 }, { "epoch": 0.9907136496628928, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6529035568237305, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8677902221679688, "num_tokens": 295662157.0, "step": 7788 }, { "epoch": 0.9908408599414833, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4834883213043213, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8617885112762451, "num_tokens": 295700215.0, "step": 7789 }, { "epoch": 0.9909680702200738, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6473028659820557, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8843240141868591, "num_tokens": 295738363.0, "step": 7790 }, { "epoch": 0.9910952804986642, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4837653636932373, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8744826316833496, "num_tokens": 295773911.0, "step": 7791 }, { "epoch": 0.9912224907772548, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6627140045166016, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8608755469322205, "num_tokens": 295809341.0, "step": 7792 }, { "epoch": 0.9913497010558453, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4710824489593506, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8524528741836548, "num_tokens": 295855952.0, "step": 7793 }, { "epoch": 0.9914769113344358, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6633787155151367, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8648425340652466, "num_tokens": 295892047.0, "step": 7794 }, { "epoch": 0.9916041216130264, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4940085411071777, "learning_rate": 1e-05, "loss": 0.5164, "mean_token_accuracy": 0.8438906669616699, "num_tokens": 295926259.0, "step": 7795 }, { "epoch": 0.9917313318916169, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6600518226623535, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8571284413337708, "num_tokens": 295965584.0, "step": 7796 }, { "epoch": 0.9918585421702073, "ewc_loss": 0.05419921875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.5762786865234375e-05, "grad_norm": 2.4705069065093994, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8709186911582947, "num_tokens": 296002754.0, "step": 7797 }, { "epoch": 0.9919857524487978, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.673572301864624, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8613367676734924, "num_tokens": 296042161.0, "step": 7798 }, { "epoch": 0.9921129627273884, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.483611583709717, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8527530431747437, "num_tokens": 296079153.0, "step": 7799 }, { "epoch": 0.9922401730059789, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.661297559738159, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8531919717788696, "num_tokens": 296124064.0, "step": 7800 }, { "epoch": 0.9923673832845694, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.464742660522461, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8685024976730347, "num_tokens": 296166035.0, "step": 7801 }, { "epoch": 0.9924945935631599, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6729953289031982, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8783407211303711, "num_tokens": 296206566.0, "step": 7802 }, { "epoch": 0.9926218038417504, "ewc_loss": 0.054443359375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.600120544433594e-05, "grad_norm": 2.4672529697418213, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.867468535900116, "num_tokens": 296243587.0, "step": 7803 }, { "epoch": 0.9927490141203409, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6928577423095703, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8593414425849915, "num_tokens": 296282704.0, "step": 7804 }, { "epoch": 0.9928762243989314, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.479595184326172, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8458230495452881, "num_tokens": 296317964.0, "step": 7805 }, { "epoch": 0.993003434677522, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.701354742050171, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8537037372589111, "num_tokens": 296352861.0, "step": 7806 }, { "epoch": 0.9931306449561125, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.482063055038452, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8609920740127563, "num_tokens": 296391302.0, "step": 7807 }, { "epoch": 0.993257855234703, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.7245993614196777, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8566895723342896, "num_tokens": 296428026.0, "step": 7808 }, { "epoch": 0.9933850655132934, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.465810775756836, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8618403077125549, "num_tokens": 296468866.0, "step": 7809 }, { "epoch": 0.993512275791884, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.703826904296875, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8628581762313843, "num_tokens": 296514806.0, "step": 7810 }, { "epoch": 0.9936394860704745, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.470858097076416, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.867751955986023, "num_tokens": 296552328.0, "step": 7811 }, { "epoch": 0.993766696349065, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7252488136291504, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8558540344238281, "num_tokens": 296589908.0, "step": 7812 }, { "epoch": 0.9938939066276555, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4789016246795654, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.851273775100708, "num_tokens": 296631147.0, "step": 7813 }, { "epoch": 0.9940211169062461, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.718528985977173, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8538092374801636, "num_tokens": 296673633.0, "step": 7814 }, { "epoch": 0.9941483271848366, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.470815658569336, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8619794845581055, "num_tokens": 296714782.0, "step": 7815 }, { "epoch": 0.994275537463427, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.728071451187134, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8563907742500305, "num_tokens": 296751556.0, "step": 7816 }, { "epoch": 0.9944027477420175, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.494565010070801, "learning_rate": 1e-05, "loss": 0.5231, "mean_token_accuracy": 0.8410214185714722, "num_tokens": 296788319.0, "step": 7817 }, { "epoch": 0.9945299580206081, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7259490489959717, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8542042970657349, "num_tokens": 296823513.0, "step": 7818 }, { "epoch": 0.9946571682991986, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.495298147201538, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8559301495552063, "num_tokens": 296862059.0, "step": 7819 }, { "epoch": 0.9947843785777891, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7148940563201904, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8654186725616455, "num_tokens": 296898051.0, "step": 7820 }, { "epoch": 0.9949115888563796, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5305449962615967, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8663867712020874, "num_tokens": 296935266.0, "step": 7821 }, { "epoch": 0.9950387991349701, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.682380437850952, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8472546935081482, "num_tokens": 296971653.0, "step": 7822 }, { "epoch": 0.9951660094135606, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5038957595825195, "learning_rate": 1e-05, "loss": 0.5455, "mean_token_accuracy": 0.8362201452255249, "num_tokens": 297015825.0, "step": 7823 }, { "epoch": 0.9952932196921511, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6610140800476074, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8487371206283569, "num_tokens": 297059115.0, "step": 7824 }, { "epoch": 0.9954204299707416, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5078935623168945, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8559638261795044, "num_tokens": 297092634.0, "step": 7825 }, { "epoch": 0.9955476402493322, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.651556968688965, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8615847229957581, "num_tokens": 297137154.0, "step": 7826 }, { "epoch": 0.9956748505279227, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5098648071289062, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.850836992263794, "num_tokens": 297175029.0, "step": 7827 }, { "epoch": 0.9958020608065131, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6546335220336914, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8631612658500671, "num_tokens": 297207397.0, "step": 7828 }, { "epoch": 0.9959292710851037, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4847490787506104, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8571726679801941, "num_tokens": 297250114.0, "step": 7829 }, { "epoch": 0.9960564813636942, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6553590297698975, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.866682231426239, "num_tokens": 297289765.0, "step": 7830 }, { "epoch": 0.9961836916422847, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.479668378829956, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8815778493881226, "num_tokens": 297327359.0, "step": 7831 }, { "epoch": 0.9963109019208752, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6744251251220703, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8676332235336304, "num_tokens": 297360851.0, "step": 7832 }, { "epoch": 0.9964381121994658, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4895899295806885, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8592687845230103, "num_tokens": 297396957.0, "step": 7833 }, { "epoch": 0.9965653224780562, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.682849645614624, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8609219789505005, "num_tokens": 297427900.0, "step": 7834 }, { "epoch": 0.9966925327566467, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.47403883934021, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8501153588294983, "num_tokens": 297464910.0, "step": 7835 }, { "epoch": 0.9968197430352372, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6881906986236572, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8741355538368225, "num_tokens": 297497585.0, "step": 7836 }, { "epoch": 0.9969469533138278, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.506455659866333, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8524230718612671, "num_tokens": 297537074.0, "step": 7837 }, { "epoch": 0.9970741635924183, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6743035316467285, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8705902695655823, "num_tokens": 297570848.0, "step": 7838 }, { "epoch": 0.9972013738710088, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4862053394317627, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.851791501045227, "num_tokens": 297607520.0, "step": 7839 }, { "epoch": 0.9973285841495992, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6584537029266357, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8658705949783325, "num_tokens": 297648442.0, "step": 7840 }, { "epoch": 0.9974557944281898, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.499660015106201, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8428446650505066, "num_tokens": 297686163.0, "step": 7841 }, { "epoch": 0.9975830047067803, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.646923780441284, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8719645738601685, "num_tokens": 297724191.0, "step": 7842 }, { "epoch": 0.9977102149853708, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4893784523010254, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8544381856918335, "num_tokens": 297766205.0, "step": 7843 }, { "epoch": 0.9978374252639614, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.651648998260498, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8707365393638611, "num_tokens": 297802153.0, "step": 7844 }, { "epoch": 0.9979646355425519, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.476799726486206, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8684141039848328, "num_tokens": 297841252.0, "step": 7845 }, { "epoch": 0.9980918458211423, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.650507926940918, "learning_rate": 1e-05, "loss": 0.3717, "mean_token_accuracy": 0.8900565505027771, "num_tokens": 297875460.0, "step": 7846 }, { "epoch": 0.9982190560997328, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.508720636367798, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8599566221237183, "num_tokens": 297908534.0, "step": 7847 }, { "epoch": 0.9983462663783234, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6505820751190186, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.8390135765075684, "num_tokens": 297950798.0, "step": 7848 }, { "epoch": 0.9984734766569139, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4777491092681885, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8674778938293457, "num_tokens": 297989696.0, "step": 7849 }, { "epoch": 0.9986006869355044, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6448862552642822, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.858206033706665, "num_tokens": 298033702.0, "step": 7850 }, { "epoch": 0.9987278972140949, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.468172788619995, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8779280185699463, "num_tokens": 298071867.0, "step": 7851 }, { "epoch": 0.9988551074926854, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.673809766769409, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8510587215423584, "num_tokens": 298108726.0, "step": 7852 }, { "epoch": 0.9989823177712759, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.476670503616333, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8560163378715515, "num_tokens": 298145174.0, "step": 7853 }, { "epoch": 0.9991095280498664, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6791787147521973, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8741916418075562, "num_tokens": 298181300.0, "step": 7854 }, { "epoch": 0.9992367383284569, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.0001850128173828125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.463158369064331, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.862028181552887, "num_tokens": 298220592.0, "step": 7855 }, { "epoch": 0.9993639486070475, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6921379566192627, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8571691513061523, "num_tokens": 298260796.0, "step": 7856 }, { "epoch": 0.999491158885638, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.515125274658203, "learning_rate": 1e-05, "loss": 0.5078, "mean_token_accuracy": 0.8441537618637085, "num_tokens": 298300640.0, "step": 7857 }, { "epoch": 0.9996183691642284, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.663297414779663, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8770174384117126, "num_tokens": 298336433.0, "step": 7858 }, { "epoch": 0.9997455794428189, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4882049560546875, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8739802241325378, "num_tokens": 298370355.0, "step": 7859 }, { "epoch": 0.9998727897214095, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.66072678565979, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8621099591255188, "num_tokens": 298406262.0, "step": 7860 }, { "epoch": 1.0, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4993245601654053, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8423924446105957, "num_tokens": 298443310.0, "step": 7861 }, { "epoch": 1.0001272102785905, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.64892578125, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8643603324890137, "num_tokens": 298482309.0, "step": 7862 }, { "epoch": 1.000254420557181, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4780471324920654, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8637091517448425, "num_tokens": 298522555.0, "step": 7863 }, { "epoch": 1.0003816308357716, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6477487087249756, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8819759488105774, "num_tokens": 298559559.0, "step": 7864 }, { "epoch": 1.0005088411143621, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5709593296051025, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8677836656570435, "num_tokens": 298596604.0, "step": 7865 }, { "epoch": 1.0006360513929526, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5919101238250732, "learning_rate": 1e-05, "loss": 0.3844, "mean_token_accuracy": 0.884428858757019, "num_tokens": 298634590.0, "step": 7866 }, { "epoch": 1.0007632616715432, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.510911226272583, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8792060613632202, "num_tokens": 298670714.0, "step": 7867 }, { "epoch": 1.0008904719501335, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5806503295898438, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8699266910552979, "num_tokens": 298713924.0, "step": 7868 }, { "epoch": 1.001017682228724, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5228657722473145, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.872075080871582, "num_tokens": 298749537.0, "step": 7869 }, { "epoch": 1.0011448925073145, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.573580503463745, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8630279898643494, "num_tokens": 298789887.0, "step": 7870 }, { "epoch": 1.001272102785905, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5195014476776123, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8709495067596436, "num_tokens": 298826790.0, "step": 7871 }, { "epoch": 1.0013993130644956, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6226818561553955, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8624472618103027, "num_tokens": 298860427.0, "step": 7872 }, { "epoch": 1.0015265233430861, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.511875629425049, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8707097768783569, "num_tokens": 298892627.0, "step": 7873 }, { "epoch": 1.0016537336216766, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.6061649322509766, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8585115671157837, "num_tokens": 298930513.0, "step": 7874 }, { "epoch": 1.0017809439002672, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4755072593688965, "learning_rate": 1e-05, "loss": 0.375, "mean_token_accuracy": 0.8894000053405762, "num_tokens": 298971234.0, "step": 7875 }, { "epoch": 1.0019081541788577, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6132965087890625, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8740965723991394, "num_tokens": 299013224.0, "step": 7876 }, { "epoch": 1.0020353644574482, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.482318639755249, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8457431197166443, "num_tokens": 299055093.0, "step": 7877 }, { "epoch": 1.0021625747360388, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.642962694168091, "learning_rate": 1e-05, "loss": 0.3881, "mean_token_accuracy": 0.8822274208068848, "num_tokens": 299085368.0, "step": 7878 }, { "epoch": 1.0022897850146293, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4754798412323, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8632937669754028, "num_tokens": 299128279.0, "step": 7879 }, { "epoch": 1.0024169952932196, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.666893482208252, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8703694343566895, "num_tokens": 299161714.0, "step": 7880 }, { "epoch": 1.0025442055718101, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4624600410461426, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8600313663482666, "num_tokens": 299200565.0, "step": 7881 }, { "epoch": 1.0026714158504006, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6809935569763184, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8693821430206299, "num_tokens": 299242242.0, "step": 7882 }, { "epoch": 1.0027986261289912, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.464808225631714, "learning_rate": 1e-05, "loss": 0.3971, "mean_token_accuracy": 0.8823573589324951, "num_tokens": 299278141.0, "step": 7883 }, { "epoch": 1.0029258364075817, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.703943967819214, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8744138479232788, "num_tokens": 299312413.0, "step": 7884 }, { "epoch": 1.0030530466861722, "ewc_loss": 0.054931640625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4550912380218506, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8785696029663086, "num_tokens": 299348158.0, "step": 7885 }, { "epoch": 1.0031802569647628, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.709282875061035, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.87611985206604, "num_tokens": 299386963.0, "step": 7886 }, { "epoch": 1.0033074672433533, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4541592597961426, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8771677017211914, "num_tokens": 299424929.0, "step": 7887 }, { "epoch": 1.0034346775219438, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7477006912231445, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8632431030273438, "num_tokens": 299461588.0, "step": 7888 }, { "epoch": 1.0035618878005343, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4627435207366943, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8786145448684692, "num_tokens": 299498226.0, "step": 7889 }, { "epoch": 1.0036890980791249, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.762998342514038, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8721692562103271, "num_tokens": 299536806.0, "step": 7890 }, { "epoch": 1.0038163083577154, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4551799297332764, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8653151988983154, "num_tokens": 299583658.0, "step": 7891 }, { "epoch": 1.0039435186363057, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.781101703643799, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8625915050506592, "num_tokens": 299621381.0, "step": 7892 }, { "epoch": 1.0040707289148962, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.478854179382324, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8656224608421326, "num_tokens": 299661055.0, "step": 7893 }, { "epoch": 1.0041979391934868, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7803990840911865, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8734045624732971, "num_tokens": 299702500.0, "step": 7894 }, { "epoch": 1.0043251494720773, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4824535846710205, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8691679239273071, "num_tokens": 299742590.0, "step": 7895 }, { "epoch": 1.0044523597506678, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.792710065841675, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8673061728477478, "num_tokens": 299777716.0, "step": 7896 }, { "epoch": 1.0045795700292584, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5129213333129883, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8810821771621704, "num_tokens": 299808829.0, "step": 7897 }, { "epoch": 1.0047067803078489, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7784578800201416, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8705786466598511, "num_tokens": 299845582.0, "step": 7898 }, { "epoch": 1.0048339905864394, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5017592906951904, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8732318878173828, "num_tokens": 299886373.0, "step": 7899 }, { "epoch": 1.00496120086503, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.778585433959961, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8728077411651611, "num_tokens": 299917464.0, "step": 7900 }, { "epoch": 1.0050884111436205, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5290088653564453, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8719028234481812, "num_tokens": 299953455.0, "step": 7901 }, { "epoch": 1.005215621422211, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7485122680664062, "learning_rate": 1e-05, "loss": 0.4844, "mean_token_accuracy": 0.8612396717071533, "num_tokens": 299993707.0, "step": 7902 }, { "epoch": 1.0053428317008015, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.521331548690796, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8671773672103882, "num_tokens": 300033021.0, "step": 7903 }, { "epoch": 1.0054700419793918, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7225356101989746, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8691571950912476, "num_tokens": 300072497.0, "step": 7904 }, { "epoch": 1.0055972522579824, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.513899803161621, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8786331415176392, "num_tokens": 300113510.0, "step": 7905 }, { "epoch": 1.0057244625365729, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7196855545043945, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8660035133361816, "num_tokens": 300151524.0, "step": 7906 }, { "epoch": 1.0058516728151634, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5514514446258545, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8695268630981445, "num_tokens": 300184281.0, "step": 7907 }, { "epoch": 1.005978883093754, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6842169761657715, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8590672016143799, "num_tokens": 300222772.0, "step": 7908 }, { "epoch": 1.0061060933723445, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5319724082946777, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8627597093582153, "num_tokens": 300257347.0, "step": 7909 }, { "epoch": 1.006233303650935, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018596649169921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6525776386260986, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.864020824432373, "num_tokens": 300300678.0, "step": 7910 }, { "epoch": 1.0063605139295255, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5184860229492188, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8764539361000061, "num_tokens": 300339851.0, "step": 7911 }, { "epoch": 1.006487724208116, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.644629955291748, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8798510432243347, "num_tokens": 300379753.0, "step": 7912 }, { "epoch": 1.0066149344867066, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.520296573638916, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8711963891983032, "num_tokens": 300418235.0, "step": 7913 }, { "epoch": 1.006742144765297, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6387887001037598, "learning_rate": 1e-05, "loss": 0.387, "mean_token_accuracy": 0.8832240104675293, "num_tokens": 300450237.0, "step": 7914 }, { "epoch": 1.0068693550438876, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5042004585266113, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8697229027748108, "num_tokens": 300489416.0, "step": 7915 }, { "epoch": 1.0069965653224782, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.640263080596924, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8791559934616089, "num_tokens": 300526180.0, "step": 7916 }, { "epoch": 1.0071237756010685, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4957656860351562, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8702622652053833, "num_tokens": 300566327.0, "step": 7917 }, { "epoch": 1.007250985879659, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6480660438537598, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8628857135772705, "num_tokens": 300610136.0, "step": 7918 }, { "epoch": 1.0073781961582495, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4949865341186523, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8685309886932373, "num_tokens": 300647238.0, "step": 7919 }, { "epoch": 1.00750540643684, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.648441791534424, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8747141361236572, "num_tokens": 300684291.0, "step": 7920 }, { "epoch": 1.0076326167154306, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5035715103149414, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8657316565513611, "num_tokens": 300718919.0, "step": 7921 }, { "epoch": 1.0077598269940211, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6507182121276855, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8675917387008667, "num_tokens": 300752353.0, "step": 7922 }, { "epoch": 1.0078870372726116, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4811599254608154, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.875112771987915, "num_tokens": 300789214.0, "step": 7923 }, { "epoch": 1.0080142475512022, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.664728879928589, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8628659248352051, "num_tokens": 300825629.0, "step": 7924 }, { "epoch": 1.0081414578297927, "ewc_loss": 0.0546875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4697186946868896, "learning_rate": 1e-05, "loss": 0.3901, "mean_token_accuracy": 0.8850513696670532, "num_tokens": 300863908.0, "step": 7925 }, { "epoch": 1.0082686681083832, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.675617218017578, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8653753399848938, "num_tokens": 300905458.0, "step": 7926 }, { "epoch": 1.0083958783869738, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.468626022338867, "learning_rate": 1e-05, "loss": 0.3917, "mean_token_accuracy": 0.8837486505508423, "num_tokens": 300943906.0, "step": 7927 }, { "epoch": 1.0085230886655643, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6885602474212646, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8740965723991394, "num_tokens": 300979371.0, "step": 7928 }, { "epoch": 1.0086502989441546, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4679930210113525, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8700273633003235, "num_tokens": 301017575.0, "step": 7929 }, { "epoch": 1.0087775092227451, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7114293575286865, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8734558820724487, "num_tokens": 301053130.0, "step": 7930 }, { "epoch": 1.0089047195013356, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.468254804611206, "learning_rate": 1e-05, "loss": 0.3843, "mean_token_accuracy": 0.8834275603294373, "num_tokens": 301090740.0, "step": 7931 }, { "epoch": 1.0090319297799262, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.713381767272949, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8713129758834839, "num_tokens": 301133539.0, "step": 7932 }, { "epoch": 1.0091591400585167, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4953465461730957, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8485718369483948, "num_tokens": 301168243.0, "step": 7933 }, { "epoch": 1.0092863503371072, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.725755214691162, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.876689076423645, "num_tokens": 301204444.0, "step": 7934 }, { "epoch": 1.0094135606156978, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.465400218963623, "learning_rate": 1e-05, "loss": 0.3835, "mean_token_accuracy": 0.8870808482170105, "num_tokens": 301245768.0, "step": 7935 }, { "epoch": 1.0095407708942883, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.737396240234375, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8579559326171875, "num_tokens": 301282381.0, "step": 7936 }, { "epoch": 1.0096679811728788, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.467365026473999, "learning_rate": 1e-05, "loss": 0.4012, "mean_token_accuracy": 0.881099283695221, "num_tokens": 301322538.0, "step": 7937 }, { "epoch": 1.0097951914514693, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7434465885162354, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8748940825462341, "num_tokens": 301357527.0, "step": 7938 }, { "epoch": 1.0099224017300599, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4823989868164062, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8528571128845215, "num_tokens": 301399609.0, "step": 7939 }, { "epoch": 1.0100496120086504, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7496306896209717, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8712677359580994, "num_tokens": 301439271.0, "step": 7940 }, { "epoch": 1.0101768222872407, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.487990617752075, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8701382875442505, "num_tokens": 301476895.0, "step": 7941 }, { "epoch": 1.0103040325658312, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.733537435531616, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8628808259963989, "num_tokens": 301519649.0, "step": 7942 }, { "epoch": 1.0104312428444218, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.510190010070801, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8618267774581909, "num_tokens": 301557132.0, "step": 7943 }, { "epoch": 1.0105584531230123, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7036643028259277, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8673536777496338, "num_tokens": 301603604.0, "step": 7944 }, { "epoch": 1.0106856634016028, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.51271390914917, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8602066040039062, "num_tokens": 301643797.0, "step": 7945 }, { "epoch": 1.0108128736801933, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.708681583404541, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8776273727416992, "num_tokens": 301676658.0, "step": 7946 }, { "epoch": 1.0109400839587839, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5021793842315674, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8800061941146851, "num_tokens": 301714749.0, "step": 7947 }, { "epoch": 1.0110672942373744, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000186920166015625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6916258335113525, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.873755931854248, "num_tokens": 301754116.0, "step": 7948 }, { "epoch": 1.011194504515965, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4956257343292236, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8811609148979187, "num_tokens": 301791441.0, "step": 7949 }, { "epoch": 1.0113217147945555, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.697047710418701, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8606138229370117, "num_tokens": 301830971.0, "step": 7950 }, { "epoch": 1.011448925073146, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.504786729812622, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8667291402816772, "num_tokens": 301867848.0, "step": 7951 }, { "epoch": 1.0115761353517365, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.674578905105591, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8686380982398987, "num_tokens": 301912700.0, "step": 7952 }, { "epoch": 1.0117033456303268, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5100739002227783, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8749316334724426, "num_tokens": 301945890.0, "step": 7953 }, { "epoch": 1.0118305559089174, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6708173751831055, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8812544345855713, "num_tokens": 301980095.0, "step": 7954 }, { "epoch": 1.0119577661875079, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5203897953033447, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.87126225233078, "num_tokens": 302011855.0, "step": 7955 }, { "epoch": 1.0120849764660984, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6564438343048096, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.869587242603302, "num_tokens": 302048299.0, "step": 7956 }, { "epoch": 1.012212186744689, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.509868860244751, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8682427406311035, "num_tokens": 302082628.0, "step": 7957 }, { "epoch": 1.0123393970232795, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6431031227111816, "learning_rate": 1e-05, "loss": 0.4003, "mean_token_accuracy": 0.8813074231147766, "num_tokens": 302119832.0, "step": 7958 }, { "epoch": 1.01246660730187, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.485255002975464, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8758159279823303, "num_tokens": 302162930.0, "step": 7959 }, { "epoch": 1.0125938175804605, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6613762378692627, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8723934888839722, "num_tokens": 302202250.0, "step": 7960 }, { "epoch": 1.012721027859051, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4939568042755127, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8651386499404907, "num_tokens": 302239446.0, "step": 7961 }, { "epoch": 1.0128482381376416, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6448235511779785, "learning_rate": 1e-05, "loss": 0.3878, "mean_token_accuracy": 0.8857948184013367, "num_tokens": 302279949.0, "step": 7962 }, { "epoch": 1.012975448416232, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.499999761581421, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8627756237983704, "num_tokens": 302318128.0, "step": 7963 }, { "epoch": 1.0131026586948226, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.653606653213501, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8701938390731812, "num_tokens": 302357353.0, "step": 7964 }, { "epoch": 1.0132298689734132, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4790191650390625, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8740352392196655, "num_tokens": 302399843.0, "step": 7965 }, { "epoch": 1.0133570792520035, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.661801815032959, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8696056604385376, "num_tokens": 302440081.0, "step": 7966 }, { "epoch": 1.013484289530594, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.490495443344116, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8599468469619751, "num_tokens": 302479061.0, "step": 7967 }, { "epoch": 1.0136114998091845, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.670278549194336, "learning_rate": 1e-05, "loss": 0.4015, "mean_token_accuracy": 0.8804429769515991, "num_tokens": 302514490.0, "step": 7968 }, { "epoch": 1.013738710087775, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4849436283111572, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8758330941200256, "num_tokens": 302552555.0, "step": 7969 }, { "epoch": 1.0138659203663656, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00018787384033203125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6799092292785645, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8615036606788635, "num_tokens": 302589296.0, "step": 7970 }, { "epoch": 1.013993130644956, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4710967540740967, "learning_rate": 1e-05, "loss": 0.3753, "mean_token_accuracy": 0.8876559138298035, "num_tokens": 302626085.0, "step": 7971 }, { "epoch": 1.0141203409235466, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6842634677886963, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8746413588523865, "num_tokens": 302662671.0, "step": 7972 }, { "epoch": 1.0142475512021372, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4797890186309814, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8725985884666443, "num_tokens": 302706112.0, "step": 7973 }, { "epoch": 1.0143747614807277, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6946897506713867, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8600869178771973, "num_tokens": 302744942.0, "step": 7974 }, { "epoch": 1.0145019717593182, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4695215225219727, "learning_rate": 1e-05, "loss": 0.3837, "mean_token_accuracy": 0.885147213935852, "num_tokens": 302784085.0, "step": 7975 }, { "epoch": 1.0146291820379088, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7253506183624268, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8761307597160339, "num_tokens": 302821336.0, "step": 7976 }, { "epoch": 1.0147563923164993, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4868712425231934, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8721799850463867, "num_tokens": 302857245.0, "step": 7977 }, { "epoch": 1.0148836025950896, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.722602605819702, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8738772869110107, "num_tokens": 302894115.0, "step": 7978 }, { "epoch": 1.0150108128736801, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.466856002807617, "learning_rate": 1e-05, "loss": 0.3591, "mean_token_accuracy": 0.8919724822044373, "num_tokens": 302930162.0, "step": 7979 }, { "epoch": 1.0151380231522706, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7258291244506836, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8720031976699829, "num_tokens": 302965184.0, "step": 7980 }, { "epoch": 1.0152652334308612, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.490994691848755, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.863806962966919, "num_tokens": 303003895.0, "step": 7981 }, { "epoch": 1.0153924437094517, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.732375383377075, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8715015053749084, "num_tokens": 303038328.0, "step": 7982 }, { "epoch": 1.0155196539880422, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5128438472747803, "learning_rate": 1e-05, "loss": 0.5079, "mean_token_accuracy": 0.8484843373298645, "num_tokens": 303075546.0, "step": 7983 }, { "epoch": 1.0156468642666328, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7108209133148193, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8687156438827515, "num_tokens": 303111825.0, "step": 7984 }, { "epoch": 1.0157740745452233, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4986042976379395, "learning_rate": 1e-05, "loss": 0.3961, "mean_token_accuracy": 0.8793979287147522, "num_tokens": 303147607.0, "step": 7985 }, { "epoch": 1.0159012848238138, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.706759452819824, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8667733073234558, "num_tokens": 303185646.0, "step": 7986 }, { "epoch": 1.0160284951024043, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5078365802764893, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8752325773239136, "num_tokens": 303218723.0, "step": 7987 }, { "epoch": 1.0161557053809949, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6866326332092285, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8669398427009583, "num_tokens": 303259012.0, "step": 7988 }, { "epoch": 1.0162829156595854, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.50537371635437, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8699507713317871, "num_tokens": 303294868.0, "step": 7989 }, { "epoch": 1.0164101259381757, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.688056707382202, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8724694848060608, "num_tokens": 303331987.0, "step": 7990 }, { "epoch": 1.0165373362167662, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.501725912094116, "learning_rate": 1e-05, "loss": 0.3595, "mean_token_accuracy": 0.8920872211456299, "num_tokens": 303364230.0, "step": 7991 }, { "epoch": 1.0166645464953568, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.673151731491089, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.8806222081184387, "num_tokens": 303404169.0, "step": 7992 }, { "epoch": 1.0167917567739473, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4929871559143066, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.875411331653595, "num_tokens": 303445137.0, "step": 7993 }, { "epoch": 1.0169189670525378, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.66218900680542, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.8852888345718384, "num_tokens": 303486515.0, "step": 7994 }, { "epoch": 1.0170461773311283, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.511449098587036, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8701742887496948, "num_tokens": 303521514.0, "step": 7995 }, { "epoch": 1.0171733876097189, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6754720211029053, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8697722554206848, "num_tokens": 303555606.0, "step": 7996 }, { "epoch": 1.0173005978883094, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.49763560295105, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8718534111976624, "num_tokens": 303592075.0, "step": 7997 }, { "epoch": 1.0174278081669, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.663681745529175, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8527204990386963, "num_tokens": 303632529.0, "step": 7998 }, { "epoch": 1.0175550184454905, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.0001888275146484375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.508702039718628, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8769419193267822, "num_tokens": 303665419.0, "step": 7999 }, { "epoch": 1.017682228724081, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6649727821350098, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8690402507781982, "num_tokens": 303698778.0, "step": 8000 }, { "epoch": 1.0178094390026715, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.492649793624878, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8700663447380066, "num_tokens": 303738713.0, "step": 8001 }, { "epoch": 1.0179366492812618, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.671835422515869, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8712080121040344, "num_tokens": 303767821.0, "step": 8002 }, { "epoch": 1.0180638595598523, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.472029209136963, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8725593090057373, "num_tokens": 303813088.0, "step": 8003 }, { "epoch": 1.0181910698384429, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7124099731445312, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.85552978515625, "num_tokens": 303849926.0, "step": 8004 }, { "epoch": 1.0183182801170334, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.472280502319336, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8764123320579529, "num_tokens": 303888681.0, "step": 8005 }, { "epoch": 1.018445490395624, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7097184658050537, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.873604953289032, "num_tokens": 303926042.0, "step": 8006 }, { "epoch": 1.0185727006742145, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4700968265533447, "learning_rate": 1e-05, "loss": 0.3954, "mean_token_accuracy": 0.8827790021896362, "num_tokens": 303962014.0, "step": 8007 }, { "epoch": 1.018699910952805, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7192766666412354, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8730518817901611, "num_tokens": 303998856.0, "step": 8008 }, { "epoch": 1.0188271212313955, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.479173183441162, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8763607740402222, "num_tokens": 304034228.0, "step": 8009 }, { "epoch": 1.018954331509986, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.743502616882324, "learning_rate": 1e-05, "loss": 0.5178, "mean_token_accuracy": 0.8516495227813721, "num_tokens": 304070477.0, "step": 8010 }, { "epoch": 1.0190815417885766, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4653024673461914, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8767795562744141, "num_tokens": 304116073.0, "step": 8011 }, { "epoch": 1.019208752067167, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.759159803390503, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8636700510978699, "num_tokens": 304149950.0, "step": 8012 }, { "epoch": 1.0193359623457576, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4672770500183105, "learning_rate": 1e-05, "loss": 0.3756, "mean_token_accuracy": 0.8881504535675049, "num_tokens": 304191824.0, "step": 8013 }, { "epoch": 1.0194631726243482, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.773674488067627, "learning_rate": 1e-05, "loss": 0.3875, "mean_token_accuracy": 0.8851603865623474, "num_tokens": 304223616.0, "step": 8014 }, { "epoch": 1.0195903829029385, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4958794116973877, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8695100545883179, "num_tokens": 304261120.0, "step": 8015 }, { "epoch": 1.019717593181529, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.78464674949646, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8589277267456055, "num_tokens": 304294796.0, "step": 8016 }, { "epoch": 1.0198448034601195, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5014050006866455, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8612390756607056, "num_tokens": 304333396.0, "step": 8017 }, { "epoch": 1.01997201373871, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.769484043121338, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8741005063056946, "num_tokens": 304371246.0, "step": 8018 }, { "epoch": 1.0200992240173006, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.50411057472229, "learning_rate": 1e-05, "loss": 0.3899, "mean_token_accuracy": 0.8848628997802734, "num_tokens": 304408179.0, "step": 8019 }, { "epoch": 1.020226434295891, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7641990184783936, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8747707605361938, "num_tokens": 304442758.0, "step": 8020 }, { "epoch": 1.0203536445744816, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5164599418640137, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.88160640001297, "num_tokens": 304479568.0, "step": 8021 }, { "epoch": 1.0204808548530722, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7432873249053955, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8720279932022095, "num_tokens": 304521255.0, "step": 8022 }, { "epoch": 1.0206080651316627, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5243680477142334, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8731452226638794, "num_tokens": 304553569.0, "step": 8023 }, { "epoch": 1.0207352754102532, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7159695625305176, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8680123090744019, "num_tokens": 304594632.0, "step": 8024 }, { "epoch": 1.0208624856888437, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5279128551483154, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8692263960838318, "num_tokens": 304635255.0, "step": 8025 }, { "epoch": 1.0209896959674343, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.692728281021118, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8724955320358276, "num_tokens": 304674472.0, "step": 8026 }, { "epoch": 1.0211169062460246, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.527927875518799, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8790497779846191, "num_tokens": 304708950.0, "step": 8027 }, { "epoch": 1.021244116524615, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.667949676513672, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8702656030654907, "num_tokens": 304752849.0, "step": 8028 }, { "epoch": 1.0213713268032056, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5232479572296143, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8690251111984253, "num_tokens": 304791936.0, "step": 8029 }, { "epoch": 1.0214985370817962, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6480815410614014, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8741178512573242, "num_tokens": 304832228.0, "step": 8030 }, { "epoch": 1.0216257473603867, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.52514910697937, "learning_rate": 1e-05, "loss": 0.3685, "mean_token_accuracy": 0.8883716464042664, "num_tokens": 304863169.0, "step": 8031 }, { "epoch": 1.0217529576389772, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6456689834594727, "learning_rate": 1e-05, "loss": 0.5207, "mean_token_accuracy": 0.8405452370643616, "num_tokens": 304904938.0, "step": 8032 }, { "epoch": 1.0218801679175677, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5197789669036865, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8722422122955322, "num_tokens": 304940590.0, "step": 8033 }, { "epoch": 1.0220073781961583, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6315438747406006, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8664984107017517, "num_tokens": 304976945.0, "step": 8034 }, { "epoch": 1.0221345884747488, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4937126636505127, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8722593784332275, "num_tokens": 305021407.0, "step": 8035 }, { "epoch": 1.0222617987533393, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.64585018157959, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8779007196426392, "num_tokens": 305053141.0, "step": 8036 }, { "epoch": 1.0223890090319299, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5009496212005615, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8780714869499207, "num_tokens": 305086795.0, "step": 8037 }, { "epoch": 1.0225162193105204, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.657270669937134, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8595502376556396, "num_tokens": 305125759.0, "step": 8038 }, { "epoch": 1.0226434295891107, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.483214855194092, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8730023503303528, "num_tokens": 305165169.0, "step": 8039 }, { "epoch": 1.0227706398677012, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6580328941345215, "learning_rate": 1e-05, "loss": 0.383, "mean_token_accuracy": 0.8902196884155273, "num_tokens": 305203457.0, "step": 8040 }, { "epoch": 1.0228978501462918, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4822845458984375, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8622848391532898, "num_tokens": 305241021.0, "step": 8041 }, { "epoch": 1.0230250604248823, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6850945949554443, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8523368239402771, "num_tokens": 305277579.0, "step": 8042 }, { "epoch": 1.0231522707034728, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.480992078781128, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8629113435745239, "num_tokens": 305318027.0, "step": 8043 }, { "epoch": 1.0232794809820633, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.694164276123047, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8724339008331299, "num_tokens": 305351249.0, "step": 8044 }, { "epoch": 1.0234066912606539, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4700446128845215, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8736937642097473, "num_tokens": 305387906.0, "step": 8045 }, { "epoch": 1.0235339015392444, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7041258811950684, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8790582418441772, "num_tokens": 305419511.0, "step": 8046 }, { "epoch": 1.023661111817835, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.472989797592163, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8724290132522583, "num_tokens": 305457483.0, "step": 8047 }, { "epoch": 1.0237883220964255, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.740656852722168, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8749279975891113, "num_tokens": 305486498.0, "step": 8048 }, { "epoch": 1.023915532375016, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.474910259246826, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8718395829200745, "num_tokens": 305525250.0, "step": 8049 }, { "epoch": 1.0240427426536065, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7427351474761963, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8683176040649414, "num_tokens": 305566229.0, "step": 8050 }, { "epoch": 1.0241699529321968, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.478821039199829, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8789860010147095, "num_tokens": 305601884.0, "step": 8051 }, { "epoch": 1.0242971632107873, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.745713949203491, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8686676621437073, "num_tokens": 305644495.0, "step": 8052 }, { "epoch": 1.0244243734893779, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4864304065704346, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8773470520973206, "num_tokens": 305680662.0, "step": 8053 }, { "epoch": 1.0245515837679684, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7355594635009766, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.876468300819397, "num_tokens": 305720821.0, "step": 8054 }, { "epoch": 1.024678794046559, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.505321502685547, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8630828261375427, "num_tokens": 305757312.0, "step": 8055 }, { "epoch": 1.0248060043251495, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7335469722747803, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8616129159927368, "num_tokens": 305794372.0, "step": 8056 }, { "epoch": 1.02493321460374, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.491342782974243, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8719513416290283, "num_tokens": 305839975.0, "step": 8057 }, { "epoch": 1.0250604248823305, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7238433361053467, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8792574405670166, "num_tokens": 305877964.0, "step": 8058 }, { "epoch": 1.025187635160921, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5125999450683594, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8705877065658569, "num_tokens": 305911077.0, "step": 8059 }, { "epoch": 1.0253148454395116, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.712178945541382, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8767653703689575, "num_tokens": 305947741.0, "step": 8060 }, { "epoch": 1.025442055718102, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5084171295166016, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8681048154830933, "num_tokens": 305986872.0, "step": 8061 }, { "epoch": 1.0255692659966926, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.70693039894104, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8631590604782104, "num_tokens": 306023437.0, "step": 8062 }, { "epoch": 1.0256964762752832, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.503784656524658, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8817419409751892, "num_tokens": 306063026.0, "step": 8063 }, { "epoch": 1.0258236865538735, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6883060932159424, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8648545742034912, "num_tokens": 306102574.0, "step": 8064 }, { "epoch": 1.025950896832464, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.512481451034546, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.865630030632019, "num_tokens": 306137194.0, "step": 8065 }, { "epoch": 1.0260781071110545, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6869804859161377, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8690410852432251, "num_tokens": 306171539.0, "step": 8066 }, { "epoch": 1.026205317389645, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5041801929473877, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8675146698951721, "num_tokens": 306212385.0, "step": 8067 }, { "epoch": 1.0263325276682356, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.677924394607544, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8731430768966675, "num_tokens": 306247624.0, "step": 8068 }, { "epoch": 1.026459737946826, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.511249303817749, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8576400279998779, "num_tokens": 306286445.0, "step": 8069 }, { "epoch": 1.0265869482254166, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6566219329833984, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8785833120346069, "num_tokens": 306327792.0, "step": 8070 }, { "epoch": 1.0267141585040072, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.514115333557129, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8541127443313599, "num_tokens": 306366811.0, "step": 8071 }, { "epoch": 1.0268413687825977, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.661863327026367, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8668893575668335, "num_tokens": 306403938.0, "step": 8072 }, { "epoch": 1.0269685790611882, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4941585063934326, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.8830891251564026, "num_tokens": 306440623.0, "step": 8073 }, { "epoch": 1.0270957893397787, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6628477573394775, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.870057225227356, "num_tokens": 306477623.0, "step": 8074 }, { "epoch": 1.0272229996183693, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.506281852722168, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8697471022605896, "num_tokens": 306513586.0, "step": 8075 }, { "epoch": 1.0273502098969596, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6569607257843018, "learning_rate": 1e-05, "loss": 0.3875, "mean_token_accuracy": 0.8866673707962036, "num_tokens": 306548121.0, "step": 8076 }, { "epoch": 1.02747742017555, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.473963975906372, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8797875642776489, "num_tokens": 306596776.0, "step": 8077 }, { "epoch": 1.0276046304541406, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6764450073242188, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8651158809661865, "num_tokens": 306635497.0, "step": 8078 }, { "epoch": 1.0277318407327312, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.540724515914917, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8589707612991333, "num_tokens": 306671933.0, "step": 8079 }, { "epoch": 1.0278590510113217, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6446328163146973, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8730946779251099, "num_tokens": 306709313.0, "step": 8080 }, { "epoch": 1.0279862612899122, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.500762701034546, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8556026220321655, "num_tokens": 306748181.0, "step": 8081 }, { "epoch": 1.0281134715685027, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.638488531112671, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8825547695159912, "num_tokens": 306786353.0, "step": 8082 }, { "epoch": 1.0282406818470933, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.00019073486328125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5027012825012207, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8709430694580078, "num_tokens": 306821660.0, "step": 8083 }, { "epoch": 1.0283678921256838, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.632896900177002, "learning_rate": 1e-05, "loss": 0.3663, "mean_token_accuracy": 0.8926891088485718, "num_tokens": 306858479.0, "step": 8084 }, { "epoch": 1.0284951024042743, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.514207601547241, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8705840110778809, "num_tokens": 306890483.0, "step": 8085 }, { "epoch": 1.0286223126828649, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.624398946762085, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8735547065734863, "num_tokens": 306931865.0, "step": 8086 }, { "epoch": 1.0287495229614554, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5125410556793213, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8669912815093994, "num_tokens": 306966907.0, "step": 8087 }, { "epoch": 1.0288767332400457, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6141068935394287, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8777973651885986, "num_tokens": 307006032.0, "step": 8088 }, { "epoch": 1.0290039435186362, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5100862979888916, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8642884492874146, "num_tokens": 307045621.0, "step": 8089 }, { "epoch": 1.0291311537972267, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6324288845062256, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8611500263214111, "num_tokens": 307078768.0, "step": 8090 }, { "epoch": 1.0292583640758173, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4932656288146973, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8744617700576782, "num_tokens": 307120089.0, "step": 8091 }, { "epoch": 1.0293855743544078, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6277029514312744, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8697248697280884, "num_tokens": 307157478.0, "step": 8092 }, { "epoch": 1.0295127846329983, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4832875728607178, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8723134994506836, "num_tokens": 307199842.0, "step": 8093 }, { "epoch": 1.0296399949115889, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.637690544128418, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8705354928970337, "num_tokens": 307240609.0, "step": 8094 }, { "epoch": 1.0297672051901794, "ewc_loss": 0.05517578125, "ewc_loss_diag": 0.00019168853759765625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.489607572555542, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8627464771270752, "num_tokens": 307279112.0, "step": 8095 }, { "epoch": 1.02989441546877, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.653460741043091, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8781814575195312, "num_tokens": 307313718.0, "step": 8096 }, { "epoch": 1.0300216257473604, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4803199768066406, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8705956935882568, "num_tokens": 307353660.0, "step": 8097 }, { "epoch": 1.030148836025951, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6624417304992676, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.863081157207489, "num_tokens": 307393974.0, "step": 8098 }, { "epoch": 1.0302760463045415, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.468939781188965, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8820245265960693, "num_tokens": 307434877.0, "step": 8099 }, { "epoch": 1.0304032565831318, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6697781085968018, "learning_rate": 1e-05, "loss": 0.3994, "mean_token_accuracy": 0.8812210559844971, "num_tokens": 307473167.0, "step": 8100 }, { "epoch": 1.0305304668617223, "ewc_loss": 0.055908203125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4634532928466797, "learning_rate": 1e-05, "loss": 0.3866, "mean_token_accuracy": 0.8817479610443115, "num_tokens": 307513781.0, "step": 8101 }, { "epoch": 1.0306576771403129, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.699733257293701, "learning_rate": 1e-05, "loss": 0.3735, "mean_token_accuracy": 0.8886745572090149, "num_tokens": 307549524.0, "step": 8102 }, { "epoch": 1.0307848874189034, "ewc_loss": 0.055419921875, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4667651653289795, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8705324530601501, "num_tokens": 307585196.0, "step": 8103 }, { "epoch": 1.030912097697494, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.722057342529297, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8600566387176514, "num_tokens": 307623002.0, "step": 8104 }, { "epoch": 1.0310393079760845, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4491584300994873, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8736132979393005, "num_tokens": 307667736.0, "step": 8105 }, { "epoch": 1.031166518254675, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.742090940475464, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8809994459152222, "num_tokens": 307706031.0, "step": 8106 }, { "epoch": 1.0312937285332655, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.467832326889038, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8588916659355164, "num_tokens": 307746333.0, "step": 8107 }, { "epoch": 1.031420938811856, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.76986026763916, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8579964637756348, "num_tokens": 307784479.0, "step": 8108 }, { "epoch": 1.0315481490904466, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.483694314956665, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.870646059513092, "num_tokens": 307818245.0, "step": 8109 }, { "epoch": 1.031675359369037, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7523770332336426, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.876518726348877, "num_tokens": 307858730.0, "step": 8110 }, { "epoch": 1.0318025696476276, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4804651737213135, "learning_rate": 1e-05, "loss": 0.39, "mean_token_accuracy": 0.8849942684173584, "num_tokens": 307897067.0, "step": 8111 }, { "epoch": 1.0319297799262181, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.762639045715332, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8702076077461243, "num_tokens": 307935896.0, "step": 8112 }, { "epoch": 1.0320569902048085, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.514536142349243, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8505113124847412, "num_tokens": 307970930.0, "step": 8113 }, { "epoch": 1.032184200483399, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7409348487854004, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8614402413368225, "num_tokens": 308015857.0, "step": 8114 }, { "epoch": 1.0323114107619895, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.499795913696289, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8724206686019897, "num_tokens": 308055432.0, "step": 8115 }, { "epoch": 1.03243862104058, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.731992244720459, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.86034095287323, "num_tokens": 308094922.0, "step": 8116 }, { "epoch": 1.0325658313191706, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.507917642593384, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8733859062194824, "num_tokens": 308132941.0, "step": 8117 }, { "epoch": 1.032693041597761, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7167859077453613, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8545024394989014, "num_tokens": 308176749.0, "step": 8118 }, { "epoch": 1.0328202518763516, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.511556386947632, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8614107370376587, "num_tokens": 308215845.0, "step": 8119 }, { "epoch": 1.0329474621549422, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7040958404541016, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8783146142959595, "num_tokens": 308254220.0, "step": 8120 }, { "epoch": 1.0330746724335327, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001926422119140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.520282506942749, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8606311678886414, "num_tokens": 308290059.0, "step": 8121 }, { "epoch": 1.0332018827121232, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7125601768493652, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.846030592918396, "num_tokens": 308320478.0, "step": 8122 }, { "epoch": 1.0333290929907137, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.507436990737915, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8751564025878906, "num_tokens": 308356094.0, "step": 8123 }, { "epoch": 1.0334563032693043, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.691032648086548, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8739922046661377, "num_tokens": 308393120.0, "step": 8124 }, { "epoch": 1.0335835135478946, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.515453815460205, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.8830351233482361, "num_tokens": 308424153.0, "step": 8125 }, { "epoch": 1.033710723826485, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6767141819000244, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8747038841247559, "num_tokens": 308462884.0, "step": 8126 }, { "epoch": 1.0338379341050756, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4970481395721436, "learning_rate": 1e-05, "loss": 0.3669, "mean_token_accuracy": 0.8911656141281128, "num_tokens": 308501185.0, "step": 8127 }, { "epoch": 1.0339651443836662, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6823432445526123, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8609181642532349, "num_tokens": 308538770.0, "step": 8128 }, { "epoch": 1.0340923546622567, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4991114139556885, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8736138343811035, "num_tokens": 308575224.0, "step": 8129 }, { "epoch": 1.0342195649408472, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.679563283920288, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8829548358917236, "num_tokens": 308611859.0, "step": 8130 }, { "epoch": 1.0343467752194377, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5076656341552734, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8561820983886719, "num_tokens": 308648212.0, "step": 8131 }, { "epoch": 1.0344739854980283, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6691269874572754, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8687506914138794, "num_tokens": 308689830.0, "step": 8132 }, { "epoch": 1.0346011957766188, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5080440044403076, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8576157689094543, "num_tokens": 308726238.0, "step": 8133 }, { "epoch": 1.0347284060552093, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6936986446380615, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.862058699131012, "num_tokens": 308755090.0, "step": 8134 }, { "epoch": 1.0348556163337999, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.491297483444214, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8569930791854858, "num_tokens": 308794025.0, "step": 8135 }, { "epoch": 1.0349828266123904, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.681321620941162, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8713696002960205, "num_tokens": 308831521.0, "step": 8136 }, { "epoch": 1.0351100368909807, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.492729663848877, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8761334419250488, "num_tokens": 308868113.0, "step": 8137 }, { "epoch": 1.0352372471695712, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.686124563217163, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.870929479598999, "num_tokens": 308905070.0, "step": 8138 }, { "epoch": 1.0353644574481617, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5010972023010254, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8637728691101074, "num_tokens": 308942703.0, "step": 8139 }, { "epoch": 1.0354916677267523, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.674499034881592, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8724217414855957, "num_tokens": 308981328.0, "step": 8140 }, { "epoch": 1.0356188780053428, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4949278831481934, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.8845801949501038, "num_tokens": 309017751.0, "step": 8141 }, { "epoch": 1.0357460882839333, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.683424711227417, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8718960285186768, "num_tokens": 309052216.0, "step": 8142 }, { "epoch": 1.0358732985625239, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.520552635192871, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8701769709587097, "num_tokens": 309091323.0, "step": 8143 }, { "epoch": 1.0360005088411144, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6700451374053955, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8706797957420349, "num_tokens": 309125324.0, "step": 8144 }, { "epoch": 1.036127719119705, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5030579566955566, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8709728717803955, "num_tokens": 309166009.0, "step": 8145 }, { "epoch": 1.0362549293982954, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.648798942565918, "learning_rate": 1e-05, "loss": 0.3975, "mean_token_accuracy": 0.883434534072876, "num_tokens": 309205822.0, "step": 8146 }, { "epoch": 1.036382139676886, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5070414543151855, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8731410503387451, "num_tokens": 309243463.0, "step": 8147 }, { "epoch": 1.0365093499554765, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.665630340576172, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8652124404907227, "num_tokens": 309276904.0, "step": 8148 }, { "epoch": 1.0366365602340668, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4932925701141357, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8694226145744324, "num_tokens": 309314292.0, "step": 8149 }, { "epoch": 1.0367637705126573, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6751136779785156, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8622775077819824, "num_tokens": 309349231.0, "step": 8150 }, { "epoch": 1.0368909807912479, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5028371810913086, "learning_rate": 1e-05, "loss": 0.5094, "mean_token_accuracy": 0.8463150262832642, "num_tokens": 309387901.0, "step": 8151 }, { "epoch": 1.0370181910698384, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6651246547698975, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8740994334220886, "num_tokens": 309427167.0, "step": 8152 }, { "epoch": 1.037145401348429, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.500001907348633, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8717508316040039, "num_tokens": 309464121.0, "step": 8153 }, { "epoch": 1.0372726116270194, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6560490131378174, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8729782700538635, "num_tokens": 309506759.0, "step": 8154 }, { "epoch": 1.03739982190561, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4885315895080566, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8800281286239624, "num_tokens": 309546059.0, "step": 8155 }, { "epoch": 1.0375270321842005, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.664541244506836, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8651658892631531, "num_tokens": 309585076.0, "step": 8156 }, { "epoch": 1.037654242462791, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.499497652053833, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8794537782669067, "num_tokens": 309623354.0, "step": 8157 }, { "epoch": 1.0377814527413816, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.652670383453369, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8570718765258789, "num_tokens": 309668420.0, "step": 8158 }, { "epoch": 1.037908663019972, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.489391803741455, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8734641671180725, "num_tokens": 309711417.0, "step": 8159 }, { "epoch": 1.0380358732985626, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6627144813537598, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8740975260734558, "num_tokens": 309749938.0, "step": 8160 }, { "epoch": 1.0381630835771531, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4958813190460205, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8753061890602112, "num_tokens": 309783328.0, "step": 8161 }, { "epoch": 1.0382902938557435, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.658611297607422, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.8767643570899963, "num_tokens": 309821347.0, "step": 8162 }, { "epoch": 1.038417504134334, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5013723373413086, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8540151715278625, "num_tokens": 309858430.0, "step": 8163 }, { "epoch": 1.0385447144129245, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.657362461090088, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8702413439750671, "num_tokens": 309896096.0, "step": 8164 }, { "epoch": 1.038671924691515, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4945149421691895, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8698073625564575, "num_tokens": 309929069.0, "step": 8165 }, { "epoch": 1.0387991349701056, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6684744358062744, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8575907945632935, "num_tokens": 309969706.0, "step": 8166 }, { "epoch": 1.038926345248696, "ewc_loss": 0.0556640625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4979896545410156, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8668919801712036, "num_tokens": 310003990.0, "step": 8167 }, { "epoch": 1.0390535555272866, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.689150094985962, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8514994382858276, "num_tokens": 310036253.0, "step": 8168 }, { "epoch": 1.0391807658058771, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.476454973220825, "learning_rate": 1e-05, "loss": 0.3741, "mean_token_accuracy": 0.885459840297699, "num_tokens": 310073189.0, "step": 8169 }, { "epoch": 1.0393079760844677, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6845080852508545, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8642085790634155, "num_tokens": 310110432.0, "step": 8170 }, { "epoch": 1.0394351863630582, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4783782958984375, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8632773160934448, "num_tokens": 310154830.0, "step": 8171 }, { "epoch": 1.0395623966416487, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.713573455810547, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8558586239814758, "num_tokens": 310187270.0, "step": 8172 }, { "epoch": 1.0396896069202393, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4772493839263916, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.871535062789917, "num_tokens": 310223736.0, "step": 8173 }, { "epoch": 1.0398168171988296, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7353644371032715, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.8534069061279297, "num_tokens": 310259262.0, "step": 8174 }, { "epoch": 1.03994402747742, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4649171829223633, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8701870441436768, "num_tokens": 310300438.0, "step": 8175 }, { "epoch": 1.0400712377560106, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.736234188079834, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8793243765830994, "num_tokens": 310336630.0, "step": 8176 }, { "epoch": 1.0401984480346012, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4902851581573486, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8660221695899963, "num_tokens": 310368755.0, "step": 8177 }, { "epoch": 1.0403256583131917, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7422091960906982, "learning_rate": 1e-05, "loss": 0.3937, "mean_token_accuracy": 0.8812610507011414, "num_tokens": 310403520.0, "step": 8178 }, { "epoch": 1.0404528685917822, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5007081031799316, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8680661916732788, "num_tokens": 310433805.0, "step": 8179 }, { "epoch": 1.0405800788703727, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7713944911956787, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8651268482208252, "num_tokens": 310469801.0, "step": 8180 }, { "epoch": 1.0407072891489633, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.484978675842285, "learning_rate": 1e-05, "loss": 0.378, "mean_token_accuracy": 0.8856592774391174, "num_tokens": 310504972.0, "step": 8181 }, { "epoch": 1.0408344994275538, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.765254497528076, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8656151294708252, "num_tokens": 310537851.0, "step": 8182 }, { "epoch": 1.0409617097061443, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5015361309051514, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8703680634498596, "num_tokens": 310572647.0, "step": 8183 }, { "epoch": 1.0410889199847349, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.748751401901245, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8654961585998535, "num_tokens": 310612239.0, "step": 8184 }, { "epoch": 1.0412161302633254, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000194549560546875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5142335891723633, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8637163043022156, "num_tokens": 310650442.0, "step": 8185 }, { "epoch": 1.0413433405419157, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.732996702194214, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8675588369369507, "num_tokens": 310687687.0, "step": 8186 }, { "epoch": 1.0414705508205062, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.532024383544922, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.861763596534729, "num_tokens": 310725165.0, "step": 8187 }, { "epoch": 1.0415977610990967, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6968250274658203, "learning_rate": 1e-05, "loss": 0.3899, "mean_token_accuracy": 0.8847912549972534, "num_tokens": 310762532.0, "step": 8188 }, { "epoch": 1.0417249713776873, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.519073724746704, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8608490824699402, "num_tokens": 310800903.0, "step": 8189 }, { "epoch": 1.0418521816562778, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6884961128234863, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8504586219787598, "num_tokens": 310841391.0, "step": 8190 }, { "epoch": 1.0419793919348683, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5346105098724365, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8618545532226562, "num_tokens": 310879027.0, "step": 8191 }, { "epoch": 1.0421066022134589, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6667747497558594, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8742667436599731, "num_tokens": 310918448.0, "step": 8192 }, { "epoch": 1.0422338124920494, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5203099250793457, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8654959201812744, "num_tokens": 310955641.0, "step": 8193 }, { "epoch": 1.04236102277064, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6633143424987793, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8467953205108643, "num_tokens": 310996487.0, "step": 8194 }, { "epoch": 1.0424882330492304, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5052382946014404, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8621917366981506, "num_tokens": 311038208.0, "step": 8195 }, { "epoch": 1.042615443327821, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.660362482070923, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8712408542633057, "num_tokens": 311073720.0, "step": 8196 }, { "epoch": 1.0427426536064115, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.488895893096924, "learning_rate": 1e-05, "loss": 0.3881, "mean_token_accuracy": 0.8835858106613159, "num_tokens": 311115488.0, "step": 8197 }, { "epoch": 1.0428698638850018, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6644253730773926, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8704704642295837, "num_tokens": 311159217.0, "step": 8198 }, { "epoch": 1.0429970741635923, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5040955543518066, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8595303893089294, "num_tokens": 311196058.0, "step": 8199 }, { "epoch": 1.0431242844421829, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6737442016601562, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8758598566055298, "num_tokens": 311231024.0, "step": 8200 }, { "epoch": 1.0432514947207734, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.493673086166382, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8707758188247681, "num_tokens": 311266883.0, "step": 8201 }, { "epoch": 1.043378704999364, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6804418563842773, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8529701232910156, "num_tokens": 311302594.0, "step": 8202 }, { "epoch": 1.0435059152779544, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4886507987976074, "learning_rate": 1e-05, "loss": 0.3691, "mean_token_accuracy": 0.8885889053344727, "num_tokens": 311334384.0, "step": 8203 }, { "epoch": 1.043633125556545, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6681764125823975, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.871971607208252, "num_tokens": 311381350.0, "step": 8204 }, { "epoch": 1.0437603358351355, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4790823459625244, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8758283853530884, "num_tokens": 311418107.0, "step": 8205 }, { "epoch": 1.043887546113726, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6882576942443848, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8702315092086792, "num_tokens": 311457881.0, "step": 8206 }, { "epoch": 1.0440147563923166, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4916481971740723, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8639225959777832, "num_tokens": 311492821.0, "step": 8207 }, { "epoch": 1.044141966670907, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6969122886657715, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8692115545272827, "num_tokens": 311527981.0, "step": 8208 }, { "epoch": 1.0442691769494976, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4731497764587402, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8610385060310364, "num_tokens": 311568853.0, "step": 8209 }, { "epoch": 1.0443963872280881, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.69305157661438, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8810229301452637, "num_tokens": 311613321.0, "step": 8210 }, { "epoch": 1.0445235975066784, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4757180213928223, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.864712655544281, "num_tokens": 311656849.0, "step": 8211 }, { "epoch": 1.044650807785269, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7287909984588623, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8575073480606079, "num_tokens": 311692231.0, "step": 8212 }, { "epoch": 1.0447780180638595, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.478243112564087, "learning_rate": 1e-05, "loss": 0.3988, "mean_token_accuracy": 0.8809146881103516, "num_tokens": 311732793.0, "step": 8213 }, { "epoch": 1.04490522834245, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7257187366485596, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8665362596511841, "num_tokens": 311773320.0, "step": 8214 }, { "epoch": 1.0450324386210406, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4725120067596436, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.866592288017273, "num_tokens": 311822187.0, "step": 8215 }, { "epoch": 1.045159648899631, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7470548152923584, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.868511438369751, "num_tokens": 311860272.0, "step": 8216 }, { "epoch": 1.0452868591782216, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4989874362945557, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8701114654541016, "num_tokens": 311894479.0, "step": 8217 }, { "epoch": 1.0454140694568121, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7372779846191406, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8803348541259766, "num_tokens": 311929401.0, "step": 8218 }, { "epoch": 1.0455412797354027, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.493027687072754, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8626886606216431, "num_tokens": 311969401.0, "step": 8219 }, { "epoch": 1.0456684900139932, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.748490571975708, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8583053946495056, "num_tokens": 312004295.0, "step": 8220 }, { "epoch": 1.0457957002925837, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5019233226776123, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8671126365661621, "num_tokens": 312038487.0, "step": 8221 }, { "epoch": 1.0459229105711743, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7255516052246094, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.870269775390625, "num_tokens": 312076011.0, "step": 8222 }, { "epoch": 1.0460501208497646, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5148701667785645, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8619797229766846, "num_tokens": 312114489.0, "step": 8223 }, { "epoch": 1.046177331128355, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7087247371673584, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8645133972167969, "num_tokens": 312152388.0, "step": 8224 }, { "epoch": 1.0463045414069456, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.506286144256592, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8749811053276062, "num_tokens": 312190269.0, "step": 8225 }, { "epoch": 1.0464317516855361, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.699328899383545, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8780273199081421, "num_tokens": 312229909.0, "step": 8226 }, { "epoch": 1.0465589619641267, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00019550323486328125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5311903953552246, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8540897369384766, "num_tokens": 312267527.0, "step": 8227 }, { "epoch": 1.0466861722427172, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6707041263580322, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8715764284133911, "num_tokens": 312310632.0, "step": 8228 }, { "epoch": 1.0468133825213077, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5141913890838623, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8600798845291138, "num_tokens": 312351570.0, "step": 8229 }, { "epoch": 1.0469405927998983, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.675271511077881, "learning_rate": 1e-05, "loss": 0.374, "mean_token_accuracy": 0.8902570009231567, "num_tokens": 312382765.0, "step": 8230 }, { "epoch": 1.0470678030784888, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5086941719055176, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.878565788269043, "num_tokens": 312415517.0, "step": 8231 }, { "epoch": 1.0471950133570793, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.672518014907837, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8762851357460022, "num_tokens": 312450574.0, "step": 8232 }, { "epoch": 1.0473222236356698, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.531395435333252, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8595836162567139, "num_tokens": 312483054.0, "step": 8233 }, { "epoch": 1.0474494339142604, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.638784885406494, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8691174983978271, "num_tokens": 312527825.0, "step": 8234 }, { "epoch": 1.0475766441928507, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5117201805114746, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8628168106079102, "num_tokens": 312566971.0, "step": 8235 }, { "epoch": 1.0477038544714412, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6429390907287598, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8702510595321655, "num_tokens": 312603527.0, "step": 8236 }, { "epoch": 1.0478310647500317, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5246336460113525, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8698257207870483, "num_tokens": 312634281.0, "step": 8237 }, { "epoch": 1.0479582750286223, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6401023864746094, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8642144203186035, "num_tokens": 312668036.0, "step": 8238 }, { "epoch": 1.0480854853072128, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.515347957611084, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.860937237739563, "num_tokens": 312707422.0, "step": 8239 }, { "epoch": 1.0482126955858033, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.7834596633911133, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8655248880386353, "num_tokens": 312740715.0, "step": 8240 }, { "epoch": 1.0483399058643939, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4801557064056396, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8719520568847656, "num_tokens": 312773196.0, "step": 8241 }, { "epoch": 1.0484671161429844, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7113025188446045, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8619665503501892, "num_tokens": 312817502.0, "step": 8242 }, { "epoch": 1.048594326421575, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001964569091796875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4670848846435547, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8665109872817993, "num_tokens": 312852810.0, "step": 8243 }, { "epoch": 1.0487215367001654, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7303051948547363, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8783447742462158, "num_tokens": 312889069.0, "step": 8244 }, { "epoch": 1.048848746978756, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5147173404693604, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8667118549346924, "num_tokens": 312929032.0, "step": 8245 }, { "epoch": 1.0489759572573465, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7127175331115723, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8718839287757874, "num_tokens": 312963792.0, "step": 8246 }, { "epoch": 1.0491031675359368, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4875383377075195, "learning_rate": 1e-05, "loss": 0.3855, "mean_token_accuracy": 0.8864358067512512, "num_tokens": 312999532.0, "step": 8247 }, { "epoch": 1.0492303778145273, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7205886840820312, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8692517280578613, "num_tokens": 313040807.0, "step": 8248 }, { "epoch": 1.0493575880931179, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5018465518951416, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8786765336990356, "num_tokens": 313074895.0, "step": 8249 }, { "epoch": 1.0494847983717084, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7077889442443848, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8730800151824951, "num_tokens": 313115299.0, "step": 8250 }, { "epoch": 1.049612008650299, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.498854875564575, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8693233132362366, "num_tokens": 313157578.0, "step": 8251 }, { "epoch": 1.0497392189288894, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.687411069869995, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8685851097106934, "num_tokens": 313202219.0, "step": 8252 }, { "epoch": 1.04986642920748, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.4979166984558105, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.8741395473480225, "num_tokens": 313242345.0, "step": 8253 }, { "epoch": 1.0499936394860705, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.700202703475952, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8755356073379517, "num_tokens": 313277052.0, "step": 8254 }, { "epoch": 1.050120849764661, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5116655826568604, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8680935502052307, "num_tokens": 313311291.0, "step": 8255 }, { "epoch": 1.0502480600432516, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6869938373565674, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8739071488380432, "num_tokens": 313349141.0, "step": 8256 }, { "epoch": 1.050375270321842, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.510507583618164, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8570936322212219, "num_tokens": 313387575.0, "step": 8257 }, { "epoch": 1.0505024806004326, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6864287853240967, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8644404411315918, "num_tokens": 313424087.0, "step": 8258 }, { "epoch": 1.0506296908790231, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4939050674438477, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8748105764389038, "num_tokens": 313464517.0, "step": 8259 }, { "epoch": 1.0507569011576134, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.683990478515625, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8604260087013245, "num_tokens": 313504146.0, "step": 8260 }, { "epoch": 1.050884111436204, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.50085711479187, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8657422065734863, "num_tokens": 313541842.0, "step": 8261 }, { "epoch": 1.0510113217147945, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6849923133850098, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8647964596748352, "num_tokens": 313580695.0, "step": 8262 }, { "epoch": 1.051138531993385, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.513011932373047, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8587645292282104, "num_tokens": 313619494.0, "step": 8263 }, { "epoch": 1.0512657422719756, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.679334878921509, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8788852691650391, "num_tokens": 313653234.0, "step": 8264 }, { "epoch": 1.051392952550566, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5063397884368896, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8691911697387695, "num_tokens": 313694395.0, "step": 8265 }, { "epoch": 1.0515201628291566, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6812081336975098, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8563221096992493, "num_tokens": 313733370.0, "step": 8266 }, { "epoch": 1.0516473731077471, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.516345977783203, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8568294048309326, "num_tokens": 313770611.0, "step": 8267 }, { "epoch": 1.0517745833863377, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6589035987854004, "learning_rate": 1e-05, "loss": 0.3697, "mean_token_accuracy": 0.8901606798171997, "num_tokens": 313807262.0, "step": 8268 }, { "epoch": 1.0519017936649282, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.514153242111206, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8637326955795288, "num_tokens": 313844294.0, "step": 8269 }, { "epoch": 1.0520290039435187, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6527295112609863, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8815439343452454, "num_tokens": 313878701.0, "step": 8270 }, { "epoch": 1.0521562142221093, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.00019741058349609375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.495962619781494, "learning_rate": 1e-05, "loss": 0.398, "mean_token_accuracy": 0.8788304328918457, "num_tokens": 313917780.0, "step": 8271 }, { "epoch": 1.0522834245006996, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0001983642578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6654741764068604, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8822875618934631, "num_tokens": 313948611.0, "step": 8272 }, { "epoch": 1.05241063477929, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001983642578125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5010344982147217, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8753195405006409, "num_tokens": 313981999.0, "step": 8273 }, { "epoch": 1.0525378450578806, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0001983642578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6618385314941406, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8772616386413574, "num_tokens": 314021199.0, "step": 8274 }, { "epoch": 1.0526650553364711, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001983642578125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5418317317962646, "learning_rate": 1e-05, "loss": 0.3924, "mean_token_accuracy": 0.8819928169250488, "num_tokens": 314058187.0, "step": 8275 }, { "epoch": 1.0527922656150617, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0001983642578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6290643215179443, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8705373406410217, "num_tokens": 314095612.0, "step": 8276 }, { "epoch": 1.0529194758936522, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0001983642578125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5340797901153564, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8636811375617981, "num_tokens": 314124895.0, "step": 8277 }, { "epoch": 1.0530466861722427, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0001983642578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.605325698852539, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.879939079284668, "num_tokens": 314167787.0, "step": 8278 }, { "epoch": 1.0531738964508333, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0001983642578125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5230026245117188, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8719490766525269, "num_tokens": 314203389.0, "step": 8279 }, { "epoch": 1.0533011067294238, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0001983642578125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6007702350616455, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8804482221603394, "num_tokens": 314245161.0, "step": 8280 }, { "epoch": 1.0534283170080143, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0001983642578125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5185861587524414, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8474020957946777, "num_tokens": 314288097.0, "step": 8281 }, { "epoch": 1.0535555272866048, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6012468338012695, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8678054809570312, "num_tokens": 314324066.0, "step": 8282 }, { "epoch": 1.0536827375651954, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.518995761871338, "learning_rate": 1e-05, "loss": 0.3911, "mean_token_accuracy": 0.8797639608383179, "num_tokens": 314358430.0, "step": 8283 }, { "epoch": 1.0538099478437857, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.607619047164917, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8670991659164429, "num_tokens": 314394199.0, "step": 8284 }, { "epoch": 1.0539371581223762, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5099451541900635, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8720515966415405, "num_tokens": 314432851.0, "step": 8285 }, { "epoch": 1.0540643684009667, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6109719276428223, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8696428537368774, "num_tokens": 314473843.0, "step": 8286 }, { "epoch": 1.0541915786795573, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4884369373321533, "learning_rate": 1e-05, "loss": 0.3916, "mean_token_accuracy": 0.8851428031921387, "num_tokens": 314518782.0, "step": 8287 }, { "epoch": 1.0543187889581478, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.622474431991577, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8678308129310608, "num_tokens": 314561140.0, "step": 8288 }, { "epoch": 1.0544459992367383, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.498795509338379, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8689188957214355, "num_tokens": 314602605.0, "step": 8289 }, { "epoch": 1.0545732095153288, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6290206909179688, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8549565076828003, "num_tokens": 314647719.0, "step": 8290 }, { "epoch": 1.0547004197939194, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.495119094848633, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8568531274795532, "num_tokens": 314690016.0, "step": 8291 }, { "epoch": 1.05482763007251, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6319329738616943, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8727462291717529, "num_tokens": 314730432.0, "step": 8292 }, { "epoch": 1.0549548403511004, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4885575771331787, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.8781468868255615, "num_tokens": 314764957.0, "step": 8293 }, { "epoch": 1.055082050629691, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6437621116638184, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8751096725463867, "num_tokens": 314803787.0, "step": 8294 }, { "epoch": 1.0552092609082815, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.4777462482452393, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8764848709106445, "num_tokens": 314841732.0, "step": 8295 }, { "epoch": 1.0553364711868718, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6608312129974365, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8744584321975708, "num_tokens": 314878212.0, "step": 8296 }, { "epoch": 1.0554636814654623, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4809815883636475, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.866126298904419, "num_tokens": 314917130.0, "step": 8297 }, { "epoch": 1.0555908917440529, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6569197177886963, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8784222602844238, "num_tokens": 314961804.0, "step": 8298 }, { "epoch": 1.0557181020226434, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.47236967086792, "learning_rate": 1e-05, "loss": 0.3754, "mean_token_accuracy": 0.8871966600418091, "num_tokens": 314997157.0, "step": 8299 }, { "epoch": 1.055845312301234, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6853036880493164, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.852595865726471, "num_tokens": 315038370.0, "step": 8300 }, { "epoch": 1.0559725225798244, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4627182483673096, "learning_rate": 1e-05, "loss": 0.3919, "mean_token_accuracy": 0.8827837109565735, "num_tokens": 315081716.0, "step": 8301 }, { "epoch": 1.056099732858415, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.716209888458252, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8510380983352661, "num_tokens": 315124528.0, "step": 8302 }, { "epoch": 1.0562269431370055, "ewc_loss": 0.056396484375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.451582670211792, "learning_rate": 1e-05, "loss": 0.3839, "mean_token_accuracy": 0.8859643340110779, "num_tokens": 315163893.0, "step": 8303 }, { "epoch": 1.056354153415596, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7547340393066406, "learning_rate": 1e-05, "loss": 0.5097, "mean_token_accuracy": 0.8489124774932861, "num_tokens": 315199462.0, "step": 8304 }, { "epoch": 1.0564813636941865, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4430484771728516, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.86976158618927, "num_tokens": 315245075.0, "step": 8305 }, { "epoch": 1.056608573972777, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.765608787536621, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8742004632949829, "num_tokens": 315289291.0, "step": 8306 }, { "epoch": 1.0567357842513676, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.465775728225708, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8778600692749023, "num_tokens": 315327750.0, "step": 8307 }, { "epoch": 1.0568629945299581, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.787301778793335, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8661200404167175, "num_tokens": 315366863.0, "step": 8308 }, { "epoch": 1.0569902048085484, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4730427265167236, "learning_rate": 1e-05, "loss": 0.3898, "mean_token_accuracy": 0.8835214376449585, "num_tokens": 315406935.0, "step": 8309 }, { "epoch": 1.057117415087139, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8006134033203125, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8643465638160706, "num_tokens": 315446483.0, "step": 8310 }, { "epoch": 1.0572446253657295, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5009443759918213, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8662404417991638, "num_tokens": 315482862.0, "step": 8311 }, { "epoch": 1.05737183564432, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7861807346343994, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8734934329986572, "num_tokens": 315522649.0, "step": 8312 }, { "epoch": 1.0574990459229106, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5080864429473877, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8646253347396851, "num_tokens": 315562642.0, "step": 8313 }, { "epoch": 1.057626256201501, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.777848720550537, "learning_rate": 1e-05, "loss": 0.367, "mean_token_accuracy": 0.8945366144180298, "num_tokens": 315594056.0, "step": 8314 }, { "epoch": 1.0577534664800916, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.526791572570801, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8662306070327759, "num_tokens": 315632474.0, "step": 8315 }, { "epoch": 1.0578806767586821, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7786240577697754, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.871878981590271, "num_tokens": 315662036.0, "step": 8316 }, { "epoch": 1.0580078870372727, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5257458686828613, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8740282654762268, "num_tokens": 315701974.0, "step": 8317 }, { "epoch": 1.0581350973158632, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7388696670532227, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8684659004211426, "num_tokens": 315745876.0, "step": 8318 }, { "epoch": 1.0582623075944537, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.531254768371582, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8697609305381775, "num_tokens": 315787386.0, "step": 8319 }, { "epoch": 1.058389517873044, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.724797010421753, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8789132237434387, "num_tokens": 315824047.0, "step": 8320 }, { "epoch": 1.0585167281516346, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5417120456695557, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8535408973693848, "num_tokens": 315863890.0, "step": 8321 }, { "epoch": 1.058643938430225, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.699815273284912, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8654638528823853, "num_tokens": 315902855.0, "step": 8322 }, { "epoch": 1.0587711487088156, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5254721641540527, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.872316300868988, "num_tokens": 315942112.0, "step": 8323 }, { "epoch": 1.0588983589874061, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6873908042907715, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8734200596809387, "num_tokens": 315981256.0, "step": 8324 }, { "epoch": 1.0590255692659967, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.523510456085205, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8651158809661865, "num_tokens": 316021192.0, "step": 8325 }, { "epoch": 1.0591527795445872, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6812798976898193, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8676754236221313, "num_tokens": 316056261.0, "step": 8326 }, { "epoch": 1.0592799898231777, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5018105506896973, "learning_rate": 1e-05, "loss": 0.3774, "mean_token_accuracy": 0.8883129358291626, "num_tokens": 316093390.0, "step": 8327 }, { "epoch": 1.0594072001017683, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.67629075050354, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.858101487159729, "num_tokens": 316133771.0, "step": 8328 }, { "epoch": 1.0595344103803588, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.507420063018799, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.868370532989502, "num_tokens": 316172123.0, "step": 8329 }, { "epoch": 1.0596616206589493, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6616032123565674, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.8834604024887085, "num_tokens": 316211866.0, "step": 8330 }, { "epoch": 1.0597888309375398, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5097525119781494, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8750752210617065, "num_tokens": 316248838.0, "step": 8331 }, { "epoch": 1.0599160412161304, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.665696859359741, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8734304904937744, "num_tokens": 316286551.0, "step": 8332 }, { "epoch": 1.0600432514947207, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.523341417312622, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8698468208312988, "num_tokens": 316321278.0, "step": 8333 }, { "epoch": 1.0601704617733112, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.651770830154419, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8701034784317017, "num_tokens": 316356282.0, "step": 8334 }, { "epoch": 1.0602976720519017, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5077805519104004, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8679088950157166, "num_tokens": 316398858.0, "step": 8335 }, { "epoch": 1.0604248823304923, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6591744422912598, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8695229291915894, "num_tokens": 316432883.0, "step": 8336 }, { "epoch": 1.0605520926090828, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.0002002716064453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.498413324356079, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8844519853591919, "num_tokens": 316469305.0, "step": 8337 }, { "epoch": 1.0606793028876733, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6551876068115234, "learning_rate": 1e-05, "loss": 0.3787, "mean_token_accuracy": 0.8888394832611084, "num_tokens": 316509072.0, "step": 8338 }, { "epoch": 1.0608065131662638, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.494859218597412, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8665605187416077, "num_tokens": 316549391.0, "step": 8339 }, { "epoch": 1.0609337234448544, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.660261392593384, "learning_rate": 1e-05, "loss": 0.3798, "mean_token_accuracy": 0.8885307908058167, "num_tokens": 316583696.0, "step": 8340 }, { "epoch": 1.061060933723445, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.49833607673645, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8641184568405151, "num_tokens": 316618421.0, "step": 8341 }, { "epoch": 1.0611881440020354, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.659045696258545, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8701040148735046, "num_tokens": 316660284.0, "step": 8342 }, { "epoch": 1.061315354280626, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.503506660461426, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8553001880645752, "num_tokens": 316699753.0, "step": 8343 }, { "epoch": 1.0614425645592165, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.666192054748535, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8693675398826599, "num_tokens": 316735336.0, "step": 8344 }, { "epoch": 1.0615697748378068, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.489636182785034, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8759870529174805, "num_tokens": 316775327.0, "step": 8345 }, { "epoch": 1.0616969851163973, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6758365631103516, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8542313575744629, "num_tokens": 316817027.0, "step": 8346 }, { "epoch": 1.0618241953949878, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.504055976867676, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8726118206977844, "num_tokens": 316850980.0, "step": 8347 }, { "epoch": 1.0619514056735784, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.661494016647339, "learning_rate": 1e-05, "loss": 0.3856, "mean_token_accuracy": 0.887052595615387, "num_tokens": 316886533.0, "step": 8348 }, { "epoch": 1.062078615952169, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4871397018432617, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8804762363433838, "num_tokens": 316927447.0, "step": 8349 }, { "epoch": 1.0622058262307594, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6771135330200195, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8619723320007324, "num_tokens": 316965983.0, "step": 8350 }, { "epoch": 1.06233303650935, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4967501163482666, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8725380301475525, "num_tokens": 316999678.0, "step": 8351 }, { "epoch": 1.0624602467879405, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.678170919418335, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8611757159233093, "num_tokens": 317038650.0, "step": 8352 }, { "epoch": 1.062587457066531, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4880127906799316, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8747410178184509, "num_tokens": 317075507.0, "step": 8353 }, { "epoch": 1.0627146673451215, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.686279535293579, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8844277262687683, "num_tokens": 317109364.0, "step": 8354 }, { "epoch": 1.062841877623712, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4730257987976074, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8759409785270691, "num_tokens": 317149959.0, "step": 8355 }, { "epoch": 1.0629690879023026, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7091000080108643, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8693325519561768, "num_tokens": 317182604.0, "step": 8356 }, { "epoch": 1.0630962981808931, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.470381259918213, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8804794549942017, "num_tokens": 317224190.0, "step": 8357 }, { "epoch": 1.0632235084594834, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.727376937866211, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8678615093231201, "num_tokens": 317261607.0, "step": 8358 }, { "epoch": 1.063350718738074, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.482666015625, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8577656149864197, "num_tokens": 317300772.0, "step": 8359 }, { "epoch": 1.0634779290166645, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7307627201080322, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8704266548156738, "num_tokens": 317336981.0, "step": 8360 }, { "epoch": 1.063605139295255, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4844377040863037, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.862126350402832, "num_tokens": 317376522.0, "step": 8361 }, { "epoch": 1.0637323495738455, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7421844005584717, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8734527826309204, "num_tokens": 317409924.0, "step": 8362 }, { "epoch": 1.063859559852436, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5041024684906006, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.8840861320495605, "num_tokens": 317440762.0, "step": 8363 }, { "epoch": 1.0639867701310266, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7495181560516357, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.873346209526062, "num_tokens": 317471606.0, "step": 8364 }, { "epoch": 1.0641139804096171, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4928128719329834, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8591980934143066, "num_tokens": 317508383.0, "step": 8365 }, { "epoch": 1.0642411906882077, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7295613288879395, "learning_rate": 1e-05, "loss": 0.3859, "mean_token_accuracy": 0.8903671503067017, "num_tokens": 317546690.0, "step": 8366 }, { "epoch": 1.0643684009667982, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.49373722076416, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.8831851482391357, "num_tokens": 317584926.0, "step": 8367 }, { "epoch": 1.0644956112453887, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7443811893463135, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8596796989440918, "num_tokens": 317619941.0, "step": 8368 }, { "epoch": 1.064622821523979, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.514927625656128, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.882364809513092, "num_tokens": 317651006.0, "step": 8369 }, { "epoch": 1.0647500318025696, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.728367805480957, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8733072280883789, "num_tokens": 317688051.0, "step": 8370 }, { "epoch": 1.06487724208116, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4994711875915527, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8775333762168884, "num_tokens": 317727447.0, "step": 8371 }, { "epoch": 1.0650044523597506, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7083418369293213, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8812071084976196, "num_tokens": 317771495.0, "step": 8372 }, { "epoch": 1.0651316626383411, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5343267917633057, "learning_rate": 1e-05, "loss": 0.3961, "mean_token_accuracy": 0.8820691108703613, "num_tokens": 317808166.0, "step": 8373 }, { "epoch": 1.0652588729169317, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6913206577301025, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.876079261302948, "num_tokens": 317850928.0, "step": 8374 }, { "epoch": 1.0653860831955222, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4992055892944336, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8783735036849976, "num_tokens": 317893156.0, "step": 8375 }, { "epoch": 1.0655132934741127, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.689988851547241, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8469897508621216, "num_tokens": 317937146.0, "step": 8376 }, { "epoch": 1.0656405037527032, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5007171630859375, "learning_rate": 1e-05, "loss": 0.3987, "mean_token_accuracy": 0.8807743787765503, "num_tokens": 317977429.0, "step": 8377 }, { "epoch": 1.0657677140312938, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.693276882171631, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8677563667297363, "num_tokens": 318013445.0, "step": 8378 }, { "epoch": 1.0658949243098843, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5069804191589355, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8629611134529114, "num_tokens": 318054082.0, "step": 8379 }, { "epoch": 1.0660221345884748, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6847827434539795, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8572957515716553, "num_tokens": 318096125.0, "step": 8380 }, { "epoch": 1.0661493448670654, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.504256010055542, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8733406066894531, "num_tokens": 318130859.0, "step": 8381 }, { "epoch": 1.0662765551456557, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.680990695953369, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8770697712898254, "num_tokens": 318171616.0, "step": 8382 }, { "epoch": 1.0664037654242462, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5121548175811768, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8663216829299927, "num_tokens": 318206741.0, "step": 8383 }, { "epoch": 1.0665309757028367, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6582446098327637, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8669025897979736, "num_tokens": 318252513.0, "step": 8384 }, { "epoch": 1.0666581859814273, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5220956802368164, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8618635535240173, "num_tokens": 318286632.0, "step": 8385 }, { "epoch": 1.0667853962600178, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6495344638824463, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8726754784584045, "num_tokens": 318326923.0, "step": 8386 }, { "epoch": 1.0669126065386083, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.579774856567383, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8518791198730469, "num_tokens": 318364159.0, "step": 8387 }, { "epoch": 1.0670398168171988, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6130521297454834, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8485626578330994, "num_tokens": 318402715.0, "step": 8388 }, { "epoch": 1.0671670270957894, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5281779766082764, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8673957586288452, "num_tokens": 318445925.0, "step": 8389 }, { "epoch": 1.06729423737438, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6050310134887695, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8630828261375427, "num_tokens": 318481756.0, "step": 8390 }, { "epoch": 1.0674214476529704, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5157618522644043, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8700338006019592, "num_tokens": 318526885.0, "step": 8391 }, { "epoch": 1.067548657931561, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.602889060974121, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8660479187965393, "num_tokens": 318569396.0, "step": 8392 }, { "epoch": 1.0676758682101515, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.519890069961548, "learning_rate": 1e-05, "loss": 0.3964, "mean_token_accuracy": 0.8860996961593628, "num_tokens": 318606404.0, "step": 8393 }, { "epoch": 1.0678030784887418, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6193478107452393, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.857391893863678, "num_tokens": 318642457.0, "step": 8394 }, { "epoch": 1.0679302887673323, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5216941833496094, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8604246377944946, "num_tokens": 318680498.0, "step": 8395 }, { "epoch": 1.0680574990459228, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.6130876541137695, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.864462673664093, "num_tokens": 318720493.0, "step": 8396 }, { "epoch": 1.0681847093245134, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.52441668510437, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8628735542297363, "num_tokens": 318755297.0, "step": 8397 }, { "epoch": 1.068311919603104, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6143128871917725, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8760418891906738, "num_tokens": 318788716.0, "step": 8398 }, { "epoch": 1.0684391298816944, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.52074933052063, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8696208596229553, "num_tokens": 318825130.0, "step": 8399 }, { "epoch": 1.068566340160285, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.615516424179077, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8722427487373352, "num_tokens": 318861216.0, "step": 8400 }, { "epoch": 1.0686935504388755, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5094025135040283, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8741833567619324, "num_tokens": 318896410.0, "step": 8401 }, { "epoch": 1.068820760717466, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6163923740386963, "learning_rate": 1e-05, "loss": 0.3914, "mean_token_accuracy": 0.8841118812561035, "num_tokens": 318933834.0, "step": 8402 }, { "epoch": 1.0689479709960565, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5136756896972656, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8530699014663696, "num_tokens": 318973752.0, "step": 8403 }, { "epoch": 1.069075181274647, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6168274879455566, "learning_rate": 1e-05, "loss": 0.3809, "mean_token_accuracy": 0.8870072960853577, "num_tokens": 319009376.0, "step": 8404 }, { "epoch": 1.0692023915532376, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5010645389556885, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8729275465011597, "num_tokens": 319050175.0, "step": 8405 }, { "epoch": 1.0693296018318281, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6439738273620605, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8735105991363525, "num_tokens": 319080808.0, "step": 8406 }, { "epoch": 1.0694568121104184, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.485161542892456, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8770807385444641, "num_tokens": 319123123.0, "step": 8407 }, { "epoch": 1.069584022389009, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.672431707382202, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8784340620040894, "num_tokens": 319155116.0, "step": 8408 }, { "epoch": 1.0697112326675995, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4864048957824707, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8787429928779602, "num_tokens": 319188087.0, "step": 8409 }, { "epoch": 1.06983844294619, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6872596740722656, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.859963059425354, "num_tokens": 319228820.0, "step": 8410 }, { "epoch": 1.0699656532247805, "ewc_loss": 0.05615234375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.62396240234375e-05, "grad_norm": 2.491441011428833, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8476927280426025, "num_tokens": 319266046.0, "step": 8411 }, { "epoch": 1.070092863503371, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6819653511047363, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8743433952331543, "num_tokens": 319305209.0, "step": 8412 }, { "epoch": 1.0702200737819616, "ewc_loss": 0.056640625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.474137544631958, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8667351603507996, "num_tokens": 319347463.0, "step": 8413 }, { "epoch": 1.0703472840605521, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7043473720550537, "learning_rate": 1e-05, "loss": 0.3964, "mean_token_accuracy": 0.881850004196167, "num_tokens": 319383965.0, "step": 8414 }, { "epoch": 1.0704744943391427, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4769976139068604, "learning_rate": 1e-05, "loss": 0.398, "mean_token_accuracy": 0.8799406886100769, "num_tokens": 319418330.0, "step": 8415 }, { "epoch": 1.0706017046177332, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.707033395767212, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.867855429649353, "num_tokens": 319460518.0, "step": 8416 }, { "epoch": 1.0707289148963237, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5210063457489014, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8653937578201294, "num_tokens": 319504044.0, "step": 8417 }, { "epoch": 1.070856125174914, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.702791929244995, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8732403516769409, "num_tokens": 319540707.0, "step": 8418 }, { "epoch": 1.0709833354535045, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4840195178985596, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8770155906677246, "num_tokens": 319580968.0, "step": 8419 }, { "epoch": 1.071110545732095, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.712902784347534, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8617144823074341, "num_tokens": 319619271.0, "step": 8420 }, { "epoch": 1.0712377560106856, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.489122152328491, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.864799976348877, "num_tokens": 319657728.0, "step": 8421 }, { "epoch": 1.0713649662892761, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7153122425079346, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8630262613296509, "num_tokens": 319696807.0, "step": 8422 }, { "epoch": 1.0714921765678667, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4924581050872803, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8613172769546509, "num_tokens": 319734397.0, "step": 8423 }, { "epoch": 1.0716193868464572, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7370288372039795, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8566457033157349, "num_tokens": 319769247.0, "step": 8424 }, { "epoch": 1.0717465971250477, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.499994993209839, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8676988482475281, "num_tokens": 319801147.0, "step": 8425 }, { "epoch": 1.0718738074036382, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.730550527572632, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.876866340637207, "num_tokens": 319834083.0, "step": 8426 }, { "epoch": 1.0720010176822288, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4930224418640137, "learning_rate": 1e-05, "loss": 0.3819, "mean_token_accuracy": 0.8836688995361328, "num_tokens": 319872586.0, "step": 8427 }, { "epoch": 1.0721282279608193, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7540438175201416, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8715742826461792, "num_tokens": 319902756.0, "step": 8428 }, { "epoch": 1.0722554382394098, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4877262115478516, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8744970560073853, "num_tokens": 319943610.0, "step": 8429 }, { "epoch": 1.0723826485180004, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020122528076171875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7558400630950928, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8647303581237793, "num_tokens": 319983762.0, "step": 8430 }, { "epoch": 1.0725098587965907, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4975361824035645, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8717169761657715, "num_tokens": 320019560.0, "step": 8431 }, { "epoch": 1.0726370690751812, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.769101858139038, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8645498752593994, "num_tokens": 320052907.0, "step": 8432 }, { "epoch": 1.0727642793537717, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.515373468399048, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8548589944839478, "num_tokens": 320087670.0, "step": 8433 }, { "epoch": 1.0728914896323622, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.744412660598755, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8708182573318481, "num_tokens": 320125768.0, "step": 8434 }, { "epoch": 1.0730186999109528, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5037105083465576, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8793215155601501, "num_tokens": 320163498.0, "step": 8435 }, { "epoch": 1.0731459101895433, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7474417686462402, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8665351271629333, "num_tokens": 320200951.0, "step": 8436 }, { "epoch": 1.0732731204681338, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.512303113937378, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8780899047851562, "num_tokens": 320239258.0, "step": 8437 }, { "epoch": 1.0734003307467244, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7348976135253906, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8667371273040771, "num_tokens": 320276872.0, "step": 8438 }, { "epoch": 1.073527541025315, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.522285223007202, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8679945468902588, "num_tokens": 320315340.0, "step": 8439 }, { "epoch": 1.0736547513039054, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.723935842514038, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8775104880332947, "num_tokens": 320351213.0, "step": 8440 }, { "epoch": 1.073781961582496, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.517427444458008, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8666117191314697, "num_tokens": 320391456.0, "step": 8441 }, { "epoch": 1.0739091718610865, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7381083965301514, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8759735822677612, "num_tokens": 320428193.0, "step": 8442 }, { "epoch": 1.0740363821396768, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5074331760406494, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8748202919960022, "num_tokens": 320466685.0, "step": 8443 }, { "epoch": 1.0741635924182673, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.726247549057007, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8622723817825317, "num_tokens": 320508936.0, "step": 8444 }, { "epoch": 1.0742908026968578, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5263960361480713, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8508554697036743, "num_tokens": 320546644.0, "step": 8445 }, { "epoch": 1.0744180129754484, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7303695678710938, "learning_rate": 1e-05, "loss": 0.3613, "mean_token_accuracy": 0.890912652015686, "num_tokens": 320571074.0, "step": 8446 }, { "epoch": 1.074545223254039, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5216758251190186, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.865915060043335, "num_tokens": 320605412.0, "step": 8447 }, { "epoch": 1.0746724335326294, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7024457454681396, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8645576238632202, "num_tokens": 320643077.0, "step": 8448 }, { "epoch": 1.07479964381122, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5090372562408447, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8775306940078735, "num_tokens": 320680630.0, "step": 8449 }, { "epoch": 1.0749268540898105, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.683168649673462, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8806184530258179, "num_tokens": 320720405.0, "step": 8450 }, { "epoch": 1.075054064368401, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5267715454101562, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8741164207458496, "num_tokens": 320753453.0, "step": 8451 }, { "epoch": 1.0751812746469915, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6605114936828613, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8727903366088867, "num_tokens": 320795316.0, "step": 8452 }, { "epoch": 1.075308484925582, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.51348614692688, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8743463754653931, "num_tokens": 320834022.0, "step": 8453 }, { "epoch": 1.0754356952041726, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.672057867050171, "learning_rate": 1e-05, "loss": 0.5426, "mean_token_accuracy": 0.8360060453414917, "num_tokens": 320878410.0, "step": 8454 }, { "epoch": 1.0755629054827631, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4930193424224854, "learning_rate": 1e-05, "loss": 0.373, "mean_token_accuracy": 0.8888734579086304, "num_tokens": 320918767.0, "step": 8455 }, { "epoch": 1.0756901157613534, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.670954704284668, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8726370334625244, "num_tokens": 320963845.0, "step": 8456 }, { "epoch": 1.075817326039944, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4860854148864746, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8755972385406494, "num_tokens": 321004811.0, "step": 8457 }, { "epoch": 1.0759445363185345, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6908082962036133, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8666059970855713, "num_tokens": 321040894.0, "step": 8458 }, { "epoch": 1.076071746597125, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.488917589187622, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8530004024505615, "num_tokens": 321083602.0, "step": 8459 }, { "epoch": 1.0761989568757155, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.700988292694092, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8683122396469116, "num_tokens": 321119444.0, "step": 8460 }, { "epoch": 1.076326167154306, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.50103497505188, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8754024505615234, "num_tokens": 321155125.0, "step": 8461 }, { "epoch": 1.0764533774328966, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.685120105743408, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8787766098976135, "num_tokens": 321193645.0, "step": 8462 }, { "epoch": 1.0765805877114871, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5159733295440674, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8562579154968262, "num_tokens": 321227118.0, "step": 8463 }, { "epoch": 1.0767077979900777, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.688042640686035, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8682494163513184, "num_tokens": 321263245.0, "step": 8464 }, { "epoch": 1.0768350082686682, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5014593601226807, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.8478485345840454, "num_tokens": 321301476.0, "step": 8465 }, { "epoch": 1.0769622185472587, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6960737705230713, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8723694682121277, "num_tokens": 321333464.0, "step": 8466 }, { "epoch": 1.077089428825849, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.50683856010437, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8695414066314697, "num_tokens": 321367807.0, "step": 8467 }, { "epoch": 1.0772166391044395, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.672008991241455, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8802719116210938, "num_tokens": 321408638.0, "step": 8468 }, { "epoch": 1.07734384938303, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4979889392852783, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8659203052520752, "num_tokens": 321451036.0, "step": 8469 }, { "epoch": 1.0774710596616206, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.666309356689453, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8822165131568909, "num_tokens": 321491423.0, "step": 8470 }, { "epoch": 1.0775982699402111, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.504763603210449, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8614914417266846, "num_tokens": 321527956.0, "step": 8471 }, { "epoch": 1.0777254802188017, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.67741322517395, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.8805804252624512, "num_tokens": 321560185.0, "step": 8472 }, { "epoch": 1.0778526904973922, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.490546464920044, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8750182390213013, "num_tokens": 321602073.0, "step": 8473 }, { "epoch": 1.0779799007759827, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6920359134674072, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8677865266799927, "num_tokens": 321636722.0, "step": 8474 }, { "epoch": 1.0781071110545732, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4845781326293945, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8712275624275208, "num_tokens": 321675348.0, "step": 8475 }, { "epoch": 1.0782343213331638, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7070224285125732, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8707029223442078, "num_tokens": 321711983.0, "step": 8476 }, { "epoch": 1.0783615316117543, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4778892993927, "learning_rate": 1e-05, "loss": 0.3875, "mean_token_accuracy": 0.8852568864822388, "num_tokens": 321748268.0, "step": 8477 }, { "epoch": 1.0784887418903448, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7271406650543213, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.866891086101532, "num_tokens": 321783082.0, "step": 8478 }, { "epoch": 1.0786159521689354, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4818618297576904, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8636744022369385, "num_tokens": 321823634.0, "step": 8479 }, { "epoch": 1.0787431624475257, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7238569259643555, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.877260684967041, "num_tokens": 321859033.0, "step": 8480 }, { "epoch": 1.0788703727261162, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.495656728744507, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8674720525741577, "num_tokens": 321896227.0, "step": 8481 }, { "epoch": 1.0789975830047067, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7267940044403076, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8699424266815186, "num_tokens": 321934329.0, "step": 8482 }, { "epoch": 1.0791247932832972, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4812381267547607, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.864331841468811, "num_tokens": 321977426.0, "step": 8483 }, { "epoch": 1.0792520035618878, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7335777282714844, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8711264133453369, "num_tokens": 322014603.0, "step": 8484 }, { "epoch": 1.0793792138404783, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.491279363632202, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8793042898178101, "num_tokens": 322051809.0, "step": 8485 }, { "epoch": 1.0795064241190688, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.734606981277466, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8682605028152466, "num_tokens": 322091628.0, "step": 8486 }, { "epoch": 1.0796336343976594, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.509422779083252, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8629428148269653, "num_tokens": 322130616.0, "step": 8487 }, { "epoch": 1.0797608446762499, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7038509845733643, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8836499452590942, "num_tokens": 322172525.0, "step": 8488 }, { "epoch": 1.0798880549548404, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.509150266647339, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8758937120437622, "num_tokens": 322213438.0, "step": 8489 }, { "epoch": 1.080015265233431, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6997528076171875, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8767392039299011, "num_tokens": 322252492.0, "step": 8490 }, { "epoch": 1.0801424755120215, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5029685497283936, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8716497421264648, "num_tokens": 322295320.0, "step": 8491 }, { "epoch": 1.0802696857906118, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.689387798309326, "learning_rate": 1e-05, "loss": 0.3955, "mean_token_accuracy": 0.8848356604576111, "num_tokens": 322335475.0, "step": 8492 }, { "epoch": 1.0803968960692023, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5101165771484375, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.87471604347229, "num_tokens": 322373559.0, "step": 8493 }, { "epoch": 1.0805241063477928, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.698277235031128, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8609702587127686, "num_tokens": 322407978.0, "step": 8494 }, { "epoch": 1.0806513166263834, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.504499673843384, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8699297904968262, "num_tokens": 322443657.0, "step": 8495 }, { "epoch": 1.080778526904974, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6964664459228516, "learning_rate": 1e-05, "loss": 0.381, "mean_token_accuracy": 0.8883833289146423, "num_tokens": 322479424.0, "step": 8496 }, { "epoch": 1.0809057371835644, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4971213340759277, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8564122915267944, "num_tokens": 322523112.0, "step": 8497 }, { "epoch": 1.081032947462155, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.703460931777954, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8753026127815247, "num_tokens": 322560444.0, "step": 8498 }, { "epoch": 1.0811601577407455, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5131571292877197, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8716703653335571, "num_tokens": 322591211.0, "step": 8499 }, { "epoch": 1.081287368019336, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.713660478591919, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.8521011471748352, "num_tokens": 322625129.0, "step": 8500 }, { "epoch": 1.0814145782979265, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4915664196014404, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8737713098526001, "num_tokens": 322663428.0, "step": 8501 }, { "epoch": 1.081541788576517, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6968185901641846, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8785312175750732, "num_tokens": 322707936.0, "step": 8502 }, { "epoch": 1.0816689988551076, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.490583896636963, "learning_rate": 1e-05, "loss": 0.4013, "mean_token_accuracy": 0.8828531503677368, "num_tokens": 322748605.0, "step": 8503 }, { "epoch": 1.0817962091336981, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7089767456054688, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8525680899620056, "num_tokens": 322788009.0, "step": 8504 }, { "epoch": 1.0819234194122884, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.503483295440674, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8669413328170776, "num_tokens": 322825159.0, "step": 8505 }, { "epoch": 1.082050629690879, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.710440158843994, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8445457220077515, "num_tokens": 322866769.0, "step": 8506 }, { "epoch": 1.0821778399694695, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5024969577789307, "learning_rate": 1e-05, "loss": 0.3873, "mean_token_accuracy": 0.8798578381538391, "num_tokens": 322901851.0, "step": 8507 }, { "epoch": 1.08230505024806, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7355597019195557, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8740093111991882, "num_tokens": 322938527.0, "step": 8508 }, { "epoch": 1.0824322605266505, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.50053071975708, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8741086721420288, "num_tokens": 322972135.0, "step": 8509 }, { "epoch": 1.082559470805241, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7271037101745605, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8597537279129028, "num_tokens": 323010451.0, "step": 8510 }, { "epoch": 1.0826866810838316, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.492555856704712, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8714768886566162, "num_tokens": 323047583.0, "step": 8511 }, { "epoch": 1.0828138913624221, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7202200889587402, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8637421727180481, "num_tokens": 323088754.0, "step": 8512 }, { "epoch": 1.0829411016410126, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.486797332763672, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8749280571937561, "num_tokens": 323128510.0, "step": 8513 }, { "epoch": 1.0830683119196032, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000202178955078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7273812294006348, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8654583096504211, "num_tokens": 323172557.0, "step": 8514 }, { "epoch": 1.0831955221981937, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4900569915771484, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.8803153038024902, "num_tokens": 323208298.0, "step": 8515 }, { "epoch": 1.083322732476784, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7248623371124268, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8606874942779541, "num_tokens": 323250824.0, "step": 8516 }, { "epoch": 1.0834499427553745, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.494394063949585, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8717814683914185, "num_tokens": 323290792.0, "step": 8517 }, { "epoch": 1.083577153033965, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7339088916778564, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8664236068725586, "num_tokens": 323327806.0, "step": 8518 }, { "epoch": 1.0837043633125556, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.511444091796875, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8699864149093628, "num_tokens": 323364275.0, "step": 8519 }, { "epoch": 1.0838315735911461, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.732503652572632, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8776052594184875, "num_tokens": 323395671.0, "step": 8520 }, { "epoch": 1.0839587838697367, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.506248712539673, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8717907667160034, "num_tokens": 323432719.0, "step": 8521 }, { "epoch": 1.0840859941483272, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7175099849700928, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8624228835105896, "num_tokens": 323474431.0, "step": 8522 }, { "epoch": 1.0842132044269177, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.502751588821411, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8741356134414673, "num_tokens": 323515439.0, "step": 8523 }, { "epoch": 1.0843404147055082, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7222816944122314, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8605115413665771, "num_tokens": 323552613.0, "step": 8524 }, { "epoch": 1.0844676249840988, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.501595973968506, "learning_rate": 1e-05, "loss": 0.3734, "mean_token_accuracy": 0.889999508857727, "num_tokens": 323590037.0, "step": 8525 }, { "epoch": 1.0845948352626893, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.703104019165039, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8812651634216309, "num_tokens": 323634540.0, "step": 8526 }, { "epoch": 1.0847220455412798, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.521834373474121, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.872896671295166, "num_tokens": 323666772.0, "step": 8527 }, { "epoch": 1.0848492558198704, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6953814029693604, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8697729110717773, "num_tokens": 323707847.0, "step": 8528 }, { "epoch": 1.0849764660984607, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5259053707122803, "learning_rate": 1e-05, "loss": 0.382, "mean_token_accuracy": 0.886329174041748, "num_tokens": 323737648.0, "step": 8529 }, { "epoch": 1.0851036763770512, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.677858591079712, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8762224912643433, "num_tokens": 323772621.0, "step": 8530 }, { "epoch": 1.0852308866556417, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.54360032081604, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8594582676887512, "num_tokens": 323806384.0, "step": 8531 }, { "epoch": 1.0853580969342322, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6436541080474854, "learning_rate": 1e-05, "loss": 0.3713, "mean_token_accuracy": 0.8904529213905334, "num_tokens": 323845287.0, "step": 8532 }, { "epoch": 1.0854853072128228, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5177149772644043, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8770921230316162, "num_tokens": 323884182.0, "step": 8533 }, { "epoch": 1.0856125174914133, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.665698766708374, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8529949188232422, "num_tokens": 323919357.0, "step": 8534 }, { "epoch": 1.0857397277700038, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.509244918823242, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8788272142410278, "num_tokens": 323957993.0, "step": 8535 }, { "epoch": 1.0858669380485944, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6581082344055176, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8821864128112793, "num_tokens": 323991726.0, "step": 8536 }, { "epoch": 1.0859941483271849, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.504453182220459, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8699179887771606, "num_tokens": 324031639.0, "step": 8537 }, { "epoch": 1.0861213586057754, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.683708667755127, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8613070249557495, "num_tokens": 324063359.0, "step": 8538 }, { "epoch": 1.086248568884366, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4920225143432617, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8671069145202637, "num_tokens": 324105058.0, "step": 8539 }, { "epoch": 1.0863757791629565, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.689359426498413, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8605256676673889, "num_tokens": 324144209.0, "step": 8540 }, { "epoch": 1.0865029894415468, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.498439311981201, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8750890493392944, "num_tokens": 324179194.0, "step": 8541 }, { "epoch": 1.0866301997201373, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.679582118988037, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8784956932067871, "num_tokens": 324218405.0, "step": 8542 }, { "epoch": 1.0867574099987278, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.499826431274414, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8460427522659302, "num_tokens": 324258298.0, "step": 8543 }, { "epoch": 1.0868846202773184, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6818435192108154, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8788331747055054, "num_tokens": 324296518.0, "step": 8544 }, { "epoch": 1.0870118305559089, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4954206943511963, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8688188195228577, "num_tokens": 324336724.0, "step": 8545 }, { "epoch": 1.0871390408344994, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.695863723754883, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8640779256820679, "num_tokens": 324373464.0, "step": 8546 }, { "epoch": 1.08726625111309, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.501412868499756, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8749669790267944, "num_tokens": 324410893.0, "step": 8547 }, { "epoch": 1.0873934613916805, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6968564987182617, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.876750111579895, "num_tokens": 324446765.0, "step": 8548 }, { "epoch": 1.087520671670271, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.497738838195801, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8724238872528076, "num_tokens": 324482745.0, "step": 8549 }, { "epoch": 1.0876478819488615, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7058939933776855, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.876815676689148, "num_tokens": 324513764.0, "step": 8550 }, { "epoch": 1.087775092227452, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.484452247619629, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8694829940795898, "num_tokens": 324549502.0, "step": 8551 }, { "epoch": 1.0879023025060426, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7168610095977783, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8707369565963745, "num_tokens": 324584456.0, "step": 8552 }, { "epoch": 1.0880295127846331, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4865975379943848, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8638776540756226, "num_tokens": 324624070.0, "step": 8553 }, { "epoch": 1.0881567230632234, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.717987298965454, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8697196245193481, "num_tokens": 324665971.0, "step": 8554 }, { "epoch": 1.088283933341814, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.474912405014038, "learning_rate": 1e-05, "loss": 0.386, "mean_token_accuracy": 0.8838923573493958, "num_tokens": 324706705.0, "step": 8555 }, { "epoch": 1.0884111436204045, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7410526275634766, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8620308637619019, "num_tokens": 324744701.0, "step": 8556 }, { "epoch": 1.088538353898995, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4974000453948975, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8687160611152649, "num_tokens": 324781447.0, "step": 8557 }, { "epoch": 1.0886655641775855, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.741739273071289, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.872776210308075, "num_tokens": 324816753.0, "step": 8558 }, { "epoch": 1.088792774456176, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5013465881347656, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8653774261474609, "num_tokens": 324855093.0, "step": 8559 }, { "epoch": 1.0889199847347666, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.738109827041626, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8709416389465332, "num_tokens": 324893107.0, "step": 8560 }, { "epoch": 1.0890471950133571, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.513026714324951, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8686779737472534, "num_tokens": 324927589.0, "step": 8561 }, { "epoch": 1.0891744052919476, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.742626428604126, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8784864544868469, "num_tokens": 324958801.0, "step": 8562 }, { "epoch": 1.0893016155705382, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5096395015716553, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8753034472465515, "num_tokens": 324995617.0, "step": 8563 }, { "epoch": 1.0894288258491287, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7208263874053955, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8658451437950134, "num_tokens": 325035702.0, "step": 8564 }, { "epoch": 1.089556036127719, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5072433948516846, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8603986501693726, "num_tokens": 325078063.0, "step": 8565 }, { "epoch": 1.0896832464063095, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.722485303878784, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8672783374786377, "num_tokens": 325113127.0, "step": 8566 }, { "epoch": 1.0898104566849, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.524644613265991, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8662869930267334, "num_tokens": 325149451.0, "step": 8567 }, { "epoch": 1.0899376669634906, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6958374977111816, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.86817467212677, "num_tokens": 325193265.0, "step": 8568 }, { "epoch": 1.0900648772420811, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.50426983833313, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8739449977874756, "num_tokens": 325235160.0, "step": 8569 }, { "epoch": 1.0901920875206716, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6965301036834717, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8673502802848816, "num_tokens": 325276299.0, "step": 8570 }, { "epoch": 1.0903192977992622, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5209295749664307, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8566581010818481, "num_tokens": 325313509.0, "step": 8571 }, { "epoch": 1.0904465080778527, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6904280185699463, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8590153455734253, "num_tokens": 325355883.0, "step": 8572 }, { "epoch": 1.0905737183564432, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.513195514678955, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8761823177337646, "num_tokens": 325392937.0, "step": 8573 }, { "epoch": 1.0907009286350338, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.706035614013672, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8715171813964844, "num_tokens": 325422132.0, "step": 8574 }, { "epoch": 1.0908281389136243, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5048561096191406, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8668652176856995, "num_tokens": 325460383.0, "step": 8575 }, { "epoch": 1.0909553491922148, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6932828426361084, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8619324564933777, "num_tokens": 325499242.0, "step": 8576 }, { "epoch": 1.0910825594708053, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4957053661346436, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8768980503082275, "num_tokens": 325535389.0, "step": 8577 }, { "epoch": 1.0912097697493957, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.695605516433716, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8778052926063538, "num_tokens": 325574459.0, "step": 8578 }, { "epoch": 1.0913369800279862, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.500986099243164, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8551133275032043, "num_tokens": 325616296.0, "step": 8579 }, { "epoch": 1.0914641903065767, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7037601470947266, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8604726791381836, "num_tokens": 325652707.0, "step": 8580 }, { "epoch": 1.0915914005851672, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4947547912597656, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8829376697540283, "num_tokens": 325692423.0, "step": 8581 }, { "epoch": 1.0917186108637578, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7014718055725098, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8712294697761536, "num_tokens": 325731638.0, "step": 8582 }, { "epoch": 1.0918458211423483, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4886064529418945, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8773890733718872, "num_tokens": 325771482.0, "step": 8583 }, { "epoch": 1.0919730314209388, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7212579250335693, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8634673953056335, "num_tokens": 325807447.0, "step": 8584 }, { "epoch": 1.0921002416995294, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5247535705566406, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8719146847724915, "num_tokens": 325848070.0, "step": 8585 }, { "epoch": 1.0922274519781199, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6965413093566895, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8767074346542358, "num_tokens": 325889027.0, "step": 8586 }, { "epoch": 1.0923546622567104, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5145487785339355, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8669301271438599, "num_tokens": 325926547.0, "step": 8587 }, { "epoch": 1.092481872535301, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6733572483062744, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8751940727233887, "num_tokens": 325968968.0, "step": 8588 }, { "epoch": 1.0926090828138915, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5156118869781494, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8714491128921509, "num_tokens": 326005060.0, "step": 8589 }, { "epoch": 1.0927362930924818, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6688218116760254, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8591060638427734, "num_tokens": 326049895.0, "step": 8590 }, { "epoch": 1.0928635033710723, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5223987102508545, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8618545532226562, "num_tokens": 326087772.0, "step": 8591 }, { "epoch": 1.0929907136496628, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6574184894561768, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8743746280670166, "num_tokens": 326126518.0, "step": 8592 }, { "epoch": 1.0931179239282534, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.518982172012329, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.861503005027771, "num_tokens": 326166672.0, "step": 8593 }, { "epoch": 1.0932451342068439, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.649060010910034, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8656023740768433, "num_tokens": 326209099.0, "step": 8594 }, { "epoch": 1.0933723444854344, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5285842418670654, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8621970415115356, "num_tokens": 326244101.0, "step": 8595 }, { "epoch": 1.093499554764025, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.640814781188965, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8685101270675659, "num_tokens": 326279997.0, "step": 8596 }, { "epoch": 1.0936267650426155, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5113606452941895, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8624202013015747, "num_tokens": 326321241.0, "step": 8597 }, { "epoch": 1.093753975321206, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.637471914291382, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8693536520004272, "num_tokens": 326359662.0, "step": 8598 }, { "epoch": 1.0938811855997965, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5081627368927, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.879135251045227, "num_tokens": 326397072.0, "step": 8599 }, { "epoch": 1.094008395878387, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6368327140808105, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.866460919380188, "num_tokens": 326440029.0, "step": 8600 }, { "epoch": 1.0941356061569776, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5033206939697266, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8727615475654602, "num_tokens": 326479586.0, "step": 8601 }, { "epoch": 1.094262816435568, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6500895023345947, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.868034839630127, "num_tokens": 326518235.0, "step": 8602 }, { "epoch": 1.0943900267141584, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.48982572555542, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8747528195381165, "num_tokens": 326559420.0, "step": 8603 }, { "epoch": 1.094517236992749, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.669527053833008, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8652248382568359, "num_tokens": 326597020.0, "step": 8604 }, { "epoch": 1.0946444472713395, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4978034496307373, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.8447105884552002, "num_tokens": 326632248.0, "step": 8605 }, { "epoch": 1.09477165754993, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.681352376937866, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.873088002204895, "num_tokens": 326666185.0, "step": 8606 }, { "epoch": 1.0948988678285205, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.536241054534912, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8515005707740784, "num_tokens": 326700356.0, "step": 8607 }, { "epoch": 1.095026078107111, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.650578260421753, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8731893301010132, "num_tokens": 326739448.0, "step": 8608 }, { "epoch": 1.0951532883857016, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.503693103790283, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8704513311386108, "num_tokens": 326776290.0, "step": 8609 }, { "epoch": 1.0952804986642921, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6452791690826416, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.874390184879303, "num_tokens": 326817975.0, "step": 8610 }, { "epoch": 1.0954077089428826, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5217041969299316, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8728708028793335, "num_tokens": 326852404.0, "step": 8611 }, { "epoch": 1.0955349192214732, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6527624130249023, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8607857823371887, "num_tokens": 326888519.0, "step": 8612 }, { "epoch": 1.0956621295000637, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5117619037628174, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8631256818771362, "num_tokens": 326925644.0, "step": 8613 }, { "epoch": 1.095789339778654, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6549832820892334, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8674261569976807, "num_tokens": 326961490.0, "step": 8614 }, { "epoch": 1.0959165500572445, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5079474449157715, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8588476777076721, "num_tokens": 326998726.0, "step": 8615 }, { "epoch": 1.096043760335835, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.654015302658081, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.855579137802124, "num_tokens": 327040717.0, "step": 8616 }, { "epoch": 1.0961709706144256, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.553915500640869, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.867199182510376, "num_tokens": 327070597.0, "step": 8617 }, { "epoch": 1.0962981808930161, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.654971122741699, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.853827953338623, "num_tokens": 327108598.0, "step": 8618 }, { "epoch": 1.0964253911716066, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.508192539215088, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8803902268409729, "num_tokens": 327142750.0, "step": 8619 }, { "epoch": 1.0965526014501972, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6374335289001465, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8598672151565552, "num_tokens": 327182536.0, "step": 8620 }, { "epoch": 1.0966798117287877, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5049641132354736, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8683484792709351, "num_tokens": 327218135.0, "step": 8621 }, { "epoch": 1.0968070220073782, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.655270576477051, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8815450668334961, "num_tokens": 327251581.0, "step": 8622 }, { "epoch": 1.0969342322859688, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4995949268341064, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8696200251579285, "num_tokens": 327288466.0, "step": 8623 }, { "epoch": 1.0970614425645593, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 3.044752359390259, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8741100430488586, "num_tokens": 327325646.0, "step": 8624 }, { "epoch": 1.0971886528431498, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.493964672088623, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8629447817802429, "num_tokens": 327358244.0, "step": 8625 }, { "epoch": 1.0973158631217403, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8202807903289795, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8773481845855713, "num_tokens": 327397315.0, "step": 8626 }, { "epoch": 1.0974430734003306, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.4823391437530518, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8699226379394531, "num_tokens": 327432642.0, "step": 8627 }, { "epoch": 1.0975702836789212, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8583014011383057, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8584284782409668, "num_tokens": 327478563.0, "step": 8628 }, { "epoch": 1.0976974939575117, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.510979175567627, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8764649629592896, "num_tokens": 327514295.0, "step": 8629 }, { "epoch": 1.0978247042361022, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8596365451812744, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8768821358680725, "num_tokens": 327550496.0, "step": 8630 }, { "epoch": 1.0979519145146928, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5355117321014404, "learning_rate": 1e-05, "loss": 0.3921, "mean_token_accuracy": 0.8861631155014038, "num_tokens": 327584692.0, "step": 8631 }, { "epoch": 1.0980791247932833, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.836276054382324, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8778257966041565, "num_tokens": 327627710.0, "step": 8632 }, { "epoch": 1.0982063350718738, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020313262939453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5518076419830322, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8633559346199036, "num_tokens": 327666958.0, "step": 8633 }, { "epoch": 1.0983335453504643, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8199269771575928, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8743699193000793, "num_tokens": 327700829.0, "step": 8634 }, { "epoch": 1.0984607556290549, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.566061019897461, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8612192869186401, "num_tokens": 327741410.0, "step": 8635 }, { "epoch": 1.0985879659076454, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7950031757354736, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8557283878326416, "num_tokens": 327778305.0, "step": 8636 }, { "epoch": 1.098715176186236, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.571296215057373, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8594809770584106, "num_tokens": 327813384.0, "step": 8637 }, { "epoch": 1.0988423864648265, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.751626491546631, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8728102445602417, "num_tokens": 327851928.0, "step": 8638 }, { "epoch": 1.0989695967434168, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5466082096099854, "learning_rate": 1e-05, "loss": 0.4004, "mean_token_accuracy": 0.8805489540100098, "num_tokens": 327890053.0, "step": 8639 }, { "epoch": 1.0990968070220073, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.733487606048584, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.868606448173523, "num_tokens": 327928947.0, "step": 8640 }, { "epoch": 1.0992240173005978, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5518393516540527, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8739562034606934, "num_tokens": 327966286.0, "step": 8641 }, { "epoch": 1.0993512275791884, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.707657814025879, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8655372262001038, "num_tokens": 328005480.0, "step": 8642 }, { "epoch": 1.0994784378577789, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.543081521987915, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8738037347793579, "num_tokens": 328040587.0, "step": 8643 }, { "epoch": 1.0996056481363694, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.695810556411743, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8641905188560486, "num_tokens": 328075856.0, "step": 8644 }, { "epoch": 1.09973285841496, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.564647674560547, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8585188984870911, "num_tokens": 328116243.0, "step": 8645 }, { "epoch": 1.0998600686935505, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.664527177810669, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8727786540985107, "num_tokens": 328151854.0, "step": 8646 }, { "epoch": 1.099987278972141, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.539640188217163, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8629299998283386, "num_tokens": 328191455.0, "step": 8647 }, { "epoch": 1.1001144892507315, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.641126871109009, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.874883770942688, "num_tokens": 328232034.0, "step": 8648 }, { "epoch": 1.100241699529322, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5465447902679443, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8705918788909912, "num_tokens": 328265581.0, "step": 8649 }, { "epoch": 1.1003689098079126, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6445043087005615, "learning_rate": 1e-05, "loss": 0.5105, "mean_token_accuracy": 0.8522669076919556, "num_tokens": 328304354.0, "step": 8650 }, { "epoch": 1.100496120086503, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.525493621826172, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8605835437774658, "num_tokens": 328347924.0, "step": 8651 }, { "epoch": 1.1006233303650934, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.624178171157837, "learning_rate": 1e-05, "loss": 0.3925, "mean_token_accuracy": 0.8842290639877319, "num_tokens": 328385603.0, "step": 8652 }, { "epoch": 1.100750540643684, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.530766010284424, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8767766356468201, "num_tokens": 328422881.0, "step": 8653 }, { "epoch": 1.1008777509222745, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.625358819961548, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.866235613822937, "num_tokens": 328461191.0, "step": 8654 }, { "epoch": 1.101004961200865, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.515021324157715, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8692407608032227, "num_tokens": 328499824.0, "step": 8655 }, { "epoch": 1.1011321714794555, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.628394603729248, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8754286766052246, "num_tokens": 328538456.0, "step": 8656 }, { "epoch": 1.101259381758046, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.531268835067749, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8553364276885986, "num_tokens": 328574722.0, "step": 8657 }, { "epoch": 1.1013865920366366, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.626909017562866, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8725572228431702, "num_tokens": 328608248.0, "step": 8658 }, { "epoch": 1.101513802315227, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5072596073150635, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8699142932891846, "num_tokens": 328651011.0, "step": 8659 }, { "epoch": 1.1016410125938176, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6201059818267822, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8645387887954712, "num_tokens": 328694438.0, "step": 8660 }, { "epoch": 1.1017682228724082, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5091567039489746, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8694149851799011, "num_tokens": 328731851.0, "step": 8661 }, { "epoch": 1.1018954331509987, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.636538505554199, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8698694705963135, "num_tokens": 328772159.0, "step": 8662 }, { "epoch": 1.102022643429589, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4877243041992188, "learning_rate": 1e-05, "loss": 0.3849, "mean_token_accuracy": 0.8856605887413025, "num_tokens": 328814712.0, "step": 8663 }, { "epoch": 1.1021498537081795, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6531922817230225, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.865646481513977, "num_tokens": 328854243.0, "step": 8664 }, { "epoch": 1.10227706398677, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4965951442718506, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8585220575332642, "num_tokens": 328894109.0, "step": 8665 }, { "epoch": 1.1024042742653606, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6493945121765137, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8807556629180908, "num_tokens": 328937275.0, "step": 8666 }, { "epoch": 1.1025314845439511, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4943344593048096, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.8781253099441528, "num_tokens": 328974572.0, "step": 8667 }, { "epoch": 1.1026586948225416, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6657421588897705, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8643820881843567, "num_tokens": 329013871.0, "step": 8668 }, { "epoch": 1.1027859051011322, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.506038188934326, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8563110828399658, "num_tokens": 329050333.0, "step": 8669 }, { "epoch": 1.1029131153797227, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.664106607437134, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8646166920661926, "num_tokens": 329091817.0, "step": 8670 }, { "epoch": 1.1030403256583132, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.495816946029663, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8747335076332092, "num_tokens": 329128523.0, "step": 8671 }, { "epoch": 1.1031675359369038, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6678049564361572, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8860293030738831, "num_tokens": 329166320.0, "step": 8672 }, { "epoch": 1.1032947462154943, "ewc_loss": 0.056884765625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5124340057373047, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8727037906646729, "num_tokens": 329195140.0, "step": 8673 }, { "epoch": 1.1034219564940848, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.669604539871216, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8791694045066833, "num_tokens": 329230814.0, "step": 8674 }, { "epoch": 1.1035491667726753, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4872336387634277, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8761882781982422, "num_tokens": 329270695.0, "step": 8675 }, { "epoch": 1.1036763770512656, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6944432258605957, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.856211245059967, "num_tokens": 329308061.0, "step": 8676 }, { "epoch": 1.1038035873298562, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4805729389190674, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8673692941665649, "num_tokens": 329347093.0, "step": 8677 }, { "epoch": 1.1039307976084467, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.705958843231201, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8669885993003845, "num_tokens": 329385009.0, "step": 8678 }, { "epoch": 1.1040580078870372, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4871137142181396, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8688762187957764, "num_tokens": 329425317.0, "step": 8679 }, { "epoch": 1.1041852181656278, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.725847005844116, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8774089217185974, "num_tokens": 329461581.0, "step": 8680 }, { "epoch": 1.1043124284442183, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4864583015441895, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8815697431564331, "num_tokens": 329494064.0, "step": 8681 }, { "epoch": 1.1044396387228088, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7340900897979736, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8756369352340698, "num_tokens": 329530807.0, "step": 8682 }, { "epoch": 1.1045668490013993, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4858896732330322, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8489732146263123, "num_tokens": 329575689.0, "step": 8683 }, { "epoch": 1.1046940592799899, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7460203170776367, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8637383580207825, "num_tokens": 329616855.0, "step": 8684 }, { "epoch": 1.1048212695585804, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4913833141326904, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8710551857948303, "num_tokens": 329654085.0, "step": 8685 }, { "epoch": 1.104948479837171, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7690508365631104, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8691948652267456, "num_tokens": 329688375.0, "step": 8686 }, { "epoch": 1.1050756901157615, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4888949394226074, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8594464063644409, "num_tokens": 329734258.0, "step": 8687 }, { "epoch": 1.1052029003943518, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7712202072143555, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8722056150436401, "num_tokens": 329772080.0, "step": 8688 }, { "epoch": 1.1053301106729423, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4825594425201416, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8777632713317871, "num_tokens": 329817809.0, "step": 8689 }, { "epoch": 1.1054573209515328, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7789814472198486, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8763450384140015, "num_tokens": 329856063.0, "step": 8690 }, { "epoch": 1.1055845312301233, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5068674087524414, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8599194884300232, "num_tokens": 329892548.0, "step": 8691 }, { "epoch": 1.1057117415087139, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7728655338287354, "learning_rate": 1e-05, "loss": 0.3906, "mean_token_accuracy": 0.8856180906295776, "num_tokens": 329931897.0, "step": 8692 }, { "epoch": 1.1058389517873044, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5231740474700928, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8652814030647278, "num_tokens": 329967477.0, "step": 8693 }, { "epoch": 1.105966162065895, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7687127590179443, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8686540126800537, "num_tokens": 330004840.0, "step": 8694 }, { "epoch": 1.1060933723444855, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.513603687286377, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8708518147468567, "num_tokens": 330048928.0, "step": 8695 }, { "epoch": 1.106220582623076, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.765580892562866, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8667383790016174, "num_tokens": 330084112.0, "step": 8696 }, { "epoch": 1.1063477929016665, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5187151432037354, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8602870106697083, "num_tokens": 330127533.0, "step": 8697 }, { "epoch": 1.106475003180257, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.967935562133789, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8498072624206543, "num_tokens": 330169379.0, "step": 8698 }, { "epoch": 1.1066022134588476, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.530733108520508, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8787654638290405, "num_tokens": 330205553.0, "step": 8699 }, { "epoch": 1.106729423737438, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8155856132507324, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8726209998130798, "num_tokens": 330241088.0, "step": 8700 }, { "epoch": 1.1068566340160284, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.52408504486084, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8715623617172241, "num_tokens": 330282520.0, "step": 8701 }, { "epoch": 1.106983844294619, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.818013906478882, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8654580116271973, "num_tokens": 330326231.0, "step": 8702 }, { "epoch": 1.1071110545732095, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.547194480895996, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8801790475845337, "num_tokens": 330363674.0, "step": 8703 }, { "epoch": 1.1072382648518, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7698519229888916, "learning_rate": 1e-05, "loss": 0.3673, "mean_token_accuracy": 0.894485354423523, "num_tokens": 330398890.0, "step": 8704 }, { "epoch": 1.1073654751303905, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5651557445526123, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8758431673049927, "num_tokens": 330434564.0, "step": 8705 }, { "epoch": 1.107492685408981, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.738065481185913, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8765670657157898, "num_tokens": 330475580.0, "step": 8706 }, { "epoch": 1.1076198956875716, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5561397075653076, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8812633156776428, "num_tokens": 330513255.0, "step": 8707 }, { "epoch": 1.107747105966162, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7175023555755615, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8603750467300415, "num_tokens": 330554101.0, "step": 8708 }, { "epoch": 1.1078743162447526, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.558504581451416, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8565738201141357, "num_tokens": 330595312.0, "step": 8709 }, { "epoch": 1.1080015265233432, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.682661533355713, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8776917457580566, "num_tokens": 330633510.0, "step": 8710 }, { "epoch": 1.1081287368019337, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5365214347839355, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8742401599884033, "num_tokens": 330673626.0, "step": 8711 }, { "epoch": 1.108255947080524, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6756200790405273, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8669942617416382, "num_tokens": 330715051.0, "step": 8712 }, { "epoch": 1.1083831573591145, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.513092041015625, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8692880868911743, "num_tokens": 330765603.0, "step": 8713 }, { "epoch": 1.108510367637705, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6965978145599365, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.867045521736145, "num_tokens": 330799427.0, "step": 8714 }, { "epoch": 1.1086375779162956, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.504047393798828, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8735864758491516, "num_tokens": 330840810.0, "step": 8715 }, { "epoch": 1.108764788194886, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7064592838287354, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8585555553436279, "num_tokens": 330878325.0, "step": 8716 }, { "epoch": 1.1088919984734766, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.499142646789551, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8824109435081482, "num_tokens": 330914817.0, "step": 8717 }, { "epoch": 1.1090192087520672, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.701511859893799, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.879573404788971, "num_tokens": 330951890.0, "step": 8718 }, { "epoch": 1.1091464190306577, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5182712078094482, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8597514033317566, "num_tokens": 330986807.0, "step": 8719 }, { "epoch": 1.1092736293092482, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6910552978515625, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8619112968444824, "num_tokens": 331027132.0, "step": 8720 }, { "epoch": 1.1094008395878387, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5104308128356934, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8766999244689941, "num_tokens": 331062669.0, "step": 8721 }, { "epoch": 1.1095280498664293, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6828773021698, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8736841678619385, "num_tokens": 331097376.0, "step": 8722 }, { "epoch": 1.1096552601450198, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5190625190734863, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8790923357009888, "num_tokens": 331130930.0, "step": 8723 }, { "epoch": 1.1097824704236103, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6646735668182373, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8683082461357117, "num_tokens": 331172012.0, "step": 8724 }, { "epoch": 1.1099096807022006, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5237011909484863, "learning_rate": 1e-05, "loss": 0.5039, "mean_token_accuracy": 0.848033607006073, "num_tokens": 331210016.0, "step": 8725 }, { "epoch": 1.1100368909807912, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6643264293670654, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8631987571716309, "num_tokens": 331251818.0, "step": 8726 }, { "epoch": 1.1101641012593817, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5324809551239014, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8579424023628235, "num_tokens": 331283034.0, "step": 8727 }, { "epoch": 1.1102913115379722, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6424171924591064, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8656536340713501, "num_tokens": 331324985.0, "step": 8728 }, { "epoch": 1.1104185218165628, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5324087142944336, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8719857931137085, "num_tokens": 331358595.0, "step": 8729 }, { "epoch": 1.1105457320951533, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.659423589706421, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8650707006454468, "num_tokens": 331387388.0, "step": 8730 }, { "epoch": 1.1106729423737438, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.518439292907715, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8641812801361084, "num_tokens": 331423562.0, "step": 8731 }, { "epoch": 1.1108001526523343, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.640073776245117, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.874424159526825, "num_tokens": 331458835.0, "step": 8732 }, { "epoch": 1.1109273629309249, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.529369592666626, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8623638153076172, "num_tokens": 331489889.0, "step": 8733 }, { "epoch": 1.1110545732095154, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6207191944122314, "learning_rate": 1e-05, "loss": 0.3822, "mean_token_accuracy": 0.882874071598053, "num_tokens": 331529103.0, "step": 8734 }, { "epoch": 1.111181783488106, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.519664764404297, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8762398362159729, "num_tokens": 331566100.0, "step": 8735 }, { "epoch": 1.1113089937666965, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.62113881111145, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8767545223236084, "num_tokens": 331606704.0, "step": 8736 }, { "epoch": 1.1114362040452868, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.537949323654175, "learning_rate": 1e-05, "loss": 0.4058, "mean_token_accuracy": 0.8796252012252808, "num_tokens": 331642019.0, "step": 8737 }, { "epoch": 1.1115634143238773, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6327733993530273, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8697644472122192, "num_tokens": 331672888.0, "step": 8738 }, { "epoch": 1.1116906246024678, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.519714593887329, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.870555579662323, "num_tokens": 331710383.0, "step": 8739 }, { "epoch": 1.1118178348810583, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6121268272399902, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8736671805381775, "num_tokens": 331754299.0, "step": 8740 }, { "epoch": 1.1119450451596489, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.523887872695923, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8722853064537048, "num_tokens": 331789329.0, "step": 8741 }, { "epoch": 1.1120722554382394, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6331777572631836, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8766365051269531, "num_tokens": 331822727.0, "step": 8742 }, { "epoch": 1.11219946571683, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020503997802734375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5038840770721436, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8741304278373718, "num_tokens": 331863091.0, "step": 8743 }, { "epoch": 1.1123266759954205, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.633592367172241, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8733936548233032, "num_tokens": 331900866.0, "step": 8744 }, { "epoch": 1.112453886274011, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.505319833755493, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8840214014053345, "num_tokens": 331936774.0, "step": 8745 }, { "epoch": 1.1125810965526015, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6501691341400146, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8739790916442871, "num_tokens": 331974629.0, "step": 8746 }, { "epoch": 1.112708306831192, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5042319297790527, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8779534101486206, "num_tokens": 332008707.0, "step": 8747 }, { "epoch": 1.1128355171097826, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.650449752807617, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.882391631603241, "num_tokens": 332046973.0, "step": 8748 }, { "epoch": 1.112962727388373, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.486480951309204, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8621366620063782, "num_tokens": 332092658.0, "step": 8749 }, { "epoch": 1.1130899376669634, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.668898105621338, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8728014230728149, "num_tokens": 332129705.0, "step": 8750 }, { "epoch": 1.113217147945554, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5034372806549072, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8707005381584167, "num_tokens": 332163129.0, "step": 8751 }, { "epoch": 1.1133443582241445, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.682101011276245, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8677548766136169, "num_tokens": 332198148.0, "step": 8752 }, { "epoch": 1.113471568502735, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4778079986572266, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8716561198234558, "num_tokens": 332238068.0, "step": 8753 }, { "epoch": 1.1135987787813255, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.702014684677124, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8602560758590698, "num_tokens": 332277455.0, "step": 8754 }, { "epoch": 1.113725989059916, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.4748589992523193, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8667814135551453, "num_tokens": 332314778.0, "step": 8755 }, { "epoch": 1.1138531993385066, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7245545387268066, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8744804263114929, "num_tokens": 332347863.0, "step": 8756 }, { "epoch": 1.113980409617097, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4673445224761963, "learning_rate": 1e-05, "loss": 0.3818, "mean_token_accuracy": 0.884354829788208, "num_tokens": 332389577.0, "step": 8757 }, { "epoch": 1.1141076198956876, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.735419750213623, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.864716649055481, "num_tokens": 332434140.0, "step": 8758 }, { "epoch": 1.1142348301742782, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.484600067138672, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8722257614135742, "num_tokens": 332468843.0, "step": 8759 }, { "epoch": 1.1143620404528687, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7407710552215576, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.8775221109390259, "num_tokens": 332509863.0, "step": 8760 }, { "epoch": 1.114489250731459, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4859189987182617, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8674842119216919, "num_tokens": 332550242.0, "step": 8761 }, { "epoch": 1.1146164610100495, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7620949745178223, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8699312806129456, "num_tokens": 332588606.0, "step": 8762 }, { "epoch": 1.11474367128864, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5026895999908447, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8730452656745911, "num_tokens": 332623787.0, "step": 8763 }, { "epoch": 1.1148708815672306, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.756105661392212, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8621557950973511, "num_tokens": 332665358.0, "step": 8764 }, { "epoch": 1.114998091845821, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.500124931335449, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8811789751052856, "num_tokens": 332700541.0, "step": 8765 }, { "epoch": 1.1151253021244116, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7694222927093506, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8618441224098206, "num_tokens": 332737555.0, "step": 8766 }, { "epoch": 1.1152525124030022, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5080080032348633, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8619083166122437, "num_tokens": 332775957.0, "step": 8767 }, { "epoch": 1.1153797226815927, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.764885187149048, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8747780323028564, "num_tokens": 332812000.0, "step": 8768 }, { "epoch": 1.1155069329601832, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5017049312591553, "learning_rate": 1e-05, "loss": 0.3894, "mean_token_accuracy": 0.8853660821914673, "num_tokens": 332848974.0, "step": 8769 }, { "epoch": 1.1156341432387737, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.75795841217041, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8745232224464417, "num_tokens": 332890362.0, "step": 8770 }, { "epoch": 1.1157613535173643, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.52700138092041, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8679964542388916, "num_tokens": 332924547.0, "step": 8771 }, { "epoch": 1.1158885637959548, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.753403902053833, "learning_rate": 1e-05, "loss": 0.3972, "mean_token_accuracy": 0.8839006423950195, "num_tokens": 332957697.0, "step": 8772 }, { "epoch": 1.1160157740745453, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.531269073486328, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.871309220790863, "num_tokens": 332992671.0, "step": 8773 }, { "epoch": 1.1161429843531356, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7248611450195312, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.875269889831543, "num_tokens": 333033092.0, "step": 8774 }, { "epoch": 1.1162701946317262, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.524975061416626, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8649191856384277, "num_tokens": 333071631.0, "step": 8775 }, { "epoch": 1.1163974049103167, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7201263904571533, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8782045841217041, "num_tokens": 333100931.0, "step": 8776 }, { "epoch": 1.1165246151889072, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5076568126678467, "learning_rate": 1e-05, "loss": 0.3801, "mean_token_accuracy": 0.891998291015625, "num_tokens": 333141063.0, "step": 8777 }, { "epoch": 1.1166518254674977, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7135348320007324, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.870980441570282, "num_tokens": 333180322.0, "step": 8778 }, { "epoch": 1.1167790357460883, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5255186557769775, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8655460476875305, "num_tokens": 333219836.0, "step": 8779 }, { "epoch": 1.1169062460246788, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7008817195892334, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.863653302192688, "num_tokens": 333263748.0, "step": 8780 }, { "epoch": 1.1170334563032693, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5307321548461914, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8716654777526855, "num_tokens": 333296110.0, "step": 8781 }, { "epoch": 1.1171606665818599, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.718143939971924, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8679121732711792, "num_tokens": 333328372.0, "step": 8782 }, { "epoch": 1.1172878768604504, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020599365234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.514611005783081, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8780136108398438, "num_tokens": 333367680.0, "step": 8783 }, { "epoch": 1.117415087139041, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6896960735321045, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8720502853393555, "num_tokens": 333406761.0, "step": 8784 }, { "epoch": 1.1175422974176314, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5089497566223145, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8751309514045715, "num_tokens": 333443268.0, "step": 8785 }, { "epoch": 1.1176695076962218, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.700486183166504, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8597006797790527, "num_tokens": 333478770.0, "step": 8786 }, { "epoch": 1.1177967179748123, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5001649856567383, "learning_rate": 1e-05, "loss": 0.3834, "mean_token_accuracy": 0.8877280354499817, "num_tokens": 333514971.0, "step": 8787 }, { "epoch": 1.1179239282534028, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6970458030700684, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8701962232589722, "num_tokens": 333553610.0, "step": 8788 }, { "epoch": 1.1180511385319933, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4825806617736816, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.87220698595047, "num_tokens": 333604283.0, "step": 8789 }, { "epoch": 1.1181783488105839, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.710313558578491, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.8821970224380493, "num_tokens": 333641621.0, "step": 8790 }, { "epoch": 1.1183055590891744, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5035452842712402, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8808228373527527, "num_tokens": 333676435.0, "step": 8791 }, { "epoch": 1.118432769367765, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.710131883621216, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8824464082717896, "num_tokens": 333715308.0, "step": 8792 }, { "epoch": 1.1185599796463555, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5184884071350098, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.856465756893158, "num_tokens": 333752086.0, "step": 8793 }, { "epoch": 1.118687189924946, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7088282108306885, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8699132800102234, "num_tokens": 333789726.0, "step": 8794 }, { "epoch": 1.1188144002035365, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5241963863372803, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8603835105895996, "num_tokens": 333828518.0, "step": 8795 }, { "epoch": 1.118941610482127, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.686920166015625, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8661550283432007, "num_tokens": 333870520.0, "step": 8796 }, { "epoch": 1.1190688207607176, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5517799854278564, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.874165415763855, "num_tokens": 333914676.0, "step": 8797 }, { "epoch": 1.119196031039308, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6613001823425293, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8764052391052246, "num_tokens": 333953381.0, "step": 8798 }, { "epoch": 1.1193232413178984, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5518312454223633, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8586152791976929, "num_tokens": 333983275.0, "step": 8799 }, { "epoch": 1.119450451596489, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6423754692077637, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8770784735679626, "num_tokens": 334018530.0, "step": 8800 }, { "epoch": 1.1195776618750795, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5295867919921875, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8819682598114014, "num_tokens": 334056606.0, "step": 8801 }, { "epoch": 1.11970487215367, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.626080274581909, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8747332692146301, "num_tokens": 334098541.0, "step": 8802 }, { "epoch": 1.1198320824322605, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.527871608734131, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8622846603393555, "num_tokens": 334140243.0, "step": 8803 }, { "epoch": 1.119959292710851, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.626405715942383, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8792065382003784, "num_tokens": 334177126.0, "step": 8804 }, { "epoch": 1.1200865029894416, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.523285150527954, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8791697025299072, "num_tokens": 334217058.0, "step": 8805 }, { "epoch": 1.120213713268032, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.625715732574463, "learning_rate": 1e-05, "loss": 0.3789, "mean_token_accuracy": 0.8903945684432983, "num_tokens": 334255060.0, "step": 8806 }, { "epoch": 1.1203409235466226, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5280003547668457, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8674290180206299, "num_tokens": 334290661.0, "step": 8807 }, { "epoch": 1.1204681338252132, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6237239837646484, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8830822706222534, "num_tokens": 334328661.0, "step": 8808 }, { "epoch": 1.1205953441038037, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.51580548286438, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8718819618225098, "num_tokens": 334367535.0, "step": 8809 }, { "epoch": 1.120722554382394, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.631281614303589, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8742030262947083, "num_tokens": 334407366.0, "step": 8810 }, { "epoch": 1.1208497646609845, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5194571018218994, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8745784163475037, "num_tokens": 334444049.0, "step": 8811 }, { "epoch": 1.120976974939575, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.624054193496704, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8657792806625366, "num_tokens": 334485405.0, "step": 8812 }, { "epoch": 1.1211041852181656, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4990642070770264, "learning_rate": 1e-05, "loss": 0.3546, "mean_token_accuracy": 0.8945903778076172, "num_tokens": 334525209.0, "step": 8813 }, { "epoch": 1.121231395496756, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6311609745025635, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8739603757858276, "num_tokens": 334569156.0, "step": 8814 }, { "epoch": 1.1213586057753466, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5125067234039307, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.87917160987854, "num_tokens": 334602692.0, "step": 8815 }, { "epoch": 1.1214858160539372, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.640350580215454, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8666481971740723, "num_tokens": 334644926.0, "step": 8816 }, { "epoch": 1.1216130263325277, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.499002456665039, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8756546974182129, "num_tokens": 334690116.0, "step": 8817 }, { "epoch": 1.1217402366111182, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6454646587371826, "learning_rate": 1e-05, "loss": 0.3968, "mean_token_accuracy": 0.885717511177063, "num_tokens": 334728403.0, "step": 8818 }, { "epoch": 1.1218674468897087, "ewc_loss": 0.057373046875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.513127088546753, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.871888279914856, "num_tokens": 334763394.0, "step": 8819 }, { "epoch": 1.1219946571682993, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6653554439544678, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8802787065505981, "num_tokens": 334795029.0, "step": 8820 }, { "epoch": 1.1221218674468898, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.500763177871704, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8690838813781738, "num_tokens": 334833298.0, "step": 8821 }, { "epoch": 1.1222490777254803, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6735758781433105, "learning_rate": 1e-05, "loss": 0.3873, "mean_token_accuracy": 0.8842734098434448, "num_tokens": 334865146.0, "step": 8822 }, { "epoch": 1.1223762880040706, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.551055431365967, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8585617542266846, "num_tokens": 334899392.0, "step": 8823 }, { "epoch": 1.1225034982826612, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6534135341644287, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.859063446521759, "num_tokens": 334936501.0, "step": 8824 }, { "epoch": 1.1226307085612517, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.5074498653411865, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8631908297538757, "num_tokens": 334975920.0, "step": 8825 }, { "epoch": 1.1227579188398422, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6576387882232666, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.871418833732605, "num_tokens": 335012104.0, "step": 8826 }, { "epoch": 1.1228851291184327, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.506979465484619, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8567221760749817, "num_tokens": 335051302.0, "step": 8827 }, { "epoch": 1.1230123393970233, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6651771068573, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8705295920372009, "num_tokens": 335091778.0, "step": 8828 }, { "epoch": 1.1231395496756138, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5343692302703857, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8573954105377197, "num_tokens": 335120687.0, "step": 8829 }, { "epoch": 1.1232667599542043, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.65498685836792, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8694146871566772, "num_tokens": 335158166.0, "step": 8830 }, { "epoch": 1.1233939702327949, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.501450300216675, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8776151537895203, "num_tokens": 335194674.0, "step": 8831 }, { "epoch": 1.1235211805113854, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6652472019195557, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8770835399627686, "num_tokens": 335229768.0, "step": 8832 }, { "epoch": 1.123648390789976, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4962522983551025, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8754032850265503, "num_tokens": 335270500.0, "step": 8833 }, { "epoch": 1.1237756010685664, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6697473526000977, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8812060356140137, "num_tokens": 335309672.0, "step": 8834 }, { "epoch": 1.1239028113471567, "ewc_loss": 0.057861328125, "ewc_loss_diag": 0.00020694732666015625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5070626735687256, "learning_rate": 1e-05, "loss": 0.391, "mean_token_accuracy": 0.8814822435379028, "num_tokens": 335341505.0, "step": 8835 }, { "epoch": 1.1240300216257473, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.687159776687622, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8583921194076538, "num_tokens": 335379013.0, "step": 8836 }, { "epoch": 1.1241572319043378, "ewc_loss": 0.05712890625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.501311779022217, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8632169961929321, "num_tokens": 335415170.0, "step": 8837 }, { "epoch": 1.1242844421829283, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6840031147003174, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8794253468513489, "num_tokens": 335455476.0, "step": 8838 }, { "epoch": 1.1244116524615189, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4877004623413086, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8680650591850281, "num_tokens": 335497185.0, "step": 8839 }, { "epoch": 1.1245388627401094, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.724759340286255, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8551113605499268, "num_tokens": 335531754.0, "step": 8840 }, { "epoch": 1.1246660730187, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4706921577453613, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8781778216362, "num_tokens": 335570486.0, "step": 8841 }, { "epoch": 1.1247932832972904, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7365574836730957, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8625761270523071, "num_tokens": 335610179.0, "step": 8842 }, { "epoch": 1.124920493575881, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5133674144744873, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8572909832000732, "num_tokens": 335639785.0, "step": 8843 }, { "epoch": 1.1250477038544715, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.724938154220581, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8711434602737427, "num_tokens": 335679012.0, "step": 8844 }, { "epoch": 1.125174914133062, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.490631580352783, "learning_rate": 1e-05, "loss": 0.3935, "mean_token_accuracy": 0.8816875219345093, "num_tokens": 335717495.0, "step": 8845 }, { "epoch": 1.1253021244116526, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.735699415206909, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.872472882270813, "num_tokens": 335751617.0, "step": 8846 }, { "epoch": 1.125429334690243, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5162580013275146, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8638202548027039, "num_tokens": 335782698.0, "step": 8847 }, { "epoch": 1.1255565449688334, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.727445602416992, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8656319379806519, "num_tokens": 335823091.0, "step": 8848 }, { "epoch": 1.125683755247424, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4950010776519775, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8701224327087402, "num_tokens": 335863318.0, "step": 8849 }, { "epoch": 1.1258109655260145, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.727079391479492, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8741451501846313, "num_tokens": 335900467.0, "step": 8850 }, { "epoch": 1.125938175804605, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.495887279510498, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8633226752281189, "num_tokens": 335942274.0, "step": 8851 }, { "epoch": 1.1260653860831955, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7453181743621826, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8683236241340637, "num_tokens": 335975073.0, "step": 8852 }, { "epoch": 1.126192596361786, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5083467960357666, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8573812246322632, "num_tokens": 336011846.0, "step": 8853 }, { "epoch": 1.1263198066403766, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7408981323242188, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8659319877624512, "num_tokens": 336050969.0, "step": 8854 }, { "epoch": 1.126447016918967, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.498056173324585, "learning_rate": 1e-05, "loss": 0.3867, "mean_token_accuracy": 0.885608971118927, "num_tokens": 336087714.0, "step": 8855 }, { "epoch": 1.1265742271975576, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7673938274383545, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8605226874351501, "num_tokens": 336124937.0, "step": 8856 }, { "epoch": 1.1267014374761481, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.4839470386505127, "learning_rate": 1e-05, "loss": 0.3888, "mean_token_accuracy": 0.8863270282745361, "num_tokens": 336167666.0, "step": 8857 }, { "epoch": 1.1268286477547387, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7660791873931885, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.8852214813232422, "num_tokens": 336202170.0, "step": 8858 }, { "epoch": 1.126955858033329, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.507936954498291, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.880230188369751, "num_tokens": 336235613.0, "step": 8859 }, { "epoch": 1.1270830683119195, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7653775215148926, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8681936860084534, "num_tokens": 336272027.0, "step": 8860 }, { "epoch": 1.12721027859051, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5168728828430176, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8569345474243164, "num_tokens": 336312554.0, "step": 8861 }, { "epoch": 1.1273374888691006, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.743762493133545, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8709008693695068, "num_tokens": 336356719.0, "step": 8862 }, { "epoch": 1.127464699147691, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5137362480163574, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8811109066009521, "num_tokens": 336395242.0, "step": 8863 }, { "epoch": 1.1275919094262816, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7412610054016113, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8676143884658813, "num_tokens": 336431738.0, "step": 8864 }, { "epoch": 1.1277191197048722, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.533735752105713, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8695014715194702, "num_tokens": 336467799.0, "step": 8865 }, { "epoch": 1.1278463299834627, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7334020137786865, "learning_rate": 1e-05, "loss": 0.5042, "mean_token_accuracy": 0.8509505391120911, "num_tokens": 336506628.0, "step": 8866 }, { "epoch": 1.1279735402620532, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.520280122756958, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8791355490684509, "num_tokens": 336546034.0, "step": 8867 }, { "epoch": 1.1281007505406437, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7289557456970215, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8695501089096069, "num_tokens": 336578561.0, "step": 8868 }, { "epoch": 1.1282279608192343, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.523900270462036, "learning_rate": 1e-05, "loss": 0.3852, "mean_token_accuracy": 0.8841885328292847, "num_tokens": 336610500.0, "step": 8869 }, { "epoch": 1.1283551710978248, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.699615716934204, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.879462480545044, "num_tokens": 336653889.0, "step": 8870 }, { "epoch": 1.1284823813764153, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5228054523468018, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8676818609237671, "num_tokens": 336690605.0, "step": 8871 }, { "epoch": 1.1286095916550056, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.69709849357605, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8661758899688721, "num_tokens": 336729159.0, "step": 8872 }, { "epoch": 1.1287368019335962, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5353856086730957, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8777749538421631, "num_tokens": 336762376.0, "step": 8873 }, { "epoch": 1.1288640122121867, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.686278820037842, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8769928216934204, "num_tokens": 336797060.0, "step": 8874 }, { "epoch": 1.1289912224907772, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.523736000061035, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8526517152786255, "num_tokens": 336838686.0, "step": 8875 }, { "epoch": 1.1291184327693677, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6726582050323486, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8544448614120483, "num_tokens": 336887535.0, "step": 8876 }, { "epoch": 1.1292456430479583, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.526021957397461, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8764927387237549, "num_tokens": 336920167.0, "step": 8877 }, { "epoch": 1.1293728533265488, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6836373805999756, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8721960186958313, "num_tokens": 336953925.0, "step": 8878 }, { "epoch": 1.1295000636051393, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.513462781906128, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8754916191101074, "num_tokens": 336992407.0, "step": 8879 }, { "epoch": 1.1296272738837299, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.68727707862854, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8583019971847534, "num_tokens": 337029495.0, "step": 8880 }, { "epoch": 1.1297544841623204, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.506591558456421, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8691128492355347, "num_tokens": 337067533.0, "step": 8881 }, { "epoch": 1.129881694440911, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6958484649658203, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8584761023521423, "num_tokens": 337105112.0, "step": 8882 }, { "epoch": 1.1300089047195012, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.522400140762329, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8613018989562988, "num_tokens": 337140807.0, "step": 8883 }, { "epoch": 1.1301361149980917, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.661482095718384, "learning_rate": 1e-05, "loss": 0.3792, "mean_token_accuracy": 0.8886581659317017, "num_tokens": 337182644.0, "step": 8884 }, { "epoch": 1.1302633252766823, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.499589204788208, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8731518983840942, "num_tokens": 337226868.0, "step": 8885 }, { "epoch": 1.1303905355552728, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7043702602386475, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.860054075717926, "num_tokens": 337262118.0, "step": 8886 }, { "epoch": 1.1305177458338633, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5136988162994385, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8527992963790894, "num_tokens": 337299575.0, "step": 8887 }, { "epoch": 1.1306449561124539, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6890010833740234, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8740825653076172, "num_tokens": 337335178.0, "step": 8888 }, { "epoch": 1.1307721663910444, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.548344373703003, "learning_rate": 1e-05, "loss": 0.5062, "mean_token_accuracy": 0.8546000123023987, "num_tokens": 337368982.0, "step": 8889 }, { "epoch": 1.130899376669635, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6708171367645264, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8656980991363525, "num_tokens": 337403708.0, "step": 8890 }, { "epoch": 1.1310265869482254, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5133304595947266, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.874343991279602, "num_tokens": 337441265.0, "step": 8891 }, { "epoch": 1.131153797226816, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6764590740203857, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8730084300041199, "num_tokens": 337475600.0, "step": 8892 }, { "epoch": 1.1312810075054065, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.509153127670288, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8811972737312317, "num_tokens": 337513051.0, "step": 8893 }, { "epoch": 1.131408217783997, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.684835910797119, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8600805997848511, "num_tokens": 337552187.0, "step": 8894 }, { "epoch": 1.1315354280625876, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4952807426452637, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.8843828439712524, "num_tokens": 337586877.0, "step": 8895 }, { "epoch": 1.131662638341178, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6874752044677734, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8718615770339966, "num_tokens": 337627979.0, "step": 8896 }, { "epoch": 1.1317898486197684, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.48486590385437, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8729882836341858, "num_tokens": 337670064.0, "step": 8897 }, { "epoch": 1.131917058898359, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.714129686355591, "learning_rate": 1e-05, "loss": 0.5343, "mean_token_accuracy": 0.8447967767715454, "num_tokens": 337709670.0, "step": 8898 }, { "epoch": 1.1320442691769494, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.503183603286743, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8683763146400452, "num_tokens": 337747507.0, "step": 8899 }, { "epoch": 1.13217147945554, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7071163654327393, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8675096035003662, "num_tokens": 337789542.0, "step": 8900 }, { "epoch": 1.1322986897341305, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.474912643432617, "learning_rate": 1e-05, "loss": 0.3829, "mean_token_accuracy": 0.8841427564620972, "num_tokens": 337832100.0, "step": 8901 }, { "epoch": 1.132425900012721, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.738548994064331, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8724603652954102, "num_tokens": 337869395.0, "step": 8902 }, { "epoch": 1.1325531102913116, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4979593753814697, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8563293814659119, "num_tokens": 337909644.0, "step": 8903 }, { "epoch": 1.132680320569902, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7422993183135986, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8571925163269043, "num_tokens": 337947255.0, "step": 8904 }, { "epoch": 1.1328075308484926, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.497697353363037, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8838611841201782, "num_tokens": 337986377.0, "step": 8905 }, { "epoch": 1.1329347411270831, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7296347618103027, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8778523802757263, "num_tokens": 338025234.0, "step": 8906 }, { "epoch": 1.1330619514056737, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.495948553085327, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8590723872184753, "num_tokens": 338069070.0, "step": 8907 }, { "epoch": 1.133189161684264, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7323904037475586, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8747509717941284, "num_tokens": 338111529.0, "step": 8908 }, { "epoch": 1.1333163719628545, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5031988620758057, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.859918475151062, "num_tokens": 338155881.0, "step": 8909 }, { "epoch": 1.133443582241445, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7479302883148193, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.863174557685852, "num_tokens": 338192570.0, "step": 8910 }, { "epoch": 1.1335707925200356, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5255074501037598, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8669394254684448, "num_tokens": 338225088.0, "step": 8911 }, { "epoch": 1.133698002798626, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.735413074493408, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8579420447349548, "num_tokens": 338267301.0, "step": 8912 }, { "epoch": 1.1338252130772166, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5099101066589355, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8765838146209717, "num_tokens": 338305130.0, "step": 8913 }, { "epoch": 1.1339524233558071, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7368083000183105, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8541406393051147, "num_tokens": 338347556.0, "step": 8914 }, { "epoch": 1.1340796336343977, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5169665813446045, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8704895973205566, "num_tokens": 338384994.0, "step": 8915 }, { "epoch": 1.1342068439129882, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7248809337615967, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8605308532714844, "num_tokens": 338427311.0, "step": 8916 }, { "epoch": 1.1343340541915787, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5120105743408203, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8786880970001221, "num_tokens": 338464091.0, "step": 8917 }, { "epoch": 1.1344612644701693, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.705220937728882, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8688917756080627, "num_tokens": 338507961.0, "step": 8918 }, { "epoch": 1.1345884747487598, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.550752878189087, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8719029426574707, "num_tokens": 338544028.0, "step": 8919 }, { "epoch": 1.1347156850273503, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.683941125869751, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8557722568511963, "num_tokens": 338587827.0, "step": 8920 }, { "epoch": 1.1348428953059406, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5192911624908447, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8728061318397522, "num_tokens": 338627729.0, "step": 8921 }, { "epoch": 1.1349701055845312, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.681495189666748, "learning_rate": 1e-05, "loss": 0.3917, "mean_token_accuracy": 0.882867693901062, "num_tokens": 338663805.0, "step": 8922 }, { "epoch": 1.1350973158631217, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5222535133361816, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8698663115501404, "num_tokens": 338702204.0, "step": 8923 }, { "epoch": 1.1352245261417122, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.660646677017212, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8747450709342957, "num_tokens": 338748769.0, "step": 8924 }, { "epoch": 1.1353517364203027, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5218987464904785, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8784809708595276, "num_tokens": 338793617.0, "step": 8925 }, { "epoch": 1.1354789466988933, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6735591888427734, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8745786547660828, "num_tokens": 338827893.0, "step": 8926 }, { "epoch": 1.1356061569774838, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5157577991485596, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8748379349708557, "num_tokens": 338866036.0, "step": 8927 }, { "epoch": 1.1357333672560743, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.685232162475586, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8641775846481323, "num_tokens": 338900595.0, "step": 8928 }, { "epoch": 1.1358605775346649, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5253348350524902, "learning_rate": 1e-05, "loss": 0.3798, "mean_token_accuracy": 0.8847600221633911, "num_tokens": 338930268.0, "step": 8929 }, { "epoch": 1.1359877878132554, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6624069213867188, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8708314895629883, "num_tokens": 338974633.0, "step": 8930 }, { "epoch": 1.136114998091846, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5240285396575928, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8597159385681152, "num_tokens": 339008884.0, "step": 8931 }, { "epoch": 1.1362422083704362, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.668661594390869, "learning_rate": 1e-05, "loss": 0.5113, "mean_token_accuracy": 0.8456854224205017, "num_tokens": 339043923.0, "step": 8932 }, { "epoch": 1.1363694186490267, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5107576847076416, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8751535415649414, "num_tokens": 339080545.0, "step": 8933 }, { "epoch": 1.1364966289276173, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6715939044952393, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8705704808235168, "num_tokens": 339114243.0, "step": 8934 }, { "epoch": 1.1366238392062078, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.515756130218506, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.877790629863739, "num_tokens": 339147609.0, "step": 8935 }, { "epoch": 1.1367510494847983, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6598966121673584, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8784184455871582, "num_tokens": 339186295.0, "step": 8936 }, { "epoch": 1.1368782597633889, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00020885467529296875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5329830646514893, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8754032850265503, "num_tokens": 339217572.0, "step": 8937 }, { "epoch": 1.1370054700419794, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.650965929031372, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8739467859268188, "num_tokens": 339257305.0, "step": 8938 }, { "epoch": 1.13713268032057, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5176780223846436, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8646884560585022, "num_tokens": 339295117.0, "step": 8939 }, { "epoch": 1.1372598905991604, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.654510736465454, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8718972206115723, "num_tokens": 339330543.0, "step": 8940 }, { "epoch": 1.137387100877751, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.498525857925415, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8712850213050842, "num_tokens": 339375040.0, "step": 8941 }, { "epoch": 1.1375143111563415, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.664504051208496, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8653985261917114, "num_tokens": 339410862.0, "step": 8942 }, { "epoch": 1.137641521434932, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5119664669036865, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8761612176895142, "num_tokens": 339441929.0, "step": 8943 }, { "epoch": 1.1377687317135226, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.655074119567871, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8728190660476685, "num_tokens": 339486156.0, "step": 8944 }, { "epoch": 1.137895941992113, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.494264602661133, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8581405878067017, "num_tokens": 339531676.0, "step": 8945 }, { "epoch": 1.1380231522707034, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6729464530944824, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.860985279083252, "num_tokens": 339571003.0, "step": 8946 }, { "epoch": 1.138150362549294, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.505352735519409, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8662270307540894, "num_tokens": 339604389.0, "step": 8947 }, { "epoch": 1.1382775728278844, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.66853666305542, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.857580304145813, "num_tokens": 339649267.0, "step": 8948 }, { "epoch": 1.138404783106475, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.512956142425537, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8612937331199646, "num_tokens": 339685564.0, "step": 8949 }, { "epoch": 1.1385319933850655, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6779303550720215, "learning_rate": 1e-05, "loss": 0.3888, "mean_token_accuracy": 0.8843940496444702, "num_tokens": 339716312.0, "step": 8950 }, { "epoch": 1.138659203663656, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4920055866241455, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8720942735671997, "num_tokens": 339760666.0, "step": 8951 }, { "epoch": 1.1387864139422466, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6818878650665283, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8789492845535278, "num_tokens": 339799432.0, "step": 8952 }, { "epoch": 1.138913624220837, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4936039447784424, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8652558326721191, "num_tokens": 339839787.0, "step": 8953 }, { "epoch": 1.1390408344994276, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7021121978759766, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8725244998931885, "num_tokens": 339876193.0, "step": 8954 }, { "epoch": 1.1391680447780181, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4878289699554443, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8701662421226501, "num_tokens": 339919791.0, "step": 8955 }, { "epoch": 1.1392952550566087, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.738187313079834, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8609634637832642, "num_tokens": 339954432.0, "step": 8956 }, { "epoch": 1.139422465335199, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4734349250793457, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8744206428527832, "num_tokens": 339996327.0, "step": 8957 }, { "epoch": 1.1395496756137895, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7492618560791016, "learning_rate": 1e-05, "loss": 0.3602, "mean_token_accuracy": 0.8991958498954773, "num_tokens": 340031604.0, "step": 8958 }, { "epoch": 1.13967688589238, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4783120155334473, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8710399270057678, "num_tokens": 340072433.0, "step": 8959 }, { "epoch": 1.1398040961709706, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7573795318603516, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8701146841049194, "num_tokens": 340114306.0, "step": 8960 }, { "epoch": 1.139931306449561, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.504408836364746, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8489168286323547, "num_tokens": 340152826.0, "step": 8961 }, { "epoch": 1.1400585167281516, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.765949010848999, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8635413646697998, "num_tokens": 340191110.0, "step": 8962 }, { "epoch": 1.1401857270067421, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.502484083175659, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8694437146186829, "num_tokens": 340230606.0, "step": 8963 }, { "epoch": 1.1403129372853327, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7578282356262207, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8617761135101318, "num_tokens": 340274831.0, "step": 8964 }, { "epoch": 1.1404401475639232, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5228097438812256, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8705962300300598, "num_tokens": 340308245.0, "step": 8965 }, { "epoch": 1.1405673578425137, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.750678062438965, "learning_rate": 1e-05, "loss": 0.3848, "mean_token_accuracy": 0.8909093141555786, "num_tokens": 340343091.0, "step": 8966 }, { "epoch": 1.1406945681211043, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5195651054382324, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8743587732315063, "num_tokens": 340382968.0, "step": 8967 }, { "epoch": 1.1408217783996948, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.72983455657959, "learning_rate": 1e-05, "loss": 0.4098, "mean_token_accuracy": 0.8783316612243652, "num_tokens": 340421740.0, "step": 8968 }, { "epoch": 1.1409489886782853, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.523542881011963, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8679775595664978, "num_tokens": 340460156.0, "step": 8969 }, { "epoch": 1.1410761989568756, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7329859733581543, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8712344169616699, "num_tokens": 340496106.0, "step": 8970 }, { "epoch": 1.1412034092354661, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.528961181640625, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8729625940322876, "num_tokens": 340530338.0, "step": 8971 }, { "epoch": 1.1413306195140567, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7226452827453613, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8724756240844727, "num_tokens": 340565590.0, "step": 8972 }, { "epoch": 1.1414578297926472, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5070626735687256, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.874109148979187, "num_tokens": 340609651.0, "step": 8973 }, { "epoch": 1.1415850400712377, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.722778081893921, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8707084059715271, "num_tokens": 340650119.0, "step": 8974 }, { "epoch": 1.1417122503498283, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5214877128601074, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8714436292648315, "num_tokens": 340686734.0, "step": 8975 }, { "epoch": 1.1418394606284188, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.71797513961792, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8785070180892944, "num_tokens": 340722405.0, "step": 8976 }, { "epoch": 1.1419666709070093, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.504725933074951, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8806700110435486, "num_tokens": 340763402.0, "step": 8977 }, { "epoch": 1.1420938811855998, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7136385440826416, "learning_rate": 1e-05, "loss": 0.36, "mean_token_accuracy": 0.8975769281387329, "num_tokens": 340799410.0, "step": 8978 }, { "epoch": 1.1422210914641904, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.522251844406128, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8616648316383362, "num_tokens": 340832970.0, "step": 8979 }, { "epoch": 1.142348301742781, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.702939510345459, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8784282803535461, "num_tokens": 340872500.0, "step": 8980 }, { "epoch": 1.1424755120213712, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5052099227905273, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8662832975387573, "num_tokens": 340913784.0, "step": 8981 }, { "epoch": 1.1426027222999617, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.710021734237671, "learning_rate": 1e-05, "loss": 0.3919, "mean_token_accuracy": 0.8821709156036377, "num_tokens": 340945660.0, "step": 8982 }, { "epoch": 1.1427299325785523, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.521052360534668, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.869354248046875, "num_tokens": 340979528.0, "step": 8983 }, { "epoch": 1.1428571428571428, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7017440795898438, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.868821382522583, "num_tokens": 341016256.0, "step": 8984 }, { "epoch": 1.1429843531357333, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5162858963012695, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8609980344772339, "num_tokens": 341056339.0, "step": 8985 }, { "epoch": 1.1431115634143239, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6992416381835938, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8779948949813843, "num_tokens": 341093060.0, "step": 8986 }, { "epoch": 1.1432387736929144, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5251100063323975, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8638143539428711, "num_tokens": 341128634.0, "step": 8987 }, { "epoch": 1.143365983971505, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.7069761753082275, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8552625775337219, "num_tokens": 341160649.0, "step": 8988 }, { "epoch": 1.1434931942500954, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5408689975738525, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8709923028945923, "num_tokens": 341196660.0, "step": 8989 }, { "epoch": 1.143620404528686, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6785173416137695, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.858933687210083, "num_tokens": 341236704.0, "step": 8990 }, { "epoch": 1.1437476148072765, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.531282901763916, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8678750991821289, "num_tokens": 341272874.0, "step": 8991 }, { "epoch": 1.143874825085867, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6745121479034424, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8644469380378723, "num_tokens": 341310209.0, "step": 8992 }, { "epoch": 1.1440020353644575, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00021076202392578125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5324928760528564, "learning_rate": 1e-05, "loss": 0.3914, "mean_token_accuracy": 0.88270103931427, "num_tokens": 341341717.0, "step": 8993 }, { "epoch": 1.144129245643048, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.666151285171509, "learning_rate": 1e-05, "loss": 0.3778, "mean_token_accuracy": 0.8875529766082764, "num_tokens": 341374298.0, "step": 8994 }, { "epoch": 1.1442564559216384, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5425755977630615, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8698291778564453, "num_tokens": 341416214.0, "step": 8995 }, { "epoch": 1.144383666200229, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.648287773132324, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.87236487865448, "num_tokens": 341457419.0, "step": 8996 }, { "epoch": 1.1445108764788194, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5252785682678223, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.878242552280426, "num_tokens": 341493251.0, "step": 8997 }, { "epoch": 1.14463808675741, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6621856689453125, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8541690111160278, "num_tokens": 341531501.0, "step": 8998 }, { "epoch": 1.1447652970360005, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5003244876861572, "learning_rate": 1e-05, "loss": 0.378, "mean_token_accuracy": 0.8865196108818054, "num_tokens": 341572114.0, "step": 8999 }, { "epoch": 1.144892507314591, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6593518257141113, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8655095100402832, "num_tokens": 341617446.0, "step": 9000 }, { "epoch": 1.1450197175931816, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.512277603149414, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8660674095153809, "num_tokens": 341653636.0, "step": 9001 }, { "epoch": 1.145146927871772, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.691873788833618, "learning_rate": 1e-05, "loss": 0.5614, "mean_token_accuracy": 0.8342500925064087, "num_tokens": 341692174.0, "step": 9002 }, { "epoch": 1.1452741381503626, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4938271045684814, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8640291094779968, "num_tokens": 341737215.0, "step": 9003 }, { "epoch": 1.1454013484289531, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6895978450775146, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.866828441619873, "num_tokens": 341778571.0, "step": 9004 }, { "epoch": 1.1455285587075437, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5116684436798096, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8576638698577881, "num_tokens": 341814725.0, "step": 9005 }, { "epoch": 1.145655768986134, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.696997880935669, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8833659291267395, "num_tokens": 341854982.0, "step": 9006 }, { "epoch": 1.1457829792647245, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.504586935043335, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8788968920707703, "num_tokens": 341888952.0, "step": 9007 }, { "epoch": 1.145910189543315, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6890416145324707, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8780813217163086, "num_tokens": 341927985.0, "step": 9008 }, { "epoch": 1.1460373998219056, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.510450839996338, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8716152906417847, "num_tokens": 341959813.0, "step": 9009 }, { "epoch": 1.146164610100496, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6940758228302, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.876447856426239, "num_tokens": 341997231.0, "step": 9010 }, { "epoch": 1.1462918203790866, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.503328323364258, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8830431699752808, "num_tokens": 342029277.0, "step": 9011 }, { "epoch": 1.1464190306576771, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6885902881622314, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8806563019752502, "num_tokens": 342069297.0, "step": 9012 }, { "epoch": 1.1465462409362677, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4896962642669678, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8682745099067688, "num_tokens": 342111525.0, "step": 9013 }, { "epoch": 1.1466734512148582, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.70825457572937, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.871092677116394, "num_tokens": 342148445.0, "step": 9014 }, { "epoch": 1.1468006614934487, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.494309902191162, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8710229396820068, "num_tokens": 342186855.0, "step": 9015 }, { "epoch": 1.1469278717720393, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.721719264984131, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8702673316001892, "num_tokens": 342224619.0, "step": 9016 }, { "epoch": 1.1470550820506298, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4903619289398193, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8789067268371582, "num_tokens": 342261775.0, "step": 9017 }, { "epoch": 1.1471822923292203, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7329864501953125, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8761670589447021, "num_tokens": 342296355.0, "step": 9018 }, { "epoch": 1.1473095026078106, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4898111820220947, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8523856401443481, "num_tokens": 342337363.0, "step": 9019 }, { "epoch": 1.1474367128864011, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7423501014709473, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8712986707687378, "num_tokens": 342376462.0, "step": 9020 }, { "epoch": 1.1475639231649917, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4893736839294434, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.873771071434021, "num_tokens": 342417341.0, "step": 9021 }, { "epoch": 1.1476911334435822, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.755789279937744, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.8811564445495605, "num_tokens": 342455489.0, "step": 9022 }, { "epoch": 1.1478183437221727, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.4952406883239746, "learning_rate": 1e-05, "loss": 0.4014, "mean_token_accuracy": 0.8780548572540283, "num_tokens": 342489594.0, "step": 9023 }, { "epoch": 1.1479455540007633, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.772625207901001, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8622922897338867, "num_tokens": 342525473.0, "step": 9024 }, { "epoch": 1.1480727642793538, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5123291015625, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8831007480621338, "num_tokens": 342557472.0, "step": 9025 }, { "epoch": 1.1481999745579443, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7642292976379395, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8762065172195435, "num_tokens": 342593386.0, "step": 9026 }, { "epoch": 1.1483271848365348, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5170738697052, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8676676750183105, "num_tokens": 342631370.0, "step": 9027 }, { "epoch": 1.1484543951151254, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7443578243255615, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8712253570556641, "num_tokens": 342675570.0, "step": 9028 }, { "epoch": 1.148581605393716, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5201501846313477, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.867613673210144, "num_tokens": 342717519.0, "step": 9029 }, { "epoch": 1.1487088156723062, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7329447269439697, "learning_rate": 1e-05, "loss": 0.3924, "mean_token_accuracy": 0.884002685546875, "num_tokens": 342754104.0, "step": 9030 }, { "epoch": 1.1488360259508967, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5220422744750977, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8600796461105347, "num_tokens": 342796059.0, "step": 9031 }, { "epoch": 1.1489632362294873, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.734041690826416, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8664504289627075, "num_tokens": 342833731.0, "step": 9032 }, { "epoch": 1.1490904465080778, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.525563955307007, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.879452109336853, "num_tokens": 342871918.0, "step": 9033 }, { "epoch": 1.1492176567866683, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7360150814056396, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8570190668106079, "num_tokens": 342903129.0, "step": 9034 }, { "epoch": 1.1493448670652588, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5219943523406982, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8616422414779663, "num_tokens": 342941503.0, "step": 9035 }, { "epoch": 1.1494720773438494, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7242586612701416, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8663939833641052, "num_tokens": 342977766.0, "step": 9036 }, { "epoch": 1.14959928762244, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5019638538360596, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8734644651412964, "num_tokens": 343019865.0, "step": 9037 }, { "epoch": 1.1497264979010304, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7392096519470215, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8609297275543213, "num_tokens": 343056500.0, "step": 9038 }, { "epoch": 1.149853708179621, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5141537189483643, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8588489294052124, "num_tokens": 343098404.0, "step": 9039 }, { "epoch": 1.1499809184582115, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.723034620285034, "learning_rate": 1e-05, "loss": 0.3907, "mean_token_accuracy": 0.8850504159927368, "num_tokens": 343137360.0, "step": 9040 }, { "epoch": 1.150108128736802, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5139975547790527, "learning_rate": 1e-05, "loss": 0.4015, "mean_token_accuracy": 0.8809557557106018, "num_tokens": 343176038.0, "step": 9041 }, { "epoch": 1.1502353390153925, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.716764450073242, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8739611506462097, "num_tokens": 343217888.0, "step": 9042 }, { "epoch": 1.150362549293983, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5180492401123047, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8746157884597778, "num_tokens": 343257217.0, "step": 9043 }, { "epoch": 1.1504897595725734, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.721013307571411, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.8771897554397583, "num_tokens": 343290607.0, "step": 9044 }, { "epoch": 1.150616969851164, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.511559009552002, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8746331334114075, "num_tokens": 343329608.0, "step": 9045 }, { "epoch": 1.1507441801297544, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7189018726348877, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8828893899917603, "num_tokens": 343365459.0, "step": 9046 }, { "epoch": 1.150871390408345, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5098984241485596, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8765676021575928, "num_tokens": 343403909.0, "step": 9047 }, { "epoch": 1.1509986006869355, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7144994735717773, "learning_rate": 1e-05, "loss": 0.3713, "mean_token_accuracy": 0.8924964070320129, "num_tokens": 343438506.0, "step": 9048 }, { "epoch": 1.151125810965526, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5191092491149902, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8675335049629211, "num_tokens": 343475059.0, "step": 9049 }, { "epoch": 1.1512530212441165, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7146947383880615, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8745424151420593, "num_tokens": 343511286.0, "step": 9050 }, { "epoch": 1.151380231522707, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.510141134262085, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.8831558227539062, "num_tokens": 343548092.0, "step": 9051 }, { "epoch": 1.1515074418012976, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.704850912094116, "learning_rate": 1e-05, "loss": 0.383, "mean_token_accuracy": 0.8868419528007507, "num_tokens": 343586487.0, "step": 9052 }, { "epoch": 1.1516346520798881, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4973509311676025, "learning_rate": 1e-05, "loss": 0.3968, "mean_token_accuracy": 0.884404182434082, "num_tokens": 343627747.0, "step": 9053 }, { "epoch": 1.1517618623584787, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.711215019226074, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8581656217575073, "num_tokens": 343672725.0, "step": 9054 }, { "epoch": 1.151889072637069, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5024421215057373, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8686434030532837, "num_tokens": 343714426.0, "step": 9055 }, { "epoch": 1.1520162829156595, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7179181575775146, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.865829586982727, "num_tokens": 343754613.0, "step": 9056 }, { "epoch": 1.15214349319425, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.505596876144409, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8730416893959045, "num_tokens": 343796187.0, "step": 9057 }, { "epoch": 1.1522707034728406, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.726227045059204, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8655163049697876, "num_tokens": 343833323.0, "step": 9058 }, { "epoch": 1.152397913751431, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.517152786254883, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8619713187217712, "num_tokens": 343872642.0, "step": 9059 }, { "epoch": 1.1525251240300216, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7197225093841553, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8690147995948792, "num_tokens": 343909104.0, "step": 9060 }, { "epoch": 1.1526523343086121, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.518115282058716, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8658258318901062, "num_tokens": 343947102.0, "step": 9061 }, { "epoch": 1.1527795445872027, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.712507963180542, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8787480592727661, "num_tokens": 343985887.0, "step": 9062 }, { "epoch": 1.1529067548657932, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5261659622192383, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8637031316757202, "num_tokens": 344019427.0, "step": 9063 }, { "epoch": 1.1530339651443837, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6964046955108643, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8641119003295898, "num_tokens": 344061706.0, "step": 9064 }, { "epoch": 1.1531611754229742, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.507021188735962, "learning_rate": 1e-05, "loss": 0.3864, "mean_token_accuracy": 0.8876088261604309, "num_tokens": 344100912.0, "step": 9065 }, { "epoch": 1.1532883857015648, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.716766119003296, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8661262392997742, "num_tokens": 344135645.0, "step": 9066 }, { "epoch": 1.1534155959801553, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.505833387374878, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8717808723449707, "num_tokens": 344176467.0, "step": 9067 }, { "epoch": 1.1535428062587456, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.719817638397217, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8659898638725281, "num_tokens": 344213498.0, "step": 9068 }, { "epoch": 1.1536700165373361, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5017926692962646, "learning_rate": 1e-05, "loss": 0.3867, "mean_token_accuracy": 0.8869664669036865, "num_tokens": 344250105.0, "step": 9069 }, { "epoch": 1.1537972268159267, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7347371578216553, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8694489002227783, "num_tokens": 344284017.0, "step": 9070 }, { "epoch": 1.1539244370945172, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.504136323928833, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.875154972076416, "num_tokens": 344317729.0, "step": 9071 }, { "epoch": 1.1540516473731077, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.730905055999756, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8545824289321899, "num_tokens": 344356032.0, "step": 9072 }, { "epoch": 1.1541788576516983, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5327446460723877, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.862724244594574, "num_tokens": 344385694.0, "step": 9073 }, { "epoch": 1.1543060679302888, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7126197814941406, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8674920797348022, "num_tokens": 344420435.0, "step": 9074 }, { "epoch": 1.1544332782088793, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.50838303565979, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8799050450325012, "num_tokens": 344461065.0, "step": 9075 }, { "epoch": 1.1545604884874698, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7129528522491455, "learning_rate": 1e-05, "loss": 0.4013, "mean_token_accuracy": 0.8783266544342041, "num_tokens": 344495419.0, "step": 9076 }, { "epoch": 1.1546876987660604, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5149364471435547, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.859443724155426, "num_tokens": 344536304.0, "step": 9077 }, { "epoch": 1.154814909044651, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.705867052078247, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8728698492050171, "num_tokens": 344577118.0, "step": 9078 }, { "epoch": 1.1549421193232412, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5362095832824707, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8615438938140869, "num_tokens": 344609919.0, "step": 9079 }, { "epoch": 1.1550693296018317, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.686541795730591, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8725128173828125, "num_tokens": 344651543.0, "step": 9080 }, { "epoch": 1.1551965398804223, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5148892402648926, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8756751418113708, "num_tokens": 344691477.0, "step": 9081 }, { "epoch": 1.1553237501590128, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.680605411529541, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8704757690429688, "num_tokens": 344731533.0, "step": 9082 }, { "epoch": 1.1554509604376033, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.505873441696167, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8831706047058105, "num_tokens": 344771628.0, "step": 9083 }, { "epoch": 1.1555781707161938, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6873326301574707, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8618261218070984, "num_tokens": 344812079.0, "step": 9084 }, { "epoch": 1.1557053809947844, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4990434646606445, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8767162561416626, "num_tokens": 344856395.0, "step": 9085 }, { "epoch": 1.155832591273375, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6932692527770996, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8792786598205566, "num_tokens": 344894578.0, "step": 9086 }, { "epoch": 1.1559598015519654, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5069117546081543, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8570146560668945, "num_tokens": 344937896.0, "step": 9087 }, { "epoch": 1.156087011830556, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.700880527496338, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8853591680526733, "num_tokens": 344972852.0, "step": 9088 }, { "epoch": 1.1562142221091465, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.505880117416382, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8664087057113647, "num_tokens": 345010322.0, "step": 9089 }, { "epoch": 1.156341432387737, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7027509212493896, "learning_rate": 1e-05, "loss": 0.3798, "mean_token_accuracy": 0.8886082768440247, "num_tokens": 345046268.0, "step": 9090 }, { "epoch": 1.1564686426663275, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5113437175750732, "learning_rate": 1e-05, "loss": 0.3787, "mean_token_accuracy": 0.8863773345947266, "num_tokens": 345078601.0, "step": 9091 }, { "epoch": 1.156595852944918, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.694719076156616, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8714371919631958, "num_tokens": 345121863.0, "step": 9092 }, { "epoch": 1.1567230632235084, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.515590190887451, "learning_rate": 1e-05, "loss": 0.3609, "mean_token_accuracy": 0.893040120601654, "num_tokens": 345153884.0, "step": 9093 }, { "epoch": 1.156850273502099, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6968061923980713, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8785752654075623, "num_tokens": 345188793.0, "step": 9094 }, { "epoch": 1.1569774837806894, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5142738819122314, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8801277875900269, "num_tokens": 345222537.0, "step": 9095 }, { "epoch": 1.15710469405928, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.689387798309326, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8689115643501282, "num_tokens": 345262006.0, "step": 9096 }, { "epoch": 1.1572319043378705, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5301146507263184, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8580062389373779, "num_tokens": 345295388.0, "step": 9097 }, { "epoch": 1.157359114616461, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.673985481262207, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.8844243288040161, "num_tokens": 345331144.0, "step": 9098 }, { "epoch": 1.1574863248950515, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5203046798706055, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8527489304542542, "num_tokens": 345373009.0, "step": 9099 }, { "epoch": 1.157613535173642, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.667617082595825, "learning_rate": 1e-05, "loss": 0.3975, "mean_token_accuracy": 0.8843450546264648, "num_tokens": 345411931.0, "step": 9100 }, { "epoch": 1.1577407454522326, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5165798664093018, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8706481456756592, "num_tokens": 345448818.0, "step": 9101 }, { "epoch": 1.1578679557308231, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.679722547531128, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8802392482757568, "num_tokens": 345484867.0, "step": 9102 }, { "epoch": 1.1579951660094137, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.5089492797851562, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8742059469223022, "num_tokens": 345525949.0, "step": 9103 }, { "epoch": 1.158122376288004, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6738619804382324, "learning_rate": 1e-05, "loss": 0.3872, "mean_token_accuracy": 0.8853132724761963, "num_tokens": 345563027.0, "step": 9104 }, { "epoch": 1.1582495865665945, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5033793449401855, "learning_rate": 1e-05, "loss": 0.4013, "mean_token_accuracy": 0.8830119371414185, "num_tokens": 345605920.0, "step": 9105 }, { "epoch": 1.158376796845185, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.689764976501465, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8654983043670654, "num_tokens": 345645843.0, "step": 9106 }, { "epoch": 1.1585040071237755, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.505807399749756, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8723794221878052, "num_tokens": 345684207.0, "step": 9107 }, { "epoch": 1.158631217402366, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7031731605529785, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8634676337242126, "num_tokens": 345721628.0, "step": 9108 }, { "epoch": 1.1587584276809566, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.0002117156982421875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4992523193359375, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8777974843978882, "num_tokens": 345761864.0, "step": 9109 }, { "epoch": 1.1588856379595471, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7036614418029785, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8735583424568176, "num_tokens": 345802478.0, "step": 9110 }, { "epoch": 1.1590128482381377, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.485517978668213, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8756536841392517, "num_tokens": 345846598.0, "step": 9111 }, { "epoch": 1.1591400585167282, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.735802412033081, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8594187498092651, "num_tokens": 345885050.0, "step": 9112 }, { "epoch": 1.1592672687953187, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5032455921173096, "learning_rate": 1e-05, "loss": 0.5177, "mean_token_accuracy": 0.8419356346130371, "num_tokens": 345927460.0, "step": 9113 }, { "epoch": 1.1593944790739092, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.74651837348938, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.869687557220459, "num_tokens": 345960855.0, "step": 9114 }, { "epoch": 1.1595216893524998, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.493945837020874, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8775145411491394, "num_tokens": 345997086.0, "step": 9115 }, { "epoch": 1.1596488996310903, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7458877563476562, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8779623508453369, "num_tokens": 346034317.0, "step": 9116 }, { "epoch": 1.1597761099096806, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.496238946914673, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8666160106658936, "num_tokens": 346072626.0, "step": 9117 }, { "epoch": 1.1599033201882711, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7605936527252197, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8664559125900269, "num_tokens": 346109792.0, "step": 9118 }, { "epoch": 1.1600305304668617, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.518310546875, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8805835247039795, "num_tokens": 346147696.0, "step": 9119 }, { "epoch": 1.1601577407454522, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.762721061706543, "learning_rate": 1e-05, "loss": 0.3917, "mean_token_accuracy": 0.8831188678741455, "num_tokens": 346183944.0, "step": 9120 }, { "epoch": 1.1602849510240427, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.506897211074829, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8795193433761597, "num_tokens": 346223352.0, "step": 9121 }, { "epoch": 1.1604121613026332, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7757668495178223, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8696548938751221, "num_tokens": 346255008.0, "step": 9122 }, { "epoch": 1.1605393715812238, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5528690814971924, "learning_rate": 1e-05, "loss": 0.5299, "mean_token_accuracy": 0.8457689881324768, "num_tokens": 346285188.0, "step": 9123 }, { "epoch": 1.1606665818598143, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.740797996520996, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8775595426559448, "num_tokens": 346318691.0, "step": 9124 }, { "epoch": 1.1607937921384048, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5223188400268555, "learning_rate": 1e-05, "loss": 0.3878, "mean_token_accuracy": 0.8858175277709961, "num_tokens": 346353800.0, "step": 9125 }, { "epoch": 1.1609210024169954, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7336785793304443, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8562408685684204, "num_tokens": 346396616.0, "step": 9126 }, { "epoch": 1.161048212695586, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00021266937255859375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.524336814880371, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8725680112838745, "num_tokens": 346438244.0, "step": 9127 }, { "epoch": 1.1611754229741762, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.743358612060547, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8711239099502563, "num_tokens": 346475081.0, "step": 9128 }, { "epoch": 1.1613026332527667, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.514118194580078, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8781498074531555, "num_tokens": 346513435.0, "step": 9129 }, { "epoch": 1.1614298435313573, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.731301784515381, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8660512566566467, "num_tokens": 346553216.0, "step": 9130 }, { "epoch": 1.1615570538099478, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5219473838806152, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8679726123809814, "num_tokens": 346589271.0, "step": 9131 }, { "epoch": 1.1616842640885383, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7197318077087402, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8725738525390625, "num_tokens": 346631806.0, "step": 9132 }, { "epoch": 1.1618114743671288, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.516589403152466, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.884190022945404, "num_tokens": 346670805.0, "step": 9133 }, { "epoch": 1.1619386846457194, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.714245557785034, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8835068345069885, "num_tokens": 346710506.0, "step": 9134 }, { "epoch": 1.16206589492431, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.528217315673828, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8627854585647583, "num_tokens": 346753410.0, "step": 9135 }, { "epoch": 1.1621931052029004, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7067995071411133, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8592258095741272, "num_tokens": 346795104.0, "step": 9136 }, { "epoch": 1.162320315481491, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5313262939453125, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8562776446342468, "num_tokens": 346836888.0, "step": 9137 }, { "epoch": 1.1624475257600815, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.693258047103882, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8688735365867615, "num_tokens": 346874449.0, "step": 9138 }, { "epoch": 1.162574736038672, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.514596939086914, "learning_rate": 1e-05, "loss": 0.3527, "mean_token_accuracy": 0.8946774005889893, "num_tokens": 346906706.0, "step": 9139 }, { "epoch": 1.1627019463172625, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.689272403717041, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8717896938323975, "num_tokens": 346947250.0, "step": 9140 }, { "epoch": 1.162829156595853, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.524794578552246, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8689373731613159, "num_tokens": 346984159.0, "step": 9141 }, { "epoch": 1.1629563668744434, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6943321228027344, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8519569635391235, "num_tokens": 347026449.0, "step": 9142 }, { "epoch": 1.163083577153034, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5201892852783203, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8850786685943604, "num_tokens": 347057903.0, "step": 9143 }, { "epoch": 1.1632107874316244, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6808927059173584, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8717963695526123, "num_tokens": 347095805.0, "step": 9144 }, { "epoch": 1.163337997710215, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5195679664611816, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8725993633270264, "num_tokens": 347133357.0, "step": 9145 }, { "epoch": 1.1634652079888055, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6877007484436035, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.860928475856781, "num_tokens": 347169679.0, "step": 9146 }, { "epoch": 1.163592418267396, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5646934509277344, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8650954961776733, "num_tokens": 347207990.0, "step": 9147 }, { "epoch": 1.1637196285459865, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.650543689727783, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8823736906051636, "num_tokens": 347244427.0, "step": 9148 }, { "epoch": 1.163846838824577, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.539588212966919, "learning_rate": 1e-05, "loss": 0.3671, "mean_token_accuracy": 0.8929979205131531, "num_tokens": 347276303.0, "step": 9149 }, { "epoch": 1.1639740491031676, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.630711793899536, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8778156042098999, "num_tokens": 347318682.0, "step": 9150 }, { "epoch": 1.1641012593817581, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5314979553222656, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8624879121780396, "num_tokens": 347359355.0, "step": 9151 }, { "epoch": 1.1642284696603487, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6439926624298096, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8682059645652771, "num_tokens": 347397617.0, "step": 9152 }, { "epoch": 1.164355679938939, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.542393684387207, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8508589267730713, "num_tokens": 347433067.0, "step": 9153 }, { "epoch": 1.1644828902175295, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6396496295928955, "learning_rate": 1e-05, "loss": 0.3859, "mean_token_accuracy": 0.8847848176956177, "num_tokens": 347467091.0, "step": 9154 }, { "epoch": 1.16461010049612, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5312087535858154, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8789496421813965, "num_tokens": 347503921.0, "step": 9155 }, { "epoch": 1.1647373107747105, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.632354497909546, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8738791346549988, "num_tokens": 347547061.0, "step": 9156 }, { "epoch": 1.164864521053301, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5379579067230225, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8655844926834106, "num_tokens": 347583845.0, "step": 9157 }, { "epoch": 1.1649917313318916, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.635953664779663, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8708776235580444, "num_tokens": 347624146.0, "step": 9158 }, { "epoch": 1.1651189416104821, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5419552326202393, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8703370094299316, "num_tokens": 347666171.0, "step": 9159 }, { "epoch": 1.1652461518890727, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.648493766784668, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8687583208084106, "num_tokens": 347706960.0, "step": 9160 }, { "epoch": 1.1653733621676632, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5267772674560547, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8687180280685425, "num_tokens": 347744830.0, "step": 9161 }, { "epoch": 1.1655005724462537, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6269137859344482, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.876997172832489, "num_tokens": 347791033.0, "step": 9162 }, { "epoch": 1.1656277827248442, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.52260684967041, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8682418465614319, "num_tokens": 347832579.0, "step": 9163 }, { "epoch": 1.1657549930034348, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6456704139709473, "learning_rate": 1e-05, "loss": 0.3725, "mean_token_accuracy": 0.8894162178039551, "num_tokens": 347863404.0, "step": 9164 }, { "epoch": 1.1658822032820253, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5187530517578125, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8663722276687622, "num_tokens": 347900882.0, "step": 9165 }, { "epoch": 1.1660094135606156, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6457390785217285, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8647927045822144, "num_tokens": 347943105.0, "step": 9166 }, { "epoch": 1.1661366238392061, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.545292377471924, "learning_rate": 1e-05, "loss": 0.5324, "mean_token_accuracy": 0.8365705013275146, "num_tokens": 347978783.0, "step": 9167 }, { "epoch": 1.1662638341177967, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.634758234024048, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8802672028541565, "num_tokens": 348016169.0, "step": 9168 }, { "epoch": 1.1663910443963872, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.526869773864746, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8641110062599182, "num_tokens": 348052431.0, "step": 9169 }, { "epoch": 1.1665182546749777, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6339406967163086, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8680504560470581, "num_tokens": 348091412.0, "step": 9170 }, { "epoch": 1.1666454649535682, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5271925926208496, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8668789863586426, "num_tokens": 348128370.0, "step": 9171 }, { "epoch": 1.1667726752321588, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.647010564804077, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.86463862657547, "num_tokens": 348167076.0, "step": 9172 }, { "epoch": 1.1668998855107493, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.513505220413208, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8770580291748047, "num_tokens": 348206926.0, "step": 9173 }, { "epoch": 1.1670270957893398, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.662083387374878, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8655760884284973, "num_tokens": 348247209.0, "step": 9174 }, { "epoch": 1.1671543060679304, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4980452060699463, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.86793053150177, "num_tokens": 348286613.0, "step": 9175 }, { "epoch": 1.1672815163465209, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6778640747070312, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8686105012893677, "num_tokens": 348322246.0, "step": 9176 }, { "epoch": 1.1674087266251112, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.49422550201416, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8777868151664734, "num_tokens": 348359435.0, "step": 9177 }, { "epoch": 1.1675359369037017, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.682935953140259, "learning_rate": 1e-05, "loss": 0.3856, "mean_token_accuracy": 0.8888965845108032, "num_tokens": 348397330.0, "step": 9178 }, { "epoch": 1.1676631471822922, "ewc_loss": 0.0576171875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.647804260253906e-05, "grad_norm": 2.497258186340332, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.859553337097168, "num_tokens": 348435413.0, "step": 9179 }, { "epoch": 1.1677903574608828, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.697963237762451, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8589733242988586, "num_tokens": 348474963.0, "step": 9180 }, { "epoch": 1.1679175677394733, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.497635841369629, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8699990510940552, "num_tokens": 348511306.0, "step": 9181 }, { "epoch": 1.1680447780180638, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7027933597564697, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8781070113182068, "num_tokens": 348552257.0, "step": 9182 }, { "epoch": 1.1681719882966544, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.496372938156128, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8676498532295227, "num_tokens": 348588143.0, "step": 9183 }, { "epoch": 1.168299198575245, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.699497938156128, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8711360692977905, "num_tokens": 348633525.0, "step": 9184 }, { "epoch": 1.1684264088538354, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.493769407272339, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8680382966995239, "num_tokens": 348676024.0, "step": 9185 }, { "epoch": 1.168553619132426, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7131240367889404, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.88002610206604, "num_tokens": 348715704.0, "step": 9186 }, { "epoch": 1.1686808294110165, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.500379800796509, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.865609884262085, "num_tokens": 348753160.0, "step": 9187 }, { "epoch": 1.168808039689607, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.715494155883789, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.873315155506134, "num_tokens": 348796423.0, "step": 9188 }, { "epoch": 1.1689352499681975, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.490676164627075, "learning_rate": 1e-05, "loss": 0.3743, "mean_token_accuracy": 0.8907250761985779, "num_tokens": 348832813.0, "step": 9189 }, { "epoch": 1.169062460246788, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7409400939941406, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8712223172187805, "num_tokens": 348868273.0, "step": 9190 }, { "epoch": 1.1691896705253784, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.483537435531616, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.8795076608657837, "num_tokens": 348908008.0, "step": 9191 }, { "epoch": 1.169316880803969, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.745328426361084, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8713676929473877, "num_tokens": 348947701.0, "step": 9192 }, { "epoch": 1.1694440910825594, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4858479499816895, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8704061508178711, "num_tokens": 348992221.0, "step": 9193 }, { "epoch": 1.16957130136115, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7658064365386963, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8801910877227783, "num_tokens": 349031982.0, "step": 9194 }, { "epoch": 1.1696985116397405, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4858813285827637, "learning_rate": 1e-05, "loss": 0.3986, "mean_token_accuracy": 0.8829900026321411, "num_tokens": 349070905.0, "step": 9195 }, { "epoch": 1.169825721918331, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000213623046875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.8221356868743896, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8713512420654297, "num_tokens": 349114736.0, "step": 9196 }, { "epoch": 1.1699529321969215, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5089287757873535, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8648750185966492, "num_tokens": 349153760.0, "step": 9197 }, { "epoch": 1.170080142475512, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7859909534454346, "learning_rate": 1e-05, "loss": 0.3906, "mean_token_accuracy": 0.8868412971496582, "num_tokens": 349192464.0, "step": 9198 }, { "epoch": 1.1702073527541026, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5145251750946045, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8703278303146362, "num_tokens": 349234444.0, "step": 9199 }, { "epoch": 1.1703345630326931, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7762744426727295, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.85835862159729, "num_tokens": 349276777.0, "step": 9200 }, { "epoch": 1.1704617733112836, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5235610008239746, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.8811548948287964, "num_tokens": 349317158.0, "step": 9201 }, { "epoch": 1.170588983589874, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7700798511505127, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8653340339660645, "num_tokens": 349357043.0, "step": 9202 }, { "epoch": 1.1707161938684645, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5382790565490723, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8546547889709473, "num_tokens": 349395843.0, "step": 9203 }, { "epoch": 1.170843404147055, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7603859901428223, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8650000095367432, "num_tokens": 349429410.0, "step": 9204 }, { "epoch": 1.1709706144256455, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.549254894256592, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.856541633605957, "num_tokens": 349466507.0, "step": 9205 }, { "epoch": 1.171097824704236, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7277660369873047, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8849523067474365, "num_tokens": 349502918.0, "step": 9206 }, { "epoch": 1.1712250349828266, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.532782554626465, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8723440170288086, "num_tokens": 349546740.0, "step": 9207 }, { "epoch": 1.1713522452614171, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.705258846282959, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8760590553283691, "num_tokens": 349590479.0, "step": 9208 }, { "epoch": 1.1714794555400077, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.549003839492798, "learning_rate": 1e-05, "loss": 0.3748, "mean_token_accuracy": 0.8881531953811646, "num_tokens": 349627726.0, "step": 9209 }, { "epoch": 1.1716066658185982, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.701859474182129, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8719395399093628, "num_tokens": 349663547.0, "step": 9210 }, { "epoch": 1.1717338760971887, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5582921504974365, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8661595582962036, "num_tokens": 349696927.0, "step": 9211 }, { "epoch": 1.1718610863757792, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6716394424438477, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8791590929031372, "num_tokens": 349736027.0, "step": 9212 }, { "epoch": 1.1719882966543698, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.530817747116089, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8661803603172302, "num_tokens": 349778225.0, "step": 9213 }, { "epoch": 1.1721155069329603, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6747019290924072, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8709831237792969, "num_tokens": 349818093.0, "step": 9214 }, { "epoch": 1.1722427172115506, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5347237586975098, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8693870902061462, "num_tokens": 349856433.0, "step": 9215 }, { "epoch": 1.1723699274901411, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.663424253463745, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8663492798805237, "num_tokens": 349898829.0, "step": 9216 }, { "epoch": 1.1724971377687317, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.53003191947937, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8742994070053101, "num_tokens": 349937059.0, "step": 9217 }, { "epoch": 1.1726243480473222, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6629109382629395, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8622828125953674, "num_tokens": 349973433.0, "step": 9218 }, { "epoch": 1.1727515583259127, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5337090492248535, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8532264828681946, "num_tokens": 350012859.0, "step": 9219 }, { "epoch": 1.1728787686045032, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.653787136077881, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8595008850097656, "num_tokens": 350055128.0, "step": 9220 }, { "epoch": 1.1730059788830938, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5396463871002197, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8572743535041809, "num_tokens": 350090058.0, "step": 9221 }, { "epoch": 1.1731331891616843, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6406240463256836, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8791454434394836, "num_tokens": 350129441.0, "step": 9222 }, { "epoch": 1.1732603994402748, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5225672721862793, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8769282102584839, "num_tokens": 350168866.0, "step": 9223 }, { "epoch": 1.1733876097188654, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.648378849029541, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8682688474655151, "num_tokens": 350207498.0, "step": 9224 }, { "epoch": 1.1735148199974559, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5180394649505615, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8598791360855103, "num_tokens": 350248729.0, "step": 9225 }, { "epoch": 1.1736420302760462, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6716256141662598, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8719854354858398, "num_tokens": 350279399.0, "step": 9226 }, { "epoch": 1.1737692405546367, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.512803316116333, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8534445762634277, "num_tokens": 350321135.0, "step": 9227 }, { "epoch": 1.1738964508332272, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6729507446289062, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8671860694885254, "num_tokens": 350359348.0, "step": 9228 }, { "epoch": 1.1740236611118178, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4966557025909424, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.875392496585846, "num_tokens": 350401140.0, "step": 9229 }, { "epoch": 1.1741508713904083, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6735665798187256, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8665041327476501, "num_tokens": 350443936.0, "step": 9230 }, { "epoch": 1.1742780816689988, "ewc_loss": 0.058349609375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4874279499053955, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.8815252780914307, "num_tokens": 350486110.0, "step": 9231 }, { "epoch": 1.1744052919475894, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7166953086853027, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8520843386650085, "num_tokens": 350521964.0, "step": 9232 }, { "epoch": 1.1745325022261799, "ewc_loss": 0.05810546875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.4920105934143066, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8846701979637146, "num_tokens": 350557578.0, "step": 9233 }, { "epoch": 1.1746597125047704, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.720775842666626, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8705703020095825, "num_tokens": 350595693.0, "step": 9234 }, { "epoch": 1.174786922783361, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4933464527130127, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.878338098526001, "num_tokens": 350633247.0, "step": 9235 }, { "epoch": 1.1749141330619515, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.731518030166626, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8803853988647461, "num_tokens": 350673048.0, "step": 9236 }, { "epoch": 1.175041343340542, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4997007846832275, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8600884675979614, "num_tokens": 350708840.0, "step": 9237 }, { "epoch": 1.1751685536191325, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7421932220458984, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8711432218551636, "num_tokens": 350748875.0, "step": 9238 }, { "epoch": 1.175295763897723, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5085713863372803, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8686097264289856, "num_tokens": 350783200.0, "step": 9239 }, { "epoch": 1.1754229741763134, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7432143688201904, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8713383078575134, "num_tokens": 350819514.0, "step": 9240 }, { "epoch": 1.175550184454904, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.4928293228149414, "learning_rate": 1e-05, "loss": 0.37, "mean_token_accuracy": 0.8919156789779663, "num_tokens": 350854261.0, "step": 9241 }, { "epoch": 1.1756773947334944, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.744760274887085, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8720581531524658, "num_tokens": 350897847.0, "step": 9242 }, { "epoch": 1.175804605012085, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.52059268951416, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8635753393173218, "num_tokens": 350931926.0, "step": 9243 }, { "epoch": 1.1759318152906755, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7321224212646484, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8825637102127075, "num_tokens": 350972352.0, "step": 9244 }, { "epoch": 1.176059025569266, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5104730129241943, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8694445490837097, "num_tokens": 351012245.0, "step": 9245 }, { "epoch": 1.1761862358478565, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.746851682662964, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8764545917510986, "num_tokens": 351043005.0, "step": 9246 }, { "epoch": 1.176313446126447, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.518325090408325, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8567707538604736, "num_tokens": 351084810.0, "step": 9247 }, { "epoch": 1.1764406564050376, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7327873706817627, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8766574859619141, "num_tokens": 351124243.0, "step": 9248 }, { "epoch": 1.1765678666836281, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.510023832321167, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.88022381067276, "num_tokens": 351163650.0, "step": 9249 }, { "epoch": 1.1766950769622184, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7482283115386963, "learning_rate": 1e-05, "loss": 0.5394, "mean_token_accuracy": 0.8460057973861694, "num_tokens": 351205912.0, "step": 9250 }, { "epoch": 1.176822287240809, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00021457672119140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.511155128479004, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.871141791343689, "num_tokens": 351250515.0, "step": 9251 }, { "epoch": 1.1769494975193995, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7522714138031006, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8683027625083923, "num_tokens": 351284222.0, "step": 9252 }, { "epoch": 1.17707670779799, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.511411666870117, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8599063158035278, "num_tokens": 351327799.0, "step": 9253 }, { "epoch": 1.1772039180765805, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.760775566101074, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8715462684631348, "num_tokens": 351362194.0, "step": 9254 }, { "epoch": 1.177331128355171, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.522071599960327, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8776986002922058, "num_tokens": 351395186.0, "step": 9255 }, { "epoch": 1.1774583386337616, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7406277656555176, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8730022311210632, "num_tokens": 351435484.0, "step": 9256 }, { "epoch": 1.1775855489123521, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.530036449432373, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8622925877571106, "num_tokens": 351474613.0, "step": 9257 }, { "epoch": 1.1777127591909426, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.738739252090454, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8779320120811462, "num_tokens": 351508612.0, "step": 9258 }, { "epoch": 1.1778399694695332, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.519777774810791, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8573987483978271, "num_tokens": 351554811.0, "step": 9259 }, { "epoch": 1.1779671797481237, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7274270057678223, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8737882375717163, "num_tokens": 351591540.0, "step": 9260 }, { "epoch": 1.1780943900267142, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.538721799850464, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8718550205230713, "num_tokens": 351630529.0, "step": 9261 }, { "epoch": 1.1782216003053048, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.720461130142212, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8596640825271606, "num_tokens": 351669412.0, "step": 9262 }, { "epoch": 1.1783488105838953, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5252578258514404, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8653117418289185, "num_tokens": 351711825.0, "step": 9263 }, { "epoch": 1.1784760208624856, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7104580402374268, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8767907023429871, "num_tokens": 351747967.0, "step": 9264 }, { "epoch": 1.1786032311410761, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5290558338165283, "learning_rate": 1e-05, "loss": 0.4058, "mean_token_accuracy": 0.8785170316696167, "num_tokens": 351786208.0, "step": 9265 }, { "epoch": 1.1787304414196667, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7040023803710938, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8687960505485535, "num_tokens": 351825711.0, "step": 9266 }, { "epoch": 1.1788576516982572, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5094833374023438, "learning_rate": 1e-05, "loss": 0.3961, "mean_token_accuracy": 0.8816077709197998, "num_tokens": 351866772.0, "step": 9267 }, { "epoch": 1.1789848619768477, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.711115837097168, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8730504512786865, "num_tokens": 351905812.0, "step": 9268 }, { "epoch": 1.1791120722554382, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5054314136505127, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8778984546661377, "num_tokens": 351950295.0, "step": 9269 }, { "epoch": 1.1792392825340288, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.725501775741577, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8588736057281494, "num_tokens": 351990980.0, "step": 9270 }, { "epoch": 1.1793664928126193, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.503066301345825, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8579306602478027, "num_tokens": 352034483.0, "step": 9271 }, { "epoch": 1.1794937030912098, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.730086088180542, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8800814151763916, "num_tokens": 352069195.0, "step": 9272 }, { "epoch": 1.1796209133698004, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5091943740844727, "learning_rate": 1e-05, "loss": 0.3804, "mean_token_accuracy": 0.8819987177848816, "num_tokens": 352103393.0, "step": 9273 }, { "epoch": 1.1797481236483909, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7329466342926025, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8786147236824036, "num_tokens": 352138032.0, "step": 9274 }, { "epoch": 1.1798753339269812, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.517953872680664, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8613497018814087, "num_tokens": 352177243.0, "step": 9275 }, { "epoch": 1.1800025442055717, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7415831089019775, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8544754981994629, "num_tokens": 352212461.0, "step": 9276 }, { "epoch": 1.1801297544841622, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.516636848449707, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8619047403335571, "num_tokens": 352250001.0, "step": 9277 }, { "epoch": 1.1802569647627528, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.734950065612793, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8571116924285889, "num_tokens": 352288893.0, "step": 9278 }, { "epoch": 1.1803841750413433, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.51488995552063, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8731092810630798, "num_tokens": 352328255.0, "step": 9279 }, { "epoch": 1.1805113853199338, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7330586910247803, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8687812089920044, "num_tokens": 352364871.0, "step": 9280 }, { "epoch": 1.1806385955985244, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5176479816436768, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.880651593208313, "num_tokens": 352398894.0, "step": 9281 }, { "epoch": 1.1807658058771149, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7223153114318848, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8666921854019165, "num_tokens": 352438327.0, "step": 9282 }, { "epoch": 1.1808930161557054, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5213818550109863, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8601651787757874, "num_tokens": 352475732.0, "step": 9283 }, { "epoch": 1.181020226434296, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7392654418945312, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8562631011009216, "num_tokens": 352509405.0, "step": 9284 }, { "epoch": 1.1811474367128865, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5238797664642334, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.862091064453125, "num_tokens": 352547172.0, "step": 9285 }, { "epoch": 1.181274646991477, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7264318466186523, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8698980212211609, "num_tokens": 352585705.0, "step": 9286 }, { "epoch": 1.1814018572700675, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5018858909606934, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8788914680480957, "num_tokens": 352629246.0, "step": 9287 }, { "epoch": 1.181529067548658, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7336623668670654, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.861359179019928, "num_tokens": 352669504.0, "step": 9288 }, { "epoch": 1.1816562778272484, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.519068956375122, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8657572865486145, "num_tokens": 352708981.0, "step": 9289 }, { "epoch": 1.1817834881058389, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.724933385848999, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8662202954292297, "num_tokens": 352751646.0, "step": 9290 }, { "epoch": 1.1819106983844294, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.517561674118042, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8702260255813599, "num_tokens": 352789959.0, "step": 9291 }, { "epoch": 1.18203790866302, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7304067611694336, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8570281863212585, "num_tokens": 352828725.0, "step": 9292 }, { "epoch": 1.1821651189416105, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.513357162475586, "learning_rate": 1e-05, "loss": 0.3758, "mean_token_accuracy": 0.8893202543258667, "num_tokens": 352866373.0, "step": 9293 }, { "epoch": 1.182292329220201, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.725123882293701, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8712426424026489, "num_tokens": 352908964.0, "step": 9294 }, { "epoch": 1.1824195394987915, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002155303955078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.519608497619629, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8792644739151001, "num_tokens": 352946292.0, "step": 9295 }, { "epoch": 1.182546749777382, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7225091457366943, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8745906352996826, "num_tokens": 352985875.0, "step": 9296 }, { "epoch": 1.1826739600559726, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5188345909118652, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8656001091003418, "num_tokens": 353032019.0, "step": 9297 }, { "epoch": 1.1828011703345631, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.734649896621704, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8687938451766968, "num_tokens": 353062624.0, "step": 9298 }, { "epoch": 1.1829283806131534, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5082666873931885, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.87436443567276, "num_tokens": 353105401.0, "step": 9299 }, { "epoch": 1.183055590891744, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7358713150024414, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.854889988899231, "num_tokens": 353143180.0, "step": 9300 }, { "epoch": 1.1831828011703345, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5317304134368896, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8709200620651245, "num_tokens": 353177800.0, "step": 9301 }, { "epoch": 1.183310011448925, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.71628737449646, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8700188994407654, "num_tokens": 353221491.0, "step": 9302 }, { "epoch": 1.1834372217275155, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5152711868286133, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8581992387771606, "num_tokens": 353263851.0, "step": 9303 }, { "epoch": 1.183564432006106, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7341485023498535, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8592960834503174, "num_tokens": 353299197.0, "step": 9304 }, { "epoch": 1.1836916422846966, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5222959518432617, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8495163917541504, "num_tokens": 353339008.0, "step": 9305 }, { "epoch": 1.1838188525632871, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7146873474121094, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8703129291534424, "num_tokens": 353378143.0, "step": 9306 }, { "epoch": 1.1839460628418776, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.516573429107666, "learning_rate": 1e-05, "loss": 0.3762, "mean_token_accuracy": 0.8891529440879822, "num_tokens": 353412659.0, "step": 9307 }, { "epoch": 1.1840732731204682, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.709916591644287, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8570383787155151, "num_tokens": 353454459.0, "step": 9308 }, { "epoch": 1.1842004833990587, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.517300844192505, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8638110756874084, "num_tokens": 353501446.0, "step": 9309 }, { "epoch": 1.1843276936776492, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.72078537940979, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8647031784057617, "num_tokens": 353538129.0, "step": 9310 }, { "epoch": 1.1844549039562398, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5083534717559814, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.879837155342102, "num_tokens": 353581488.0, "step": 9311 }, { "epoch": 1.1845821142348303, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7203283309936523, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8724334239959717, "num_tokens": 353622729.0, "step": 9312 }, { "epoch": 1.1847093245134206, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.495163917541504, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8782630562782288, "num_tokens": 353663696.0, "step": 9313 }, { "epoch": 1.1848365347920111, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7306034564971924, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.860162615776062, "num_tokens": 353705934.0, "step": 9314 }, { "epoch": 1.1849637450706016, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.513214588165283, "learning_rate": 1e-05, "loss": 0.3918, "mean_token_accuracy": 0.8867907524108887, "num_tokens": 353741535.0, "step": 9315 }, { "epoch": 1.1850909553491922, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.721283197402954, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8687376379966736, "num_tokens": 353784446.0, "step": 9316 }, { "epoch": 1.1852181656277827, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.52349853515625, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8687604665756226, "num_tokens": 353821955.0, "step": 9317 }, { "epoch": 1.1853453759063732, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.718858242034912, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8705627918243408, "num_tokens": 353859994.0, "step": 9318 }, { "epoch": 1.1854725861849638, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5136806964874268, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8762292861938477, "num_tokens": 353903301.0, "step": 9319 }, { "epoch": 1.1855997964635543, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.722459554672241, "learning_rate": 1e-05, "loss": 0.5124, "mean_token_accuracy": 0.8468328714370728, "num_tokens": 353944527.0, "step": 9320 }, { "epoch": 1.1857270067421448, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5266189575195312, "learning_rate": 1e-05, "loss": 0.4021, "mean_token_accuracy": 0.8800559639930725, "num_tokens": 353977402.0, "step": 9321 }, { "epoch": 1.1858542170207353, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.70703125, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8781754970550537, "num_tokens": 354016681.0, "step": 9322 }, { "epoch": 1.1859814272993259, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5216493606567383, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8597330451011658, "num_tokens": 354057923.0, "step": 9323 }, { "epoch": 1.1861086375779162, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6974637508392334, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.869259238243103, "num_tokens": 354097787.0, "step": 9324 }, { "epoch": 1.1862358478565067, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5374701023101807, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8769142031669617, "num_tokens": 354136694.0, "step": 9325 }, { "epoch": 1.1863630581350972, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.680520534515381, "learning_rate": 1e-05, "loss": 0.386, "mean_token_accuracy": 0.8888756036758423, "num_tokens": 354176326.0, "step": 9326 }, { "epoch": 1.1864902684136878, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5249602794647217, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8655081391334534, "num_tokens": 354214350.0, "step": 9327 }, { "epoch": 1.1866174786922783, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.682079792022705, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8692508339881897, "num_tokens": 354252910.0, "step": 9328 }, { "epoch": 1.1867446889708688, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.526102066040039, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8707173466682434, "num_tokens": 354291105.0, "step": 9329 }, { "epoch": 1.1868718992494594, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.67533016204834, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8750607371330261, "num_tokens": 354329767.0, "step": 9330 }, { "epoch": 1.1869991095280499, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.509812355041504, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8794853687286377, "num_tokens": 354369826.0, "step": 9331 }, { "epoch": 1.1871263198066404, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6905932426452637, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8819810152053833, "num_tokens": 354406088.0, "step": 9332 }, { "epoch": 1.187253530085231, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5093019008636475, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8829928040504456, "num_tokens": 354445675.0, "step": 9333 }, { "epoch": 1.1873807403638215, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7001492977142334, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8522714972496033, "num_tokens": 354482733.0, "step": 9334 }, { "epoch": 1.187507950642412, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5300166606903076, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8649746775627136, "num_tokens": 354515483.0, "step": 9335 }, { "epoch": 1.1876351609210025, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6902012825012207, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.878347635269165, "num_tokens": 354551972.0, "step": 9336 }, { "epoch": 1.187762371199593, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.000217437744140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5135555267333984, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8704896569252014, "num_tokens": 354592089.0, "step": 9337 }, { "epoch": 1.1878895814781834, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6865460872650146, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.880778431892395, "num_tokens": 354629582.0, "step": 9338 }, { "epoch": 1.1880167917567739, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.525228977203369, "learning_rate": 1e-05, "loss": 0.5129, "mean_token_accuracy": 0.84734046459198, "num_tokens": 354677114.0, "step": 9339 }, { "epoch": 1.1881440020353644, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6726434230804443, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8677302598953247, "num_tokens": 354723125.0, "step": 9340 }, { "epoch": 1.188271212313955, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5253493785858154, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8608101606369019, "num_tokens": 354761032.0, "step": 9341 }, { "epoch": 1.1883984225925455, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6751022338867188, "learning_rate": 1e-05, "loss": 0.4156, "mean_token_accuracy": 0.8797057867050171, "num_tokens": 354798522.0, "step": 9342 }, { "epoch": 1.188525632871136, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.533994436264038, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8719485998153687, "num_tokens": 354833752.0, "step": 9343 }, { "epoch": 1.1886528431497265, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.678835868835449, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8746579885482788, "num_tokens": 354869901.0, "step": 9344 }, { "epoch": 1.188780053428317, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5301709175109863, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8707370758056641, "num_tokens": 354902487.0, "step": 9345 }, { "epoch": 1.1889072637069076, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6726362705230713, "learning_rate": 1e-05, "loss": 0.5025, "mean_token_accuracy": 0.8485746383666992, "num_tokens": 354948778.0, "step": 9346 }, { "epoch": 1.189034473985498, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5151853561401367, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8696928024291992, "num_tokens": 354990242.0, "step": 9347 }, { "epoch": 1.1891616842640884, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.665832757949829, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8827682137489319, "num_tokens": 355025462.0, "step": 9348 }, { "epoch": 1.189288894542679, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.522702693939209, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8608980178833008, "num_tokens": 355067488.0, "step": 9349 }, { "epoch": 1.1894161048212695, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6847050189971924, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8667740225791931, "num_tokens": 355101734.0, "step": 9350 }, { "epoch": 1.18954331509986, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5173118114471436, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8601687550544739, "num_tokens": 355140209.0, "step": 9351 }, { "epoch": 1.1896705253784505, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.684873104095459, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8677164316177368, "num_tokens": 355177424.0, "step": 9352 }, { "epoch": 1.189797735657041, "ewc_loss": 0.058837890625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.4981319904327393, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8727234601974487, "num_tokens": 355215112.0, "step": 9353 }, { "epoch": 1.1899249459356316, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.702082633972168, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8814972639083862, "num_tokens": 355250402.0, "step": 9354 }, { "epoch": 1.1900521562142221, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4935693740844727, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8733439445495605, "num_tokens": 355289887.0, "step": 9355 }, { "epoch": 1.1901793664928126, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.713811159133911, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8721379041671753, "num_tokens": 355330946.0, "step": 9356 }, { "epoch": 1.1903065767714032, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4903855323791504, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8724031448364258, "num_tokens": 355369029.0, "step": 9357 }, { "epoch": 1.1904337870499937, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.748673915863037, "learning_rate": 1e-05, "loss": 0.5109, "mean_token_accuracy": 0.8500406742095947, "num_tokens": 355406145.0, "step": 9358 }, { "epoch": 1.1905609973285842, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4823622703552246, "learning_rate": 1e-05, "loss": 0.3696, "mean_token_accuracy": 0.8903031945228577, "num_tokens": 355445702.0, "step": 9359 }, { "epoch": 1.1906882076071748, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.75333833694458, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8675068616867065, "num_tokens": 355485592.0, "step": 9360 }, { "epoch": 1.1908154178857653, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4891111850738525, "learning_rate": 1e-05, "loss": 0.3988, "mean_token_accuracy": 0.8853150010108948, "num_tokens": 355523208.0, "step": 9361 }, { "epoch": 1.1909426281643556, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7787671089172363, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8706129789352417, "num_tokens": 355557729.0, "step": 9362 }, { "epoch": 1.1910698384429461, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4875199794769287, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.881821870803833, "num_tokens": 355597222.0, "step": 9363 }, { "epoch": 1.1911970487215366, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.820507764816284, "learning_rate": 1e-05, "loss": 0.3972, "mean_token_accuracy": 0.8871500492095947, "num_tokens": 355630477.0, "step": 9364 }, { "epoch": 1.1913242590001272, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5116820335388184, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8702126741409302, "num_tokens": 355664506.0, "step": 9365 }, { "epoch": 1.1914514692787177, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.79866099357605, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8709115982055664, "num_tokens": 355710268.0, "step": 9366 }, { "epoch": 1.1915786795573082, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.53184175491333, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8562910556793213, "num_tokens": 355746396.0, "step": 9367 }, { "epoch": 1.1917058898358988, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.78383731842041, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8615624904632568, "num_tokens": 355787483.0, "step": 9368 }, { "epoch": 1.1918331001144893, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5370311737060547, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8643110990524292, "num_tokens": 355824138.0, "step": 9369 }, { "epoch": 1.1919603103930798, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7814743518829346, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8606200814247131, "num_tokens": 355854258.0, "step": 9370 }, { "epoch": 1.1920875206716703, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5456390380859375, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8682452440261841, "num_tokens": 355889750.0, "step": 9371 }, { "epoch": 1.1922147309502609, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7534477710723877, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8806102275848389, "num_tokens": 355923950.0, "step": 9372 }, { "epoch": 1.1923419412288512, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5439515113830566, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8775811791419983, "num_tokens": 355959807.0, "step": 9373 }, { "epoch": 1.1924691515074417, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7389471530914307, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8773989677429199, "num_tokens": 355992574.0, "step": 9374 }, { "epoch": 1.1925963617860322, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.536457061767578, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8629723787307739, "num_tokens": 356033726.0, "step": 9375 }, { "epoch": 1.1927235720646228, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7354049682617188, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.880927562713623, "num_tokens": 356066959.0, "step": 9376 }, { "epoch": 1.1928507823432133, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.543787717819214, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.871934711933136, "num_tokens": 356102316.0, "step": 9377 }, { "epoch": 1.1929779926218038, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.695822238922119, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.881560742855072, "num_tokens": 356146203.0, "step": 9378 }, { "epoch": 1.1931052029003943, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.547987699508667, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8600889444351196, "num_tokens": 356184673.0, "step": 9379 }, { "epoch": 1.1932324131789849, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.684647560119629, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8680385947227478, "num_tokens": 356224693.0, "step": 9380 }, { "epoch": 1.1933596234575754, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5385820865631104, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8638553619384766, "num_tokens": 356265454.0, "step": 9381 }, { "epoch": 1.193486833736166, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.686556816101074, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8697664737701416, "num_tokens": 356302866.0, "step": 9382 }, { "epoch": 1.1936140440147565, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5265049934387207, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.8736461997032166, "num_tokens": 356341897.0, "step": 9383 }, { "epoch": 1.193741254293347, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.682936429977417, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8779436349868774, "num_tokens": 356380637.0, "step": 9384 }, { "epoch": 1.1938684645719375, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.523102045059204, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8602341413497925, "num_tokens": 356422584.0, "step": 9385 }, { "epoch": 1.193995674850528, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.688558578491211, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8785794973373413, "num_tokens": 356457007.0, "step": 9386 }, { "epoch": 1.1941228851291183, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.526423454284668, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8745953440666199, "num_tokens": 356491837.0, "step": 9387 }, { "epoch": 1.1942500954077089, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6915197372436523, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8714603185653687, "num_tokens": 356527892.0, "step": 9388 }, { "epoch": 1.1943773056862994, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.519676446914673, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8660150766372681, "num_tokens": 356564310.0, "step": 9389 }, { "epoch": 1.19450451596489, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6781136989593506, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.878658652305603, "num_tokens": 356606476.0, "step": 9390 }, { "epoch": 1.1946317262434805, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5129663944244385, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.869243323802948, "num_tokens": 356646104.0, "step": 9391 }, { "epoch": 1.194758936522071, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.680023670196533, "learning_rate": 1e-05, "loss": 0.3811, "mean_token_accuracy": 0.8872330784797668, "num_tokens": 356687196.0, "step": 9392 }, { "epoch": 1.1948861468006615, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.531400442123413, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8702816963195801, "num_tokens": 356717634.0, "step": 9393 }, { "epoch": 1.195013357079252, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.687121868133545, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8580040335655212, "num_tokens": 356755494.0, "step": 9394 }, { "epoch": 1.1951405673578426, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.505610942840576, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8778830766677856, "num_tokens": 356800798.0, "step": 9395 }, { "epoch": 1.195267777636433, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.687117099761963, "learning_rate": 1e-05, "loss": 0.379, "mean_token_accuracy": 0.8918188810348511, "num_tokens": 356837819.0, "step": 9396 }, { "epoch": 1.1953949879150234, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5025177001953125, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8785117864608765, "num_tokens": 356876371.0, "step": 9397 }, { "epoch": 1.195522198193614, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7014076709747314, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8710846900939941, "num_tokens": 356911427.0, "step": 9398 }, { "epoch": 1.1956494084722045, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4950766563415527, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8798646926879883, "num_tokens": 356951229.0, "step": 9399 }, { "epoch": 1.195776618750795, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7098636627197266, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8781550526618958, "num_tokens": 356988954.0, "step": 9400 }, { "epoch": 1.1959038290293855, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5037317276000977, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8562465310096741, "num_tokens": 357027982.0, "step": 9401 }, { "epoch": 1.196031039307976, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7264962196350098, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.869149923324585, "num_tokens": 357061434.0, "step": 9402 }, { "epoch": 1.1961582495865666, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.498741388320923, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8675051331520081, "num_tokens": 357099255.0, "step": 9403 }, { "epoch": 1.196285459865157, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7509453296661377, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8782514333724976, "num_tokens": 357131917.0, "step": 9404 }, { "epoch": 1.1964126701437476, "ewc_loss": 0.059326171875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4821791648864746, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8799846172332764, "num_tokens": 357171520.0, "step": 9405 }, { "epoch": 1.1965398804223382, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.770141839981079, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8715406656265259, "num_tokens": 357208517.0, "step": 9406 }, { "epoch": 1.1966670907009287, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4937169551849365, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8816184401512146, "num_tokens": 357245300.0, "step": 9407 }, { "epoch": 1.1967943009795192, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7640223503112793, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8866312503814697, "num_tokens": 357284765.0, "step": 9408 }, { "epoch": 1.1969215112581097, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5117247104644775, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8621963262557983, "num_tokens": 357322968.0, "step": 9409 }, { "epoch": 1.1970487215367003, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7650258541107178, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8751562833786011, "num_tokens": 357360546.0, "step": 9410 }, { "epoch": 1.1971759318152906, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5104928016662598, "learning_rate": 1e-05, "loss": 0.3819, "mean_token_accuracy": 0.8851097822189331, "num_tokens": 357394433.0, "step": 9411 }, { "epoch": 1.1973031420938811, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.765930414199829, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8692432641983032, "num_tokens": 357431498.0, "step": 9412 }, { "epoch": 1.1974303523724716, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5320115089416504, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8668413162231445, "num_tokens": 357466930.0, "step": 9413 }, { "epoch": 1.1975575626510622, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7443997859954834, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8612174987792969, "num_tokens": 357509343.0, "step": 9414 }, { "epoch": 1.1976847729296527, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.54097843170166, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8722636699676514, "num_tokens": 357541980.0, "step": 9415 }, { "epoch": 1.1978119832082432, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.72324800491333, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8776695132255554, "num_tokens": 357579380.0, "step": 9416 }, { "epoch": 1.1979391934868338, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5315940380096436, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8816712498664856, "num_tokens": 357614666.0, "step": 9417 }, { "epoch": 1.1980664037654243, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.722982883453369, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8780199885368347, "num_tokens": 357648532.0, "step": 9418 }, { "epoch": 1.1981936140440148, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0002193450927734375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5370678901672363, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8768779635429382, "num_tokens": 357683585.0, "step": 9419 }, { "epoch": 1.1983208243226053, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.708350896835327, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.848922848701477, "num_tokens": 357725580.0, "step": 9420 }, { "epoch": 1.1984480346011959, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5273375511169434, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8694324493408203, "num_tokens": 357763017.0, "step": 9421 }, { "epoch": 1.1985752448797862, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7076809406280518, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8822950720787048, "num_tokens": 357797433.0, "step": 9422 }, { "epoch": 1.1987024551583767, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5225987434387207, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8687403202056885, "num_tokens": 357835348.0, "step": 9423 }, { "epoch": 1.1988296654369672, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6923820972442627, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8655440807342529, "num_tokens": 357877376.0, "step": 9424 }, { "epoch": 1.1989568757155578, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.528810739517212, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8704989552497864, "num_tokens": 357912983.0, "step": 9425 }, { "epoch": 1.1990840859941483, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6963391304016113, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8790604472160339, "num_tokens": 357948068.0, "step": 9426 }, { "epoch": 1.1992112962727388, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5301766395568848, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8772756457328796, "num_tokens": 357984019.0, "step": 9427 }, { "epoch": 1.1993385065513293, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6779556274414062, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8789342045783997, "num_tokens": 358022440.0, "step": 9428 }, { "epoch": 1.1994657168299199, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5273730754852295, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8704379796981812, "num_tokens": 358061615.0, "step": 9429 }, { "epoch": 1.1995929271085104, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6926140785217285, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8674353361129761, "num_tokens": 358097042.0, "step": 9430 }, { "epoch": 1.199720137387101, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.550675392150879, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8656113147735596, "num_tokens": 358134067.0, "step": 9431 }, { "epoch": 1.1998473476656915, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6584887504577637, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8700948357582092, "num_tokens": 358178975.0, "step": 9432 }, { "epoch": 1.199974557944282, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5403871536254883, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8726948499679565, "num_tokens": 358212879.0, "step": 9433 }, { "epoch": 1.2001017682228725, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6567490100860596, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8837100267410278, "num_tokens": 358245711.0, "step": 9434 }, { "epoch": 1.200228978501463, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5308761596679688, "learning_rate": 1e-05, "loss": 0.3921, "mean_token_accuracy": 0.8841665983200073, "num_tokens": 358283073.0, "step": 9435 }, { "epoch": 1.2003561887800533, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6522819995880127, "learning_rate": 1e-05, "loss": 0.4058, "mean_token_accuracy": 0.880246639251709, "num_tokens": 358320338.0, "step": 9436 }, { "epoch": 1.2004833990586439, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5334906578063965, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8710002899169922, "num_tokens": 358358704.0, "step": 9437 }, { "epoch": 1.2006106093372344, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6550519466400146, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8781728148460388, "num_tokens": 358393512.0, "step": 9438 }, { "epoch": 1.200737819615825, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.533728837966919, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8633546829223633, "num_tokens": 358435843.0, "step": 9439 }, { "epoch": 1.2008650298944155, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.65043568611145, "learning_rate": 1e-05, "loss": 0.3752, "mean_token_accuracy": 0.8872972726821899, "num_tokens": 358468918.0, "step": 9440 }, { "epoch": 1.200992240173006, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.523730754852295, "learning_rate": 1e-05, "loss": 0.5026, "mean_token_accuracy": 0.8517462015151978, "num_tokens": 358511275.0, "step": 9441 }, { "epoch": 1.2011194504515965, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6571311950683594, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8633586764335632, "num_tokens": 358552665.0, "step": 9442 }, { "epoch": 1.201246660730187, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.529001474380493, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8787970542907715, "num_tokens": 358584725.0, "step": 9443 }, { "epoch": 1.2013738710087776, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.665494918823242, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8770033121109009, "num_tokens": 358620026.0, "step": 9444 }, { "epoch": 1.201501081287368, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.517742156982422, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8700429201126099, "num_tokens": 358657386.0, "step": 9445 }, { "epoch": 1.2016282915659584, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.668503522872925, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8740115165710449, "num_tokens": 358695069.0, "step": 9446 }, { "epoch": 1.201755501844549, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.523740768432617, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8789739012718201, "num_tokens": 358728091.0, "step": 9447 }, { "epoch": 1.2018827121231395, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.669600248336792, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8580443263053894, "num_tokens": 358767930.0, "step": 9448 }, { "epoch": 1.20200992240173, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.515075445175171, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8677700757980347, "num_tokens": 358803705.0, "step": 9449 }, { "epoch": 1.2021371326803205, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.68217396736145, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8719450831413269, "num_tokens": 358840730.0, "step": 9450 }, { "epoch": 1.202264342958911, "ewc_loss": 0.05859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.6716461181640625e-05, "grad_norm": 2.501217842102051, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8653196096420288, "num_tokens": 358881457.0, "step": 9451 }, { "epoch": 1.2023915532375016, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.710547924041748, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.864971399307251, "num_tokens": 358917034.0, "step": 9452 }, { "epoch": 1.202518763516092, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4987969398498535, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8635399341583252, "num_tokens": 358957851.0, "step": 9453 }, { "epoch": 1.2026459737946826, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.714398145675659, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8630509376525879, "num_tokens": 358999816.0, "step": 9454 }, { "epoch": 1.2027731840732732, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5004591941833496, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8504904508590698, "num_tokens": 359041067.0, "step": 9455 }, { "epoch": 1.2029003943518637, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7231972217559814, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8696932792663574, "num_tokens": 359080495.0, "step": 9456 }, { "epoch": 1.2030276046304542, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.499150037765503, "learning_rate": 1e-05, "loss": 0.3965, "mean_token_accuracy": 0.8835163712501526, "num_tokens": 359118245.0, "step": 9457 }, { "epoch": 1.2031548149090447, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.736128091812134, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.860012412071228, "num_tokens": 359159840.0, "step": 9458 }, { "epoch": 1.2032820251876353, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.502500534057617, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8673737049102783, "num_tokens": 359198917.0, "step": 9459 }, { "epoch": 1.2034092354662256, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7580854892730713, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8683966398239136, "num_tokens": 359231904.0, "step": 9460 }, { "epoch": 1.203536445744816, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4989571571350098, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8575630187988281, "num_tokens": 359271713.0, "step": 9461 }, { "epoch": 1.2036636560234066, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.755384922027588, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8693386316299438, "num_tokens": 359312548.0, "step": 9462 }, { "epoch": 1.2037908663019972, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.496347427368164, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.8803045153617859, "num_tokens": 359348171.0, "step": 9463 }, { "epoch": 1.2039180765805877, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7924835681915283, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8608831167221069, "num_tokens": 359383863.0, "step": 9464 }, { "epoch": 1.2040452868591782, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.4978275299072266, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8649778962135315, "num_tokens": 359421595.0, "step": 9465 }, { "epoch": 1.2041724971377687, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7920804023742676, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8825297355651855, "num_tokens": 359457889.0, "step": 9466 }, { "epoch": 1.2042997074163593, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5140066146850586, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8656489849090576, "num_tokens": 359496938.0, "step": 9467 }, { "epoch": 1.2044269176949498, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7885591983795166, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8635067939758301, "num_tokens": 359536124.0, "step": 9468 }, { "epoch": 1.2045541279735403, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5293705463409424, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8629255294799805, "num_tokens": 359573525.0, "step": 9469 }, { "epoch": 1.2046813382521309, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.76583194732666, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8739917874336243, "num_tokens": 359614960.0, "step": 9470 }, { "epoch": 1.2048085485307212, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5400335788726807, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8642388582229614, "num_tokens": 359653367.0, "step": 9471 }, { "epoch": 1.2049357588093117, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.766733407974243, "learning_rate": 1e-05, "loss": 0.503, "mean_token_accuracy": 0.8509780764579773, "num_tokens": 359692983.0, "step": 9472 }, { "epoch": 1.2050629690879022, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.548583984375, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8621702790260315, "num_tokens": 359729487.0, "step": 9473 }, { "epoch": 1.2051901793664928, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7319462299346924, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.878797709941864, "num_tokens": 359764805.0, "step": 9474 }, { "epoch": 1.2053173896450833, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5476675033569336, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.859451174736023, "num_tokens": 359803014.0, "step": 9475 }, { "epoch": 1.2054445999236738, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.735957384109497, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8718674778938293, "num_tokens": 359837866.0, "step": 9476 }, { "epoch": 1.2055718102022643, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5331690311431885, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8847904205322266, "num_tokens": 359874052.0, "step": 9477 }, { "epoch": 1.2056990204808549, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.706721544265747, "learning_rate": 1e-05, "loss": 0.3858, "mean_token_accuracy": 0.8877450227737427, "num_tokens": 359915746.0, "step": 9478 }, { "epoch": 1.2058262307594454, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.544287919998169, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8829888105392456, "num_tokens": 359945857.0, "step": 9479 }, { "epoch": 1.205953441038036, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.693964719772339, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8771403431892395, "num_tokens": 359988455.0, "step": 9480 }, { "epoch": 1.2060806513166265, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.524245262145996, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8768582940101624, "num_tokens": 360030610.0, "step": 9481 }, { "epoch": 1.206207861595217, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.695622444152832, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8628950119018555, "num_tokens": 360068919.0, "step": 9482 }, { "epoch": 1.2063350718738075, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5490667819976807, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8691411018371582, "num_tokens": 360099073.0, "step": 9483 }, { "epoch": 1.206462282152398, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.673759698867798, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.872402548789978, "num_tokens": 360141947.0, "step": 9484 }, { "epoch": 1.2065894924309883, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.544062376022339, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8680030107498169, "num_tokens": 360179435.0, "step": 9485 }, { "epoch": 1.2067167027095789, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6620969772338867, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8773972392082214, "num_tokens": 360221254.0, "step": 9486 }, { "epoch": 1.2068439129881694, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5321502685546875, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8792672753334045, "num_tokens": 360261657.0, "step": 9487 }, { "epoch": 1.20697112326676, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6627659797668457, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8737587928771973, "num_tokens": 360294408.0, "step": 9488 }, { "epoch": 1.2070983335453505, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5193827152252197, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8792824745178223, "num_tokens": 360332275.0, "step": 9489 }, { "epoch": 1.207225543823941, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.659573554992676, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8786827325820923, "num_tokens": 360372381.0, "step": 9490 }, { "epoch": 1.2073527541025315, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.6035869121551514, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8655998110771179, "num_tokens": 360404066.0, "step": 9491 }, { "epoch": 1.207479964381122, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.630045175552368, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.8798802495002747, "num_tokens": 360434457.0, "step": 9492 }, { "epoch": 1.2076071746597126, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5641255378723145, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8717514276504517, "num_tokens": 360467905.0, "step": 9493 }, { "epoch": 1.207734384938303, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.607893943786621, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8739826679229736, "num_tokens": 360505190.0, "step": 9494 }, { "epoch": 1.2078615952168934, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5614688396453857, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8738884329795837, "num_tokens": 360539191.0, "step": 9495 }, { "epoch": 1.207988805495484, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.639484167098999, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8590694069862366, "num_tokens": 360569508.0, "step": 9496 }, { "epoch": 1.2081160157740745, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5442628860473633, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8805496692657471, "num_tokens": 360604669.0, "step": 9497 }, { "epoch": 1.208243226052665, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6183745861053467, "learning_rate": 1e-05, "loss": 0.3599, "mean_token_accuracy": 0.8934537172317505, "num_tokens": 360639009.0, "step": 9498 }, { "epoch": 1.2083704363312555, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022125244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5409131050109863, "learning_rate": 1e-05, "loss": 0.3724, "mean_token_accuracy": 0.8865357637405396, "num_tokens": 360673857.0, "step": 9499 }, { "epoch": 1.208497646609846, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6201610565185547, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8625882267951965, "num_tokens": 360717553.0, "step": 9500 }, { "epoch": 1.2086248568884366, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5318939685821533, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8729103803634644, "num_tokens": 360761013.0, "step": 9501 }, { "epoch": 1.208752067167027, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.624614715576172, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8712449073791504, "num_tokens": 360804302.0, "step": 9502 }, { "epoch": 1.2088792774456176, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5237762928009033, "learning_rate": 1e-05, "loss": 0.3524, "mean_token_accuracy": 0.8920011520385742, "num_tokens": 360837045.0, "step": 9503 }, { "epoch": 1.2090064877242082, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6377112865448, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8722734451293945, "num_tokens": 360876243.0, "step": 9504 }, { "epoch": 1.2091336980027987, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5271637439727783, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8681974411010742, "num_tokens": 360915112.0, "step": 9505 }, { "epoch": 1.2092609082813892, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6370017528533936, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8745605945587158, "num_tokens": 360951856.0, "step": 9506 }, { "epoch": 1.2093881185599797, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5162956714630127, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8787939548492432, "num_tokens": 360987243.0, "step": 9507 }, { "epoch": 1.2095153288385703, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6449944972991943, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.877754807472229, "num_tokens": 361025099.0, "step": 9508 }, { "epoch": 1.2096425391171606, "ewc_loss": 0.05908203125, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.538389205932617, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8548707365989685, "num_tokens": 361055232.0, "step": 9509 }, { "epoch": 1.209769749395751, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6494948863983154, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8719203472137451, "num_tokens": 361092500.0, "step": 9510 }, { "epoch": 1.2098969596743416, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.526841163635254, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8776552677154541, "num_tokens": 361127044.0, "step": 9511 }, { "epoch": 1.2100241699529322, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.7112369537353516, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8911949396133423, "num_tokens": 361163065.0, "step": 9512 }, { "epoch": 1.2101513802315227, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.489729881286621, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8636263608932495, "num_tokens": 361204465.0, "step": 9513 }, { "epoch": 1.2102785905101132, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.71935772895813, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.850125789642334, "num_tokens": 361243788.0, "step": 9514 }, { "epoch": 1.2104058007887037, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4853415489196777, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8777624368667603, "num_tokens": 361281427.0, "step": 9515 }, { "epoch": 1.2105330110672943, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.73646879196167, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8568393588066101, "num_tokens": 361324695.0, "step": 9516 }, { "epoch": 1.2106602213458848, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4705986976623535, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8697820901870728, "num_tokens": 361370446.0, "step": 9517 }, { "epoch": 1.2107874316244753, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7554166316986084, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8696208000183105, "num_tokens": 361414622.0, "step": 9518 }, { "epoch": 1.2109146419030659, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.480828046798706, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.882226288318634, "num_tokens": 361452363.0, "step": 9519 }, { "epoch": 1.2110418521816562, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.809616804122925, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8563603162765503, "num_tokens": 361484195.0, "step": 9520 }, { "epoch": 1.2111690624602467, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.484938859939575, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8700261116027832, "num_tokens": 361523604.0, "step": 9521 }, { "epoch": 1.2112962727388372, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.815476179122925, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8563004732131958, "num_tokens": 361563723.0, "step": 9522 }, { "epoch": 1.2114234830174277, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.498784303665161, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8729882836341858, "num_tokens": 361602578.0, "step": 9523 }, { "epoch": 1.2115506932960183, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8265910148620605, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8694698810577393, "num_tokens": 361644255.0, "step": 9524 }, { "epoch": 1.2116779035746088, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.522719621658325, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8726229667663574, "num_tokens": 361683424.0, "step": 9525 }, { "epoch": 1.2118051138531993, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8201377391815186, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.881202757358551, "num_tokens": 361720965.0, "step": 9526 }, { "epoch": 1.2119323241317899, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5615453720092773, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8577509522438049, "num_tokens": 361753900.0, "step": 9527 }, { "epoch": 1.2120595344103804, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.791611433029175, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8707584738731384, "num_tokens": 361793072.0, "step": 9528 }, { "epoch": 1.212186744688971, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002231597900390625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5531625747680664, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8540226221084595, "num_tokens": 361831082.0, "step": 9529 }, { "epoch": 1.2123139549675614, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.761321783065796, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8741429448127747, "num_tokens": 361864834.0, "step": 9530 }, { "epoch": 1.212441165246152, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5556206703186035, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8722275495529175, "num_tokens": 361905567.0, "step": 9531 }, { "epoch": 1.2125683755247425, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7310116291046143, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8804364204406738, "num_tokens": 361945718.0, "step": 9532 }, { "epoch": 1.212695585803333, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5536906719207764, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8641724586486816, "num_tokens": 361985317.0, "step": 9533 }, { "epoch": 1.2128227960819233, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.735062599182129, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8727710247039795, "num_tokens": 362020725.0, "step": 9534 }, { "epoch": 1.2129500063605139, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5637829303741455, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8613000512123108, "num_tokens": 362054398.0, "step": 9535 }, { "epoch": 1.2130772166391044, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.704970359802246, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8554748892784119, "num_tokens": 362092075.0, "step": 9536 }, { "epoch": 1.213204426917695, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5358519554138184, "learning_rate": 1e-05, "loss": 0.3781, "mean_token_accuracy": 0.8920629620552063, "num_tokens": 362128286.0, "step": 9537 }, { "epoch": 1.2133316371962855, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.702751636505127, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8639905452728271, "num_tokens": 362169965.0, "step": 9538 }, { "epoch": 1.213458847474876, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5317559242248535, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8702231049537659, "num_tokens": 362211638.0, "step": 9539 }, { "epoch": 1.2135860577534665, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.703836679458618, "learning_rate": 1e-05, "loss": 0.3811, "mean_token_accuracy": 0.8877679705619812, "num_tokens": 362241597.0, "step": 9540 }, { "epoch": 1.213713268032057, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5217535495758057, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8690351247787476, "num_tokens": 362280028.0, "step": 9541 }, { "epoch": 1.2138404783106476, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.694983959197998, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8701058030128479, "num_tokens": 362323611.0, "step": 9542 }, { "epoch": 1.213967688589238, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5273563861846924, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8719836473464966, "num_tokens": 362359097.0, "step": 9543 }, { "epoch": 1.2140948988678284, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.687674045562744, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8691928386688232, "num_tokens": 362402175.0, "step": 9544 }, { "epoch": 1.214222109146419, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.527161121368408, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8716055154800415, "num_tokens": 362442581.0, "step": 9545 }, { "epoch": 1.2143493194250095, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.68035888671875, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.880972146987915, "num_tokens": 362480691.0, "step": 9546 }, { "epoch": 1.2144765297036, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.518972873687744, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8727277517318726, "num_tokens": 362522864.0, "step": 9547 }, { "epoch": 1.2146037399821905, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6773147583007812, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8727525472640991, "num_tokens": 362564431.0, "step": 9548 }, { "epoch": 1.214730950260781, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5160813331604004, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8840959668159485, "num_tokens": 362599670.0, "step": 9549 }, { "epoch": 1.2148581605393716, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6956002712249756, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8675323724746704, "num_tokens": 362637588.0, "step": 9550 }, { "epoch": 1.214985370817962, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5185723304748535, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8733400106430054, "num_tokens": 362674679.0, "step": 9551 }, { "epoch": 1.2151125810965526, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.706172466278076, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8665941953659058, "num_tokens": 362705840.0, "step": 9552 }, { "epoch": 1.2152397913751432, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5272939205169678, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8692315816879272, "num_tokens": 362739880.0, "step": 9553 }, { "epoch": 1.2153670016537337, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.690171480178833, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8790693283081055, "num_tokens": 362779101.0, "step": 9554 }, { "epoch": 1.2154942119323242, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.499290704727173, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8750969171524048, "num_tokens": 362821034.0, "step": 9555 }, { "epoch": 1.2156214222109147, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7137415409088135, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8810458183288574, "num_tokens": 362856895.0, "step": 9556 }, { "epoch": 1.2157486324895053, "ewc_loss": 0.059814453125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5054752826690674, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8703218698501587, "num_tokens": 362894955.0, "step": 9557 }, { "epoch": 1.2158758427680956, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.725188732147217, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8664301037788391, "num_tokens": 362930397.0, "step": 9558 }, { "epoch": 1.216003053046686, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.489703416824341, "learning_rate": 1e-05, "loss": 0.383, "mean_token_accuracy": 0.8888086080551147, "num_tokens": 362970728.0, "step": 9559 }, { "epoch": 1.2161302633252766, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.741101026535034, "learning_rate": 1e-05, "loss": 0.3956, "mean_token_accuracy": 0.8861989378929138, "num_tokens": 363007803.0, "step": 9560 }, { "epoch": 1.2162574736038672, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5036544799804688, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8698900938034058, "num_tokens": 363047815.0, "step": 9561 }, { "epoch": 1.2163846838824577, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7468507289886475, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8794873952865601, "num_tokens": 363084156.0, "step": 9562 }, { "epoch": 1.2165118941610482, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5026333332061768, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8679394721984863, "num_tokens": 363125390.0, "step": 9563 }, { "epoch": 1.2166391044396387, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.751055955886841, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8725587129592896, "num_tokens": 363164210.0, "step": 9564 }, { "epoch": 1.2167663147182293, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5207955837249756, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.854033350944519, "num_tokens": 363200356.0, "step": 9565 }, { "epoch": 1.2168935249968198, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7489397525787354, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.852985143661499, "num_tokens": 363239622.0, "step": 9566 }, { "epoch": 1.2170207352754103, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5223803520202637, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8759593963623047, "num_tokens": 363277348.0, "step": 9567 }, { "epoch": 1.2171479455540009, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.751209020614624, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.87195885181427, "num_tokens": 363310077.0, "step": 9568 }, { "epoch": 1.2172751558325912, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.531675100326538, "learning_rate": 1e-05, "loss": 0.534, "mean_token_accuracy": 0.8380385637283325, "num_tokens": 363347919.0, "step": 9569 }, { "epoch": 1.2174023661111817, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.722860336303711, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8731696605682373, "num_tokens": 363390818.0, "step": 9570 }, { "epoch": 1.2175295763897722, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5554251670837402, "learning_rate": 1e-05, "loss": 0.524, "mean_token_accuracy": 0.8463083505630493, "num_tokens": 363424028.0, "step": 9571 }, { "epoch": 1.2176567866683627, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7125015258789062, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8693637847900391, "num_tokens": 363456862.0, "step": 9572 }, { "epoch": 1.2177839969469533, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5212574005126953, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8746212720870972, "num_tokens": 363498538.0, "step": 9573 }, { "epoch": 1.2179112072255438, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7083077430725098, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8754924535751343, "num_tokens": 363533050.0, "step": 9574 }, { "epoch": 1.2180384175041343, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5222020149230957, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8702808618545532, "num_tokens": 363574151.0, "step": 9575 }, { "epoch": 1.2181656277827249, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.70298171043396, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8671464920043945, "num_tokens": 363615829.0, "step": 9576 }, { "epoch": 1.2182928380613154, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.524179697036743, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8699709177017212, "num_tokens": 363649974.0, "step": 9577 }, { "epoch": 1.218420048339906, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.701630115509033, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8729262948036194, "num_tokens": 363691196.0, "step": 9578 }, { "epoch": 1.2185472586184964, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5134117603302, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8835315108299255, "num_tokens": 363726496.0, "step": 9579 }, { "epoch": 1.218674468897087, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7019481658935547, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8616054654121399, "num_tokens": 363769119.0, "step": 9580 }, { "epoch": 1.2188016791756775, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5369343757629395, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8563363552093506, "num_tokens": 363803363.0, "step": 9581 }, { "epoch": 1.218928889454268, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6943254470825195, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8757407665252686, "num_tokens": 363842602.0, "step": 9582 }, { "epoch": 1.2190560997328583, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.519587278366089, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8695919513702393, "num_tokens": 363882758.0, "step": 9583 }, { "epoch": 1.2191833100114489, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.687535285949707, "learning_rate": 1e-05, "loss": 0.3937, "mean_token_accuracy": 0.884247899055481, "num_tokens": 363923738.0, "step": 9584 }, { "epoch": 1.2193105202900394, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5217645168304443, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8792625665664673, "num_tokens": 363957999.0, "step": 9585 }, { "epoch": 1.21943773056863, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6972239017486572, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8734278678894043, "num_tokens": 363992653.0, "step": 9586 }, { "epoch": 1.2195649408472204, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.515744686126709, "learning_rate": 1e-05, "loss": 0.3852, "mean_token_accuracy": 0.8852569460868835, "num_tokens": 364027376.0, "step": 9587 }, { "epoch": 1.219692151125811, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7009940147399902, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8705560564994812, "num_tokens": 364063299.0, "step": 9588 }, { "epoch": 1.2198193614044015, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.513920783996582, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8437941074371338, "num_tokens": 364105724.0, "step": 9589 }, { "epoch": 1.219946571682992, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7163257598876953, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8611598014831543, "num_tokens": 364140427.0, "step": 9590 }, { "epoch": 1.2200737819615826, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5023021697998047, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8603106737136841, "num_tokens": 364185262.0, "step": 9591 }, { "epoch": 1.220200992240173, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7253308296203613, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8727380037307739, "num_tokens": 364219006.0, "step": 9592 }, { "epoch": 1.2203282025187634, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.503948926925659, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.876686155796051, "num_tokens": 364259699.0, "step": 9593 }, { "epoch": 1.220455412797354, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.725832223892212, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8774638175964355, "num_tokens": 364297759.0, "step": 9594 }, { "epoch": 1.2205826230759445, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5048770904541016, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8677515983581543, "num_tokens": 364338683.0, "step": 9595 }, { "epoch": 1.220709833354535, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.740229368209839, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8714228868484497, "num_tokens": 364377587.0, "step": 9596 }, { "epoch": 1.2208370436331255, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5064730644226074, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8644918203353882, "num_tokens": 364416740.0, "step": 9597 }, { "epoch": 1.220964253911716, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.74284291267395, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8609617948532104, "num_tokens": 364458695.0, "step": 9598 }, { "epoch": 1.2210914641903066, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022411346435546875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5054283142089844, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8636089563369751, "num_tokens": 364501395.0, "step": 9599 }, { "epoch": 1.221218674468897, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7432289123535156, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8776752948760986, "num_tokens": 364534126.0, "step": 9600 }, { "epoch": 1.2213458847474876, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5081725120544434, "learning_rate": 1e-05, "loss": 0.3928, "mean_token_accuracy": 0.8849354982376099, "num_tokens": 364573776.0, "step": 9601 }, { "epoch": 1.2214730950260781, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7375526428222656, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8668870329856873, "num_tokens": 364617900.0, "step": 9602 }, { "epoch": 1.2216003053046687, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5073461532592773, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8791024684906006, "num_tokens": 364658992.0, "step": 9603 }, { "epoch": 1.2217275155832592, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.745487689971924, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8857415914535522, "num_tokens": 364695224.0, "step": 9604 }, { "epoch": 1.2218547258618497, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5288610458374023, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8744022846221924, "num_tokens": 364731476.0, "step": 9605 }, { "epoch": 1.2219819361404403, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.728013277053833, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8724187612533569, "num_tokens": 364774129.0, "step": 9606 }, { "epoch": 1.2221091464190306, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5383834838867188, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8714737296104431, "num_tokens": 364808097.0, "step": 9607 }, { "epoch": 1.222236356697621, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7075319290161133, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8791659474372864, "num_tokens": 364847592.0, "step": 9608 }, { "epoch": 1.2223635669762116, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5274906158447266, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8742319345474243, "num_tokens": 364885846.0, "step": 9609 }, { "epoch": 1.2224907772548022, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7194602489471436, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8572779297828674, "num_tokens": 364927536.0, "step": 9610 }, { "epoch": 1.2226179875333927, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5349678993225098, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8712380528450012, "num_tokens": 364961050.0, "step": 9611 }, { "epoch": 1.2227451978119832, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7048492431640625, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8655288219451904, "num_tokens": 365001058.0, "step": 9612 }, { "epoch": 1.2228724080905737, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.522388219833374, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8785579800605774, "num_tokens": 365041345.0, "step": 9613 }, { "epoch": 1.2229996183691643, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.693138360977173, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.869135320186615, "num_tokens": 365085669.0, "step": 9614 }, { "epoch": 1.2231268286477548, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.507702589035034, "learning_rate": 1e-05, "loss": 0.3893, "mean_token_accuracy": 0.8864305019378662, "num_tokens": 365125630.0, "step": 9615 }, { "epoch": 1.2232540389263453, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.711186170578003, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8737027645111084, "num_tokens": 365162904.0, "step": 9616 }, { "epoch": 1.2233812492049359, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.523254871368408, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8735693693161011, "num_tokens": 365196509.0, "step": 9617 }, { "epoch": 1.2235084594835262, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7196037769317627, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.857237696647644, "num_tokens": 365232947.0, "step": 9618 }, { "epoch": 1.2236356697621167, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.514134407043457, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8729183673858643, "num_tokens": 365273434.0, "step": 9619 }, { "epoch": 1.2237628800407072, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7123959064483643, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8648680448532104, "num_tokens": 365316638.0, "step": 9620 }, { "epoch": 1.2238900903192977, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5291271209716797, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8667261600494385, "num_tokens": 365348519.0, "step": 9621 }, { "epoch": 1.2240173005978883, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.708413600921631, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8837769031524658, "num_tokens": 365383782.0, "step": 9622 }, { "epoch": 1.2241445108764788, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5264625549316406, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8615130186080933, "num_tokens": 365421363.0, "step": 9623 }, { "epoch": 1.2242717211550693, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000225067138671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.688413619995117, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.871304988861084, "num_tokens": 365468207.0, "step": 9624 }, { "epoch": 1.2243989314336599, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5370302200317383, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8777562379837036, "num_tokens": 365498318.0, "step": 9625 }, { "epoch": 1.2245261417122504, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.683675765991211, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8732128143310547, "num_tokens": 365536140.0, "step": 9626 }, { "epoch": 1.224653351990841, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.531859874725342, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8580440282821655, "num_tokens": 365576490.0, "step": 9627 }, { "epoch": 1.2247805622694314, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.679804801940918, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8682416081428528, "num_tokens": 365618043.0, "step": 9628 }, { "epoch": 1.224907772548022, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5564045906066895, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8643999099731445, "num_tokens": 365648303.0, "step": 9629 }, { "epoch": 1.2250349828266125, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.664870262145996, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8706960678100586, "num_tokens": 365684739.0, "step": 9630 }, { "epoch": 1.225162193105203, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.525474786758423, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8568257093429565, "num_tokens": 365726542.0, "step": 9631 }, { "epoch": 1.2252894033837933, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6606321334838867, "learning_rate": 1e-05, "loss": 0.4004, "mean_token_accuracy": 0.880268931388855, "num_tokens": 365769949.0, "step": 9632 }, { "epoch": 1.2254166136623839, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5316996574401855, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8729535341262817, "num_tokens": 365805891.0, "step": 9633 }, { "epoch": 1.2255438239409744, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6798036098480225, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8523048162460327, "num_tokens": 365847011.0, "step": 9634 }, { "epoch": 1.225671034219565, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.518740653991699, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8736748695373535, "num_tokens": 365884546.0, "step": 9635 }, { "epoch": 1.2257982444981554, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6823267936706543, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.880597710609436, "num_tokens": 365919619.0, "step": 9636 }, { "epoch": 1.225925454776746, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5108580589294434, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8787835836410522, "num_tokens": 365956020.0, "step": 9637 }, { "epoch": 1.2260526650553365, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.699836492538452, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8492445945739746, "num_tokens": 365994945.0, "step": 9638 }, { "epoch": 1.226179875333927, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.518476963043213, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8633874654769897, "num_tokens": 366036476.0, "step": 9639 }, { "epoch": 1.2263070856125176, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6863901615142822, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8784601092338562, "num_tokens": 366072966.0, "step": 9640 }, { "epoch": 1.226434295891108, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.512000560760498, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8730083703994751, "num_tokens": 366111535.0, "step": 9641 }, { "epoch": 1.2265615061696984, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6947009563446045, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8662248253822327, "num_tokens": 366151104.0, "step": 9642 }, { "epoch": 1.226688716448289, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.534292697906494, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8630033731460571, "num_tokens": 366184062.0, "step": 9643 }, { "epoch": 1.2268159267268794, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6854665279388428, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8752077221870422, "num_tokens": 366217774.0, "step": 9644 }, { "epoch": 1.22694313700547, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5186312198638916, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8783226609230042, "num_tokens": 366252776.0, "step": 9645 }, { "epoch": 1.2270703472840605, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6936540603637695, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8686372637748718, "num_tokens": 366289031.0, "step": 9646 }, { "epoch": 1.227197557562651, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.504467248916626, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8826455473899841, "num_tokens": 366327919.0, "step": 9647 }, { "epoch": 1.2273247678412416, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.709953784942627, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8575254678726196, "num_tokens": 366368655.0, "step": 9648 }, { "epoch": 1.227451978119832, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.511137008666992, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.872075080871582, "num_tokens": 366406987.0, "step": 9649 }, { "epoch": 1.2275791883984226, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.713543653488159, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8582817316055298, "num_tokens": 366447360.0, "step": 9650 }, { "epoch": 1.2277063986770131, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4985930919647217, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8781408071517944, "num_tokens": 366487640.0, "step": 9651 }, { "epoch": 1.2278336089556037, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7383525371551514, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8617463111877441, "num_tokens": 366526569.0, "step": 9652 }, { "epoch": 1.2279608192341942, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.491997003555298, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8660539984703064, "num_tokens": 366567503.0, "step": 9653 }, { "epoch": 1.2280880295127847, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.745438575744629, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8693467378616333, "num_tokens": 366609230.0, "step": 9654 }, { "epoch": 1.2282152397913753, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.507023572921753, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8729459047317505, "num_tokens": 366646982.0, "step": 9655 }, { "epoch": 1.2283424500699656, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7604761123657227, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8501666188240051, "num_tokens": 366684051.0, "step": 9656 }, { "epoch": 1.228469660348556, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5073795318603516, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.855258584022522, "num_tokens": 366724617.0, "step": 9657 }, { "epoch": 1.2285968706271466, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.766509771347046, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8862193822860718, "num_tokens": 366756642.0, "step": 9658 }, { "epoch": 1.2287240809057371, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5309858322143555, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8699782490730286, "num_tokens": 366789197.0, "step": 9659 }, { "epoch": 1.2288512911843277, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.758418560028076, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8678518533706665, "num_tokens": 366825684.0, "step": 9660 }, { "epoch": 1.2289785014629182, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.524280309677124, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8590270280838013, "num_tokens": 366871597.0, "step": 9661 }, { "epoch": 1.2291057117415087, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7411911487579346, "learning_rate": 1e-05, "loss": 0.3897, "mean_token_accuracy": 0.88609778881073, "num_tokens": 366910961.0, "step": 9662 }, { "epoch": 1.2292329220200993, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5150868892669678, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8841500878334045, "num_tokens": 366951701.0, "step": 9663 }, { "epoch": 1.2293601322986898, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7544097900390625, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8613331317901611, "num_tokens": 366992472.0, "step": 9664 }, { "epoch": 1.2294873425772803, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5441784858703613, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8628401756286621, "num_tokens": 367031220.0, "step": 9665 }, { "epoch": 1.2296145528558708, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.742004632949829, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8640292882919312, "num_tokens": 367067275.0, "step": 9666 }, { "epoch": 1.2297417631344612, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.522955894470215, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.8547329306602478, "num_tokens": 367111851.0, "step": 9667 }, { "epoch": 1.2298689734130517, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.736809015274048, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8650974631309509, "num_tokens": 367150895.0, "step": 9668 }, { "epoch": 1.2299961836916422, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5328338146209717, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8584860563278198, "num_tokens": 367187874.0, "step": 9669 }, { "epoch": 1.2301233939702327, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.725969076156616, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8756510615348816, "num_tokens": 367228859.0, "step": 9670 }, { "epoch": 1.2302506042488233, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5268421173095703, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8862283229827881, "num_tokens": 367264686.0, "step": 9671 }, { "epoch": 1.2303778145274138, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.707498073577881, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8751014471054077, "num_tokens": 367309909.0, "step": 9672 }, { "epoch": 1.2305050248060043, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5401458740234375, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8785498142242432, "num_tokens": 367343996.0, "step": 9673 }, { "epoch": 1.2306322350845948, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.700087070465088, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.867638349533081, "num_tokens": 367386267.0, "step": 9674 }, { "epoch": 1.2307594453631854, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5190937519073486, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8629021644592285, "num_tokens": 367430256.0, "step": 9675 }, { "epoch": 1.230886655641776, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7090952396392822, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.855975866317749, "num_tokens": 367469189.0, "step": 9676 }, { "epoch": 1.2310138659203664, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.559654474258423, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8563997745513916, "num_tokens": 367501323.0, "step": 9677 }, { "epoch": 1.231141076198957, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6832265853881836, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8655530214309692, "num_tokens": 367542361.0, "step": 9678 }, { "epoch": 1.2312682864775475, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5595250129699707, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8534423112869263, "num_tokens": 367575859.0, "step": 9679 }, { "epoch": 1.231395496756138, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6662158966064453, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8708887100219727, "num_tokens": 367611330.0, "step": 9680 }, { "epoch": 1.2315227070347283, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5276761054992676, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8821156024932861, "num_tokens": 367653110.0, "step": 9681 }, { "epoch": 1.2316499173133189, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6746063232421875, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8777520656585693, "num_tokens": 367687799.0, "step": 9682 }, { "epoch": 1.2317771275919094, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5285439491271973, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8717148900032043, "num_tokens": 367728253.0, "step": 9683 }, { "epoch": 1.2319043378705, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.654202938079834, "learning_rate": 1e-05, "loss": 0.3537, "mean_token_accuracy": 0.8995825052261353, "num_tokens": 367770397.0, "step": 9684 }, { "epoch": 1.2320315481490904, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.527967691421509, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8690224885940552, "num_tokens": 367810630.0, "step": 9685 }, { "epoch": 1.232158758427681, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6635873317718506, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8634747266769409, "num_tokens": 367853563.0, "step": 9686 }, { "epoch": 1.2322859687062715, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5512192249298096, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8586164116859436, "num_tokens": 367888113.0, "step": 9687 }, { "epoch": 1.232413178984862, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6546757221221924, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8654547929763794, "num_tokens": 367925839.0, "step": 9688 }, { "epoch": 1.2325403892634526, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.531031370162964, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8594614267349243, "num_tokens": 367966944.0, "step": 9689 }, { "epoch": 1.232667599542043, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.657965898513794, "learning_rate": 1e-05, "loss": 0.4156, "mean_token_accuracy": 0.8725172281265259, "num_tokens": 367999470.0, "step": 9690 }, { "epoch": 1.2327948098206334, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5060508251190186, "learning_rate": 1e-05, "loss": 0.3767, "mean_token_accuracy": 0.8892704844474792, "num_tokens": 368042780.0, "step": 9691 }, { "epoch": 1.232922020099224, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6722257137298584, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.875316858291626, "num_tokens": 368078062.0, "step": 9692 }, { "epoch": 1.2330492303778144, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.6006391048431396, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8708194494247437, "num_tokens": 368121570.0, "step": 9693 }, { "epoch": 1.233176440656405, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6394810676574707, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.8770670890808105, "num_tokens": 368158576.0, "step": 9694 }, { "epoch": 1.2333036509349955, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5266754627227783, "learning_rate": 1e-05, "loss": 0.3794, "mean_token_accuracy": 0.8881125450134277, "num_tokens": 368199939.0, "step": 9695 }, { "epoch": 1.233430861213586, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6351358890533447, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8638694286346436, "num_tokens": 368240156.0, "step": 9696 }, { "epoch": 1.2335580714921766, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5568881034851074, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8522855639457703, "num_tokens": 368270852.0, "step": 9697 }, { "epoch": 1.233685281770767, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6278209686279297, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8649946451187134, "num_tokens": 368311986.0, "step": 9698 }, { "epoch": 1.2338124920493576, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.537885904312134, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8641433715820312, "num_tokens": 368354330.0, "step": 9699 }, { "epoch": 1.2339397023279481, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.633521795272827, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8789513111114502, "num_tokens": 368392712.0, "step": 9700 }, { "epoch": 1.2340669126065387, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.538681745529175, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8630557060241699, "num_tokens": 368430913.0, "step": 9701 }, { "epoch": 1.2341941228851292, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6457555294036865, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8844980001449585, "num_tokens": 368462353.0, "step": 9702 }, { "epoch": 1.2343213331637197, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5247113704681396, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8637949228286743, "num_tokens": 368506061.0, "step": 9703 }, { "epoch": 1.2344485434423103, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6467978954315186, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8773531913757324, "num_tokens": 368539435.0, "step": 9704 }, { "epoch": 1.2345757537209006, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.524456262588501, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.868394672870636, "num_tokens": 368578249.0, "step": 9705 }, { "epoch": 1.234702963999491, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6499836444854736, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8850675821304321, "num_tokens": 368617766.0, "step": 9706 }, { "epoch": 1.2348301742780816, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5197207927703857, "learning_rate": 1e-05, "loss": 0.4003, "mean_token_accuracy": 0.8787463903427124, "num_tokens": 368651458.0, "step": 9707 }, { "epoch": 1.2349573845566721, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6763854026794434, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8533623218536377, "num_tokens": 368688689.0, "step": 9708 }, { "epoch": 1.2350845948352627, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5077099800109863, "learning_rate": 1e-05, "loss": 0.49, "mean_token_accuracy": 0.8500964641571045, "num_tokens": 368731996.0, "step": 9709 }, { "epoch": 1.2352118051138532, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.696218252182007, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8804136514663696, "num_tokens": 368772712.0, "step": 9710 }, { "epoch": 1.2353390153924437, "ewc_loss": 0.0595703125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.4974937438964844, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8749421238899231, "num_tokens": 368816582.0, "step": 9711 }, { "epoch": 1.2354662256710343, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7075655460357666, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8751983642578125, "num_tokens": 368854284.0, "step": 9712 }, { "epoch": 1.2355934359496248, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.4954421520233154, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8729468584060669, "num_tokens": 368891564.0, "step": 9713 }, { "epoch": 1.2357206462282153, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.731337308883667, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8542790412902832, "num_tokens": 368931176.0, "step": 9714 }, { "epoch": 1.2358478565068058, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.500554084777832, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8764957189559937, "num_tokens": 368967730.0, "step": 9715 }, { "epoch": 1.2359750667853961, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7285308837890625, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8737436532974243, "num_tokens": 369009378.0, "step": 9716 }, { "epoch": 1.2361022770639867, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5040283203125, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8655630350112915, "num_tokens": 369045716.0, "step": 9717 }, { "epoch": 1.2362294873425772, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7378008365631104, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8805168867111206, "num_tokens": 369082022.0, "step": 9718 }, { "epoch": 1.2363566976211677, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.499361515045166, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8628166913986206, "num_tokens": 369122456.0, "step": 9719 }, { "epoch": 1.2364839078997583, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.782742977142334, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8583724498748779, "num_tokens": 369156977.0, "step": 9720 }, { "epoch": 1.2366111181783488, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.510314702987671, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8595604300498962, "num_tokens": 369195113.0, "step": 9721 }, { "epoch": 1.2367383284569393, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7485344409942627, "learning_rate": 1e-05, "loss": 0.4061, "mean_token_accuracy": 0.8821345567703247, "num_tokens": 369238650.0, "step": 9722 }, { "epoch": 1.2368655387355298, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.522378444671631, "learning_rate": 1e-05, "loss": 0.4939, "mean_token_accuracy": 0.8513543009757996, "num_tokens": 369275753.0, "step": 9723 }, { "epoch": 1.2369927490141204, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.8006176948547363, "learning_rate": 1e-05, "loss": 0.5052, "mean_token_accuracy": 0.8471271395683289, "num_tokens": 369317364.0, "step": 9724 }, { "epoch": 1.237119959292711, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022602081298828125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5139496326446533, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8748301267623901, "num_tokens": 369355889.0, "step": 9725 }, { "epoch": 1.2372471695713014, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7607243061065674, "learning_rate": 1e-05, "loss": 0.3898, "mean_token_accuracy": 0.8840532898902893, "num_tokens": 369393749.0, "step": 9726 }, { "epoch": 1.237374379849892, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5151419639587402, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8699873685836792, "num_tokens": 369432442.0, "step": 9727 }, { "epoch": 1.2375015901284825, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.779226303100586, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8684879541397095, "num_tokens": 369469513.0, "step": 9728 }, { "epoch": 1.237628800407073, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5300683975219727, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8704472780227661, "num_tokens": 369505720.0, "step": 9729 }, { "epoch": 1.2377560106856633, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.774048328399658, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.878868818283081, "num_tokens": 369535225.0, "step": 9730 }, { "epoch": 1.2378832209642538, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.536807060241699, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8738691806793213, "num_tokens": 369572580.0, "step": 9731 }, { "epoch": 1.2380104312428444, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7481460571289062, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8772996664047241, "num_tokens": 369611566.0, "step": 9732 }, { "epoch": 1.238137641521435, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5304372310638428, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8717999458312988, "num_tokens": 369652541.0, "step": 9733 }, { "epoch": 1.2382648518000254, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7667484283447266, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8589047789573669, "num_tokens": 369686744.0, "step": 9734 }, { "epoch": 1.238392062078616, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.530043125152588, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8799827694892883, "num_tokens": 369725090.0, "step": 9735 }, { "epoch": 1.2385192723572065, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7331619262695312, "learning_rate": 1e-05, "loss": 0.3826, "mean_token_accuracy": 0.8890398740768433, "num_tokens": 369764255.0, "step": 9736 }, { "epoch": 1.238646482635797, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5355825424194336, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8763399124145508, "num_tokens": 369803589.0, "step": 9737 }, { "epoch": 1.2387736929143875, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.733168125152588, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8828538656234741, "num_tokens": 369837984.0, "step": 9738 }, { "epoch": 1.238900903192978, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5229101181030273, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8710217475891113, "num_tokens": 369881398.0, "step": 9739 }, { "epoch": 1.2390281134715684, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.738211154937744, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8696520328521729, "num_tokens": 369915230.0, "step": 9740 }, { "epoch": 1.239155323750159, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5181853771209717, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8710389137268066, "num_tokens": 369955237.0, "step": 9741 }, { "epoch": 1.2392825340287494, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7265777587890625, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.86681067943573, "num_tokens": 369995938.0, "step": 9742 }, { "epoch": 1.23940974430734, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.52239990234375, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8627582788467407, "num_tokens": 370037346.0, "step": 9743 }, { "epoch": 1.2395369545859305, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.740401029586792, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8701295852661133, "num_tokens": 370067222.0, "step": 9744 }, { "epoch": 1.239664164864521, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5227909088134766, "learning_rate": 1e-05, "loss": 0.4012, "mean_token_accuracy": 0.8820315003395081, "num_tokens": 370101438.0, "step": 9745 }, { "epoch": 1.2397913751431116, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7325077056884766, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8682594895362854, "num_tokens": 370136887.0, "step": 9746 }, { "epoch": 1.239918585421702, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5166923999786377, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8799986839294434, "num_tokens": 370176797.0, "step": 9747 }, { "epoch": 1.2400457957002926, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7601265907287598, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8643739819526672, "num_tokens": 370206246.0, "step": 9748 }, { "epoch": 1.2401730059788831, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5221774578094482, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8705790638923645, "num_tokens": 370246132.0, "step": 9749 }, { "epoch": 1.2403002162574737, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7482573986053467, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8760536313056946, "num_tokens": 370280039.0, "step": 9750 }, { "epoch": 1.2404274265360642, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5224690437316895, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8759424686431885, "num_tokens": 370317522.0, "step": 9751 }, { "epoch": 1.2405546368146547, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.733214855194092, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8739579916000366, "num_tokens": 370352350.0, "step": 9752 }, { "epoch": 1.2406818470932452, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5450501441955566, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8818761706352234, "num_tokens": 370382645.0, "step": 9753 }, { "epoch": 1.2408090573718356, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.720407485961914, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8743981122970581, "num_tokens": 370422679.0, "step": 9754 }, { "epoch": 1.240936267650426, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.537092447280884, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.867812991142273, "num_tokens": 370458475.0, "step": 9755 }, { "epoch": 1.2410634779290166, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7241437435150146, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8668684363365173, "num_tokens": 370489641.0, "step": 9756 }, { "epoch": 1.2411906882076071, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5405941009521484, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8705687522888184, "num_tokens": 370526649.0, "step": 9757 }, { "epoch": 1.2413178984861977, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.711737632751465, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8663400411605835, "num_tokens": 370564664.0, "step": 9758 }, { "epoch": 1.2414451087647882, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5435380935668945, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8812113404273987, "num_tokens": 370598000.0, "step": 9759 }, { "epoch": 1.2415723190433787, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7074856758117676, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8599063158035278, "num_tokens": 370639318.0, "step": 9760 }, { "epoch": 1.2416995293219693, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5472466945648193, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8607052564620972, "num_tokens": 370676064.0, "step": 9761 }, { "epoch": 1.2418267396005598, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6889524459838867, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8750386238098145, "num_tokens": 370713190.0, "step": 9762 }, { "epoch": 1.2419539498791503, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5350821018218994, "learning_rate": 1e-05, "loss": 0.3994, "mean_token_accuracy": 0.8843445181846619, "num_tokens": 370748433.0, "step": 9763 }, { "epoch": 1.2420811601577408, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6834332942962646, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8775069713592529, "num_tokens": 370781522.0, "step": 9764 }, { "epoch": 1.2422083704363311, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5277020931243896, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8736704587936401, "num_tokens": 370821878.0, "step": 9765 }, { "epoch": 1.2423355807149217, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6898348331451416, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8669684529304504, "num_tokens": 370857754.0, "step": 9766 }, { "epoch": 1.2424627909935122, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5299854278564453, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8663919568061829, "num_tokens": 370893672.0, "step": 9767 }, { "epoch": 1.2425900012721027, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7470524311065674, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8663396835327148, "num_tokens": 370928118.0, "step": 9768 }, { "epoch": 1.2427172115506933, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00022792816162109375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5357561111450195, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8536442518234253, "num_tokens": 370960748.0, "step": 9769 }, { "epoch": 1.2428444218292838, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.707990884780884, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8829623460769653, "num_tokens": 370992408.0, "step": 9770 }, { "epoch": 1.2429716321078743, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.506763458251953, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8557834625244141, "num_tokens": 371035020.0, "step": 9771 }, { "epoch": 1.2430988423864648, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7139599323272705, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8777607679367065, "num_tokens": 371072738.0, "step": 9772 }, { "epoch": 1.2432260526650554, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5037715435028076, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8781445026397705, "num_tokens": 371109246.0, "step": 9773 }, { "epoch": 1.243353262943646, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.718229055404663, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8805965185165405, "num_tokens": 371148322.0, "step": 9774 }, { "epoch": 1.2434804732222364, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5206689834594727, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8598207235336304, "num_tokens": 371186534.0, "step": 9775 }, { "epoch": 1.243607683500827, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.719942331314087, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8709278106689453, "num_tokens": 371224973.0, "step": 9776 }, { "epoch": 1.2437348937794175, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.519442319869995, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8716764450073242, "num_tokens": 371262629.0, "step": 9777 }, { "epoch": 1.243862104058008, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.729957342147827, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8691663146018982, "num_tokens": 371298421.0, "step": 9778 }, { "epoch": 1.2439893143365983, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5234076976776123, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8470922708511353, "num_tokens": 371337393.0, "step": 9779 }, { "epoch": 1.2441165246151888, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.722914457321167, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8749185800552368, "num_tokens": 371372436.0, "step": 9780 }, { "epoch": 1.2442437348937794, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.535323143005371, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8742243051528931, "num_tokens": 371404269.0, "step": 9781 }, { "epoch": 1.24437094517237, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.704343557357788, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8727469444274902, "num_tokens": 371446494.0, "step": 9782 }, { "epoch": 1.2444981554509604, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5358030796051025, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8636474609375, "num_tokens": 371482524.0, "step": 9783 }, { "epoch": 1.244625365729551, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7199718952178955, "learning_rate": 1e-05, "loss": 0.5179, "mean_token_accuracy": 0.8506327271461487, "num_tokens": 371519745.0, "step": 9784 }, { "epoch": 1.2447525760081415, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.518982410430908, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8746557235717773, "num_tokens": 371563593.0, "step": 9785 }, { "epoch": 1.244879786286732, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.718424081802368, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8769717812538147, "num_tokens": 371600646.0, "step": 9786 }, { "epoch": 1.2450069965653225, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5214898586273193, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8794212341308594, "num_tokens": 371635071.0, "step": 9787 }, { "epoch": 1.245134206843913, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7209365367889404, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.870917797088623, "num_tokens": 371672441.0, "step": 9788 }, { "epoch": 1.2452614171225034, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5067801475524902, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8768029808998108, "num_tokens": 371712751.0, "step": 9789 }, { "epoch": 1.245388627401094, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7368788719177246, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8734850883483887, "num_tokens": 371748863.0, "step": 9790 }, { "epoch": 1.2455158376796844, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.527437686920166, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8610376715660095, "num_tokens": 371784763.0, "step": 9791 }, { "epoch": 1.245643047958275, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.742640972137451, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8495373725891113, "num_tokens": 371822722.0, "step": 9792 }, { "epoch": 1.2457702582368655, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5091240406036377, "learning_rate": 1e-05, "loss": 0.37, "mean_token_accuracy": 0.891122043132782, "num_tokens": 371855423.0, "step": 9793 }, { "epoch": 1.245897468515456, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7384355068206787, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.865807056427002, "num_tokens": 371897900.0, "step": 9794 }, { "epoch": 1.2460246787940465, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.502506732940674, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8737319707870483, "num_tokens": 371942394.0, "step": 9795 }, { "epoch": 1.246151889072637, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.757389545440674, "learning_rate": 1e-05, "loss": 0.502, "mean_token_accuracy": 0.8500477075576782, "num_tokens": 371982391.0, "step": 9796 }, { "epoch": 1.2462790993512276, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.510373830795288, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8709369897842407, "num_tokens": 372018599.0, "step": 9797 }, { "epoch": 1.2464063096298181, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7584946155548096, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8660430908203125, "num_tokens": 372058650.0, "step": 9798 }, { "epoch": 1.2465335199084087, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.50982928276062, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.866697371006012, "num_tokens": 372102272.0, "step": 9799 }, { "epoch": 1.2466607301869992, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7510082721710205, "learning_rate": 1e-05, "loss": 0.383, "mean_token_accuracy": 0.8892905116081238, "num_tokens": 372139086.0, "step": 9800 }, { "epoch": 1.2467879404655897, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.509401559829712, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.872834324836731, "num_tokens": 372178418.0, "step": 9801 }, { "epoch": 1.2469151507441802, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.75757098197937, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8823279142379761, "num_tokens": 372218837.0, "step": 9802 }, { "epoch": 1.2470423610227706, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5507314205169678, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8568370938301086, "num_tokens": 372256620.0, "step": 9803 }, { "epoch": 1.247169571301361, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.750009298324585, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8655145764350891, "num_tokens": 372292379.0, "step": 9804 }, { "epoch": 1.2472967815799516, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.51313853263855, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8800150752067566, "num_tokens": 372336319.0, "step": 9805 }, { "epoch": 1.2474239918585421, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.756983995437622, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8648045063018799, "num_tokens": 372372965.0, "step": 9806 }, { "epoch": 1.2475512021371327, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5196330547332764, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8737544417381287, "num_tokens": 372415453.0, "step": 9807 }, { "epoch": 1.2476784124157232, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7575933933258057, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8663326501846313, "num_tokens": 372452822.0, "step": 9808 }, { "epoch": 1.2478056226943137, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.514927387237549, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8689531683921814, "num_tokens": 372494316.0, "step": 9809 }, { "epoch": 1.2479328329729042, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7562248706817627, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8809242844581604, "num_tokens": 372531334.0, "step": 9810 }, { "epoch": 1.2480600432514948, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.529935359954834, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8787810802459717, "num_tokens": 372568303.0, "step": 9811 }, { "epoch": 1.2481872535300853, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7500908374786377, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8815903067588806, "num_tokens": 372602783.0, "step": 9812 }, { "epoch": 1.2483144638086758, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5251805782318115, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8734734058380127, "num_tokens": 372643883.0, "step": 9813 }, { "epoch": 1.2484416740872661, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.749645471572876, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8675764799118042, "num_tokens": 372681864.0, "step": 9814 }, { "epoch": 1.2485688843658567, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.542968511581421, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8673239946365356, "num_tokens": 372717657.0, "step": 9815 }, { "epoch": 1.2486960946444472, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.727381706237793, "learning_rate": 1e-05, "loss": 0.3874, "mean_token_accuracy": 0.8871018886566162, "num_tokens": 372756702.0, "step": 9816 }, { "epoch": 1.2488233049230377, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.53304386138916, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8680043816566467, "num_tokens": 372795620.0, "step": 9817 }, { "epoch": 1.2489505152016283, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.725987672805786, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8656885623931885, "num_tokens": 372833803.0, "step": 9818 }, { "epoch": 1.2490777254802188, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5496878623962402, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.8768095970153809, "num_tokens": 372864229.0, "step": 9819 }, { "epoch": 1.2492049357588093, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7066054344177246, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8729875087738037, "num_tokens": 372900321.0, "step": 9820 }, { "epoch": 1.2493321460373998, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.527273654937744, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8701767921447754, "num_tokens": 372941237.0, "step": 9821 }, { "epoch": 1.2494593563159904, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.699618339538574, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8666916489601135, "num_tokens": 372980057.0, "step": 9822 }, { "epoch": 1.249586566594581, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.54296612739563, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8718273639678955, "num_tokens": 373014746.0, "step": 9823 }, { "epoch": 1.2497137768731714, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6991934776306152, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8622983694076538, "num_tokens": 373053298.0, "step": 9824 }, { "epoch": 1.249840987151762, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5145044326782227, "learning_rate": 1e-05, "loss": 0.385, "mean_token_accuracy": 0.8862032294273376, "num_tokens": 373100155.0, "step": 9825 }, { "epoch": 1.2499681974303525, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.709282159805298, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8648436069488525, "num_tokens": 373138554.0, "step": 9826 }, { "epoch": 1.250095407708943, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5405311584472656, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8584998250007629, "num_tokens": 373173023.0, "step": 9827 }, { "epoch": 1.2502226179875333, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7010092735290527, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8663477301597595, "num_tokens": 373210630.0, "step": 9828 }, { "epoch": 1.2503498282661238, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5309524536132812, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8628005981445312, "num_tokens": 373247859.0, "step": 9829 }, { "epoch": 1.2504770385447144, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.684434652328491, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8726148009300232, "num_tokens": 373285663.0, "step": 9830 }, { "epoch": 1.250604248823305, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5366523265838623, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.863852858543396, "num_tokens": 373321743.0, "step": 9831 }, { "epoch": 1.2507314591018954, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7025961875915527, "learning_rate": 1e-05, "loss": 0.5077, "mean_token_accuracy": 0.8479069471359253, "num_tokens": 373355768.0, "step": 9832 }, { "epoch": 1.250858669380486, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5199875831604004, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8669706583023071, "num_tokens": 373399335.0, "step": 9833 }, { "epoch": 1.2509858796590765, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.714111328125, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8640872240066528, "num_tokens": 373433708.0, "step": 9834 }, { "epoch": 1.251113089937667, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.521758556365967, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8751858472824097, "num_tokens": 373472145.0, "step": 9835 }, { "epoch": 1.2512403002162575, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7467801570892334, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8759543895721436, "num_tokens": 373506578.0, "step": 9836 }, { "epoch": 1.2513675104948478, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.4951741695404053, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8791230916976929, "num_tokens": 373545988.0, "step": 9837 }, { "epoch": 1.2514947207734384, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.756047010421753, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8755747079849243, "num_tokens": 373581001.0, "step": 9838 }, { "epoch": 1.251621931052029, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5207879543304443, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8661442995071411, "num_tokens": 373614961.0, "step": 9839 }, { "epoch": 1.2517491413306194, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002288818359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7473156452178955, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.8837025761604309, "num_tokens": 373652385.0, "step": 9840 }, { "epoch": 1.25187635160921, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5145070552825928, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8657522201538086, "num_tokens": 373692468.0, "step": 9841 }, { "epoch": 1.2520035618878005, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.742635488510132, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8749881982803345, "num_tokens": 373736885.0, "step": 9842 }, { "epoch": 1.252130772166391, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.521733283996582, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8653337955474854, "num_tokens": 373777354.0, "step": 9843 }, { "epoch": 1.2522579824449815, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.771690607070923, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8619507551193237, "num_tokens": 373812400.0, "step": 9844 }, { "epoch": 1.252385192723572, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.520643472671509, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.8848655223846436, "num_tokens": 373846033.0, "step": 9845 }, { "epoch": 1.2525124030021626, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7483763694763184, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8790212273597717, "num_tokens": 373887888.0, "step": 9846 }, { "epoch": 1.2526396132807531, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5157017707824707, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8834748864173889, "num_tokens": 373928082.0, "step": 9847 }, { "epoch": 1.2527668235593437, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7575583457946777, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8606436848640442, "num_tokens": 373964078.0, "step": 9848 }, { "epoch": 1.2528940338379342, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5430095195770264, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8529353141784668, "num_tokens": 373998163.0, "step": 9849 }, { "epoch": 1.2530212441165247, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.741579532623291, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8688719868659973, "num_tokens": 374036322.0, "step": 9850 }, { "epoch": 1.2531484543951152, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5234460830688477, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8664390444755554, "num_tokens": 374079144.0, "step": 9851 }, { "epoch": 1.2532756646737058, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.745398759841919, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8689631819725037, "num_tokens": 374116200.0, "step": 9852 }, { "epoch": 1.253402874952296, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.545535087585449, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.86585533618927, "num_tokens": 374151009.0, "step": 9853 }, { "epoch": 1.2535300852308866, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.730259418487549, "learning_rate": 1e-05, "loss": 0.5146, "mean_token_accuracy": 0.847367525100708, "num_tokens": 374187974.0, "step": 9854 }, { "epoch": 1.2536572955094771, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002307891845703125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5366148948669434, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8598559498786926, "num_tokens": 374229203.0, "step": 9855 }, { "epoch": 1.2537845057880677, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7267565727233887, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8807666301727295, "num_tokens": 374261606.0, "step": 9856 }, { "epoch": 1.2539117160666582, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5395429134368896, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8634220361709595, "num_tokens": 374300274.0, "step": 9857 }, { "epoch": 1.2540389263452487, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7330684661865234, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.848747968673706, "num_tokens": 374336942.0, "step": 9858 }, { "epoch": 1.2541661366238392, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5513741970062256, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8616144061088562, "num_tokens": 374370213.0, "step": 9859 }, { "epoch": 1.2542933469024298, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6976823806762695, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8617205023765564, "num_tokens": 374410653.0, "step": 9860 }, { "epoch": 1.2544205571810203, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5317625999450684, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8567776679992676, "num_tokens": 374454449.0, "step": 9861 }, { "epoch": 1.2545477674596106, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6824138164520264, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8743653297424316, "num_tokens": 374497494.0, "step": 9862 }, { "epoch": 1.2546749777382011, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.536463737487793, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8754537105560303, "num_tokens": 374534739.0, "step": 9863 }, { "epoch": 1.2548021880167917, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.689469337463379, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8644386529922485, "num_tokens": 374573609.0, "step": 9864 }, { "epoch": 1.2549293982953822, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.541517496109009, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8765605688095093, "num_tokens": 374610269.0, "step": 9865 }, { "epoch": 1.2550566085739727, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6882152557373047, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8655418157577515, "num_tokens": 374644888.0, "step": 9866 }, { "epoch": 1.2551838188525632, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.527129888534546, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.871868908405304, "num_tokens": 374685845.0, "step": 9867 }, { "epoch": 1.2553110291311538, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6853103637695312, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.87346351146698, "num_tokens": 374723501.0, "step": 9868 }, { "epoch": 1.2554382394097443, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.536703586578369, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8643030524253845, "num_tokens": 374763082.0, "step": 9869 }, { "epoch": 1.2555654496883348, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.667187213897705, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8768386840820312, "num_tokens": 374802829.0, "step": 9870 }, { "epoch": 1.2556926599669254, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.525176763534546, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8768560886383057, "num_tokens": 374838802.0, "step": 9871 }, { "epoch": 1.255819870245516, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6679115295410156, "learning_rate": 1e-05, "loss": 0.3817, "mean_token_accuracy": 0.8897271752357483, "num_tokens": 374876831.0, "step": 9872 }, { "epoch": 1.2559470805241064, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.51669979095459, "learning_rate": 1e-05, "loss": 0.3819, "mean_token_accuracy": 0.8865708708763123, "num_tokens": 374918120.0, "step": 9873 }, { "epoch": 1.256074290802697, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.67767596244812, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8653827905654907, "num_tokens": 374958247.0, "step": 9874 }, { "epoch": 1.2562015010812875, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5287654399871826, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8652685284614563, "num_tokens": 374998461.0, "step": 9875 }, { "epoch": 1.256328711359878, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.686601161956787, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8784321546554565, "num_tokens": 375029813.0, "step": 9876 }, { "epoch": 1.2564559216384683, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5193841457366943, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8756210803985596, "num_tokens": 375070724.0, "step": 9877 }, { "epoch": 1.2565831319170588, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6754965782165527, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8685034513473511, "num_tokens": 375110634.0, "step": 9878 }, { "epoch": 1.2567103421956494, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.514191150665283, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8780868053436279, "num_tokens": 375150054.0, "step": 9879 }, { "epoch": 1.25683755247424, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.706244707107544, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.860459566116333, "num_tokens": 375185406.0, "step": 9880 }, { "epoch": 1.2569647627528304, "ewc_loss": 0.060302734375, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.504025936126709, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.876895546913147, "num_tokens": 375227544.0, "step": 9881 }, { "epoch": 1.257091973031421, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.704843759536743, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8757502436637878, "num_tokens": 375263523.0, "step": 9882 }, { "epoch": 1.2572191833100115, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.507713556289673, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8778322339057922, "num_tokens": 375304764.0, "step": 9883 }, { "epoch": 1.257346393588602, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.713441848754883, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8759429454803467, "num_tokens": 375340989.0, "step": 9884 }, { "epoch": 1.2574736038671925, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5122668743133545, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8691381812095642, "num_tokens": 375374987.0, "step": 9885 }, { "epoch": 1.2576008141457828, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7176926136016846, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.86420738697052, "num_tokens": 375415847.0, "step": 9886 }, { "epoch": 1.2577280244243734, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5057919025421143, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.8836308121681213, "num_tokens": 375455761.0, "step": 9887 }, { "epoch": 1.257855234702964, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.713573932647705, "learning_rate": 1e-05, "loss": 0.3891, "mean_token_accuracy": 0.8877220153808594, "num_tokens": 375493774.0, "step": 9888 }, { "epoch": 1.2579824449815544, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.511012315750122, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8642520308494568, "num_tokens": 375531947.0, "step": 9889 }, { "epoch": 1.258109655260145, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.714348554611206, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8811588287353516, "num_tokens": 375573546.0, "step": 9890 }, { "epoch": 1.2582368655387355, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.517211437225342, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.849478542804718, "num_tokens": 375617351.0, "step": 9891 }, { "epoch": 1.258364075817326, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7313456535339355, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8701030611991882, "num_tokens": 375653939.0, "step": 9892 }, { "epoch": 1.2584912860959165, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.510430335998535, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8634840250015259, "num_tokens": 375695345.0, "step": 9893 }, { "epoch": 1.258618496374507, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.719484329223633, "learning_rate": 1e-05, "loss": 0.3839, "mean_token_accuracy": 0.8892569541931152, "num_tokens": 375731786.0, "step": 9894 }, { "epoch": 1.2587457066530976, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5158679485321045, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8707414269447327, "num_tokens": 375773258.0, "step": 9895 }, { "epoch": 1.2588729169316881, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.727499485015869, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8841466307640076, "num_tokens": 375805712.0, "step": 9896 }, { "epoch": 1.2590001272102787, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.537179470062256, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8687781095504761, "num_tokens": 375835927.0, "step": 9897 }, { "epoch": 1.2591273374888692, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7208824157714844, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8797519207000732, "num_tokens": 375870568.0, "step": 9898 }, { "epoch": 1.2592545477674597, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5319645404815674, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8706939220428467, "num_tokens": 375904941.0, "step": 9899 }, { "epoch": 1.2593817580460502, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7171764373779297, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.853460431098938, "num_tokens": 375941193.0, "step": 9900 }, { "epoch": 1.2595089683246408, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.523345947265625, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8767414093017578, "num_tokens": 375977952.0, "step": 9901 }, { "epoch": 1.259636178603231, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7176671028137207, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.860925555229187, "num_tokens": 376016711.0, "step": 9902 }, { "epoch": 1.2597633888818216, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.522282123565674, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8597087860107422, "num_tokens": 376061753.0, "step": 9903 }, { "epoch": 1.2598905991604121, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.719924211502075, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8618662357330322, "num_tokens": 376100148.0, "step": 9904 }, { "epoch": 1.2600178094390027, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5155975818634033, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8743475079536438, "num_tokens": 376140663.0, "step": 9905 }, { "epoch": 1.2601450197175932, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.728533983230591, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.852153480052948, "num_tokens": 376186702.0, "step": 9906 }, { "epoch": 1.2602722299961837, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.524536609649658, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8707784414291382, "num_tokens": 376222736.0, "step": 9907 }, { "epoch": 1.2603994402747742, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.737990140914917, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8573731184005737, "num_tokens": 376258871.0, "step": 9908 }, { "epoch": 1.2605266505533648, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5229358673095703, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8692778944969177, "num_tokens": 376295486.0, "step": 9909 }, { "epoch": 1.2606538608319553, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7238287925720215, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8745388984680176, "num_tokens": 376333138.0, "step": 9910 }, { "epoch": 1.2607810711105456, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5275793075561523, "learning_rate": 1e-05, "loss": 0.3998, "mean_token_accuracy": 0.8805210590362549, "num_tokens": 376367284.0, "step": 9911 }, { "epoch": 1.2609082813891361, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7510199546813965, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8591252565383911, "num_tokens": 376398318.0, "step": 9912 }, { "epoch": 1.2610354916677267, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5359132289886475, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8788719177246094, "num_tokens": 376430599.0, "step": 9913 }, { "epoch": 1.2611627019463172, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7310359477996826, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8724684715270996, "num_tokens": 376467238.0, "step": 9914 }, { "epoch": 1.2612899122249077, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5322341918945312, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8645428419113159, "num_tokens": 376505728.0, "step": 9915 }, { "epoch": 1.2614171225034982, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7214226722717285, "learning_rate": 1e-05, "loss": 0.5034, "mean_token_accuracy": 0.8520035743713379, "num_tokens": 376550387.0, "step": 9916 }, { "epoch": 1.2615443327820888, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.533452033996582, "learning_rate": 1e-05, "loss": 0.3837, "mean_token_accuracy": 0.8878716230392456, "num_tokens": 376583365.0, "step": 9917 }, { "epoch": 1.2616715430606793, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7118802070617676, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8788120746612549, "num_tokens": 376620625.0, "step": 9918 }, { "epoch": 1.2617987533392698, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5352659225463867, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8740770220756531, "num_tokens": 376657048.0, "step": 9919 }, { "epoch": 1.2619259636178604, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7237210273742676, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8649568557739258, "num_tokens": 376697000.0, "step": 9920 }, { "epoch": 1.2620531738964509, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5216119289398193, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8706962466239929, "num_tokens": 376739306.0, "step": 9921 }, { "epoch": 1.2621803841750414, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7200639247894287, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8773655891418457, "num_tokens": 376777136.0, "step": 9922 }, { "epoch": 1.262307594453632, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5112643241882324, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8799278736114502, "num_tokens": 376823875.0, "step": 9923 }, { "epoch": 1.2624348047322225, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.73691725730896, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.868067741394043, "num_tokens": 376861335.0, "step": 9924 }, { "epoch": 1.262562015010813, "ewc_loss": 0.060791015625, "ewc_loss_diag": 0.00023174285888671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.542534351348877, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8575155735015869, "num_tokens": 376896076.0, "step": 9925 }, { "epoch": 1.2626892252894033, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.715028762817383, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8615692257881165, "num_tokens": 376943761.0, "step": 9926 }, { "epoch": 1.2628164355679938, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5259509086608887, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.865837037563324, "num_tokens": 376983526.0, "step": 9927 }, { "epoch": 1.2629436458465844, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7067160606384277, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.8842836618423462, "num_tokens": 377025293.0, "step": 9928 }, { "epoch": 1.263070856125175, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.513843536376953, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8699121475219727, "num_tokens": 377071380.0, "step": 9929 }, { "epoch": 1.2631980664037654, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7208127975463867, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8606258630752563, "num_tokens": 377110215.0, "step": 9930 }, { "epoch": 1.263325276682356, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5168843269348145, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.8821309804916382, "num_tokens": 377150191.0, "step": 9931 }, { "epoch": 1.2634524869609465, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7523353099823, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8502882719039917, "num_tokens": 377190636.0, "step": 9932 }, { "epoch": 1.263579697239537, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5051321983337402, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8702760338783264, "num_tokens": 377230314.0, "step": 9933 }, { "epoch": 1.2637069075181275, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7423996925354004, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8651005625724792, "num_tokens": 377270830.0, "step": 9934 }, { "epoch": 1.2638341177967178, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.512188673019409, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8523941040039062, "num_tokens": 377313862.0, "step": 9935 }, { "epoch": 1.2639613280753084, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7454326152801514, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8684173226356506, "num_tokens": 377355451.0, "step": 9936 }, { "epoch": 1.264088538353899, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5119431018829346, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8659423589706421, "num_tokens": 377397608.0, "step": 9937 }, { "epoch": 1.2642157486324894, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7595694065093994, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.871087908744812, "num_tokens": 377436268.0, "step": 9938 }, { "epoch": 1.26434295891108, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5397589206695557, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8689205646514893, "num_tokens": 377470896.0, "step": 9939 }, { "epoch": 1.2644701691896705, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7345399856567383, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8769282102584839, "num_tokens": 377508828.0, "step": 9940 }, { "epoch": 1.264597379468261, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.556372880935669, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8557611703872681, "num_tokens": 377542007.0, "step": 9941 }, { "epoch": 1.2647245897468515, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7115156650543213, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8667411804199219, "num_tokens": 377582671.0, "step": 9942 }, { "epoch": 1.264851800025442, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.531278371810913, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8741328716278076, "num_tokens": 377624840.0, "step": 9943 }, { "epoch": 1.2649790103040326, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7091403007507324, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8682712316513062, "num_tokens": 377661955.0, "step": 9944 }, { "epoch": 1.2651062205826231, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5336084365844727, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8762520551681519, "num_tokens": 377698574.0, "step": 9945 }, { "epoch": 1.2652334308612136, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.701746940612793, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.861396074295044, "num_tokens": 377739874.0, "step": 9946 }, { "epoch": 1.2653606411398042, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5460774898529053, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8539267778396606, "num_tokens": 377775644.0, "step": 9947 }, { "epoch": 1.2654878514183947, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.701204776763916, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8633333444595337, "num_tokens": 377812648.0, "step": 9948 }, { "epoch": 1.2656150616969852, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5408623218536377, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8581346869468689, "num_tokens": 377849928.0, "step": 9949 }, { "epoch": 1.2657422719755758, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6757802963256836, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8759016990661621, "num_tokens": 377891154.0, "step": 9950 }, { "epoch": 1.265869482254166, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.544320583343506, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8672041296958923, "num_tokens": 377926858.0, "step": 9951 }, { "epoch": 1.2659966925327566, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6512155532836914, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8772884607315063, "num_tokens": 377976895.0, "step": 9952 }, { "epoch": 1.2661239028113471, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.55556058883667, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8716029524803162, "num_tokens": 378010300.0, "step": 9953 }, { "epoch": 1.2662511130899377, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6504464149475098, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8702572584152222, "num_tokens": 378051764.0, "step": 9954 }, { "epoch": 1.2663783233685282, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.548146963119507, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8639671206474304, "num_tokens": 378087976.0, "step": 9955 }, { "epoch": 1.2665055336471187, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6505329608917236, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8719914555549622, "num_tokens": 378124062.0, "step": 9956 }, { "epoch": 1.2666327439257092, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.6400415897369385, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8684648275375366, "num_tokens": 378161248.0, "step": 9957 }, { "epoch": 1.2667599542042998, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.604902505874634, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8620501756668091, "num_tokens": 378197959.0, "step": 9958 }, { "epoch": 1.2668871644828903, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5915327072143555, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8641088604927063, "num_tokens": 378233030.0, "step": 9959 }, { "epoch": 1.2670143747614806, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6070826053619385, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8643722534179688, "num_tokens": 378265147.0, "step": 9960 }, { "epoch": 1.2671415850400711, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.593799352645874, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.865628182888031, "num_tokens": 378298759.0, "step": 9961 }, { "epoch": 1.2672687953186617, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.589646100997925, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8526696562767029, "num_tokens": 378338456.0, "step": 9962 }, { "epoch": 1.2673960055972522, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5909698009490967, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8716480135917664, "num_tokens": 378374755.0, "step": 9963 }, { "epoch": 1.2675232158758427, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5650572776794434, "learning_rate": 1e-05, "loss": 0.3878, "mean_token_accuracy": 0.8858407735824585, "num_tokens": 378415302.0, "step": 9964 }, { "epoch": 1.2676504261544332, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.58868670463562, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8655744791030884, "num_tokens": 378454400.0, "step": 9965 }, { "epoch": 1.2677776364330238, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5767881870269775, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8723533153533936, "num_tokens": 378496986.0, "step": 9966 }, { "epoch": 1.2679048467116143, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.588958978652954, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8786584138870239, "num_tokens": 378533937.0, "step": 9967 }, { "epoch": 1.2680320569902048, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5836968421936035, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8775677680969238, "num_tokens": 378567420.0, "step": 9968 }, { "epoch": 1.2681592672687954, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.60097336769104, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8760260343551636, "num_tokens": 378600939.0, "step": 9969 }, { "epoch": 1.2682864775473859, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.610013484954834, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8630540370941162, "num_tokens": 378647858.0, "step": 9970 }, { "epoch": 1.2684136878259764, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5815041065216064, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8685880303382874, "num_tokens": 378686695.0, "step": 9971 }, { "epoch": 1.268540898104567, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.590837240219116, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8743682503700256, "num_tokens": 378732583.0, "step": 9972 }, { "epoch": 1.2686681083831575, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.567415952682495, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8794949054718018, "num_tokens": 378768134.0, "step": 9973 }, { "epoch": 1.268795318661748, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.6062679290771484, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8852277994155884, "num_tokens": 378799684.0, "step": 9974 }, { "epoch": 1.2689225289403383, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.558112144470215, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.863477349281311, "num_tokens": 378837831.0, "step": 9975 }, { "epoch": 1.2690497392189288, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.603091239929199, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8688482642173767, "num_tokens": 378876697.0, "step": 9976 }, { "epoch": 1.2691769494975194, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5497734546661377, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.8826091289520264, "num_tokens": 378916301.0, "step": 9977 }, { "epoch": 1.2693041597761099, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.610870122909546, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8639604449272156, "num_tokens": 378956294.0, "step": 9978 }, { "epoch": 1.2694313700547004, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.560197114944458, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.86305832862854, "num_tokens": 378992704.0, "step": 9979 }, { "epoch": 1.269558580333291, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6047816276550293, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8777268528938293, "num_tokens": 379034955.0, "step": 9980 }, { "epoch": 1.2696857906118815, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.555277109146118, "learning_rate": 1e-05, "loss": 0.4013, "mean_token_accuracy": 0.8806832432746887, "num_tokens": 379075026.0, "step": 9981 }, { "epoch": 1.269813000890472, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6041972637176514, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8686158657073975, "num_tokens": 379113452.0, "step": 9982 }, { "epoch": 1.2699402111690625, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.561500310897827, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.8460087180137634, "num_tokens": 379155632.0, "step": 9983 }, { "epoch": 1.2700674214476528, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.622250556945801, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8779594898223877, "num_tokens": 379188987.0, "step": 9984 }, { "epoch": 1.2701946317262434, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5387051105499268, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8761228322982788, "num_tokens": 379229099.0, "step": 9985 }, { "epoch": 1.270321842004834, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6240687370300293, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8815262317657471, "num_tokens": 379265365.0, "step": 9986 }, { "epoch": 1.2704490522834244, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5399489402770996, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.877813458442688, "num_tokens": 379301170.0, "step": 9987 }, { "epoch": 1.270576262562015, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6399431228637695, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.867817759513855, "num_tokens": 379334983.0, "step": 9988 }, { "epoch": 1.2707034728406055, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5321919918060303, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8594801425933838, "num_tokens": 379371584.0, "step": 9989 }, { "epoch": 1.270830683119196, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6272077560424805, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8716323375701904, "num_tokens": 379415363.0, "step": 9990 }, { "epoch": 1.2709578933977865, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5354087352752686, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8819156885147095, "num_tokens": 379447019.0, "step": 9991 }, { "epoch": 1.271085103676377, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.624647378921509, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8780532479286194, "num_tokens": 379490229.0, "step": 9992 }, { "epoch": 1.2712123139549676, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5381994247436523, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8695656061172485, "num_tokens": 379528608.0, "step": 9993 }, { "epoch": 1.2713395242335581, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.643066883087158, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8616559505462646, "num_tokens": 379566916.0, "step": 9994 }, { "epoch": 1.2714667345121486, "ewc_loss": 0.06005859375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.695487976074219e-05, "grad_norm": 2.5331923961639404, "learning_rate": 1e-05, "loss": 0.3893, "mean_token_accuracy": 0.8847614526748657, "num_tokens": 379600648.0, "step": 9995 }, { "epoch": 1.2715939447907392, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6498982906341553, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8638912439346313, "num_tokens": 379639519.0, "step": 9996 }, { "epoch": 1.2717211550693297, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5272629261016846, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8769991993904114, "num_tokens": 379674426.0, "step": 9997 }, { "epoch": 1.2718483653479202, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6572461128234863, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8777242302894592, "num_tokens": 379713121.0, "step": 9998 }, { "epoch": 1.2719755756265108, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.52247953414917, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8775442242622375, "num_tokens": 379749952.0, "step": 9999 }, { "epoch": 1.272102785905101, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6767332553863525, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8602262735366821, "num_tokens": 379786057.0, "step": 10000 }, { "epoch": 1.2722299961836916, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.536536931991577, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.862528383731842, "num_tokens": 379820034.0, "step": 10001 }, { "epoch": 1.2723572064622821, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6623804569244385, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.874829113483429, "num_tokens": 379863969.0, "step": 10002 }, { "epoch": 1.2724844167408726, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5177109241485596, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8783442974090576, "num_tokens": 379903360.0, "step": 10003 }, { "epoch": 1.2726116270194632, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6886346340179443, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8732409477233887, "num_tokens": 379935874.0, "step": 10004 }, { "epoch": 1.2727388372980537, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.510068416595459, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8662019371986389, "num_tokens": 379972525.0, "step": 10005 }, { "epoch": 1.2728660475766442, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.68473482131958, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8732505440711975, "num_tokens": 380012925.0, "step": 10006 }, { "epoch": 1.2729932578552348, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5226168632507324, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8601714372634888, "num_tokens": 380048235.0, "step": 10007 }, { "epoch": 1.2731204681338253, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.706531286239624, "learning_rate": 1e-05, "loss": 0.5096, "mean_token_accuracy": 0.8500646352767944, "num_tokens": 380084607.0, "step": 10008 }, { "epoch": 1.2732476784124156, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5150365829467773, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8599422574043274, "num_tokens": 380121514.0, "step": 10009 }, { "epoch": 1.2733748886910061, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7077009677886963, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8641273975372314, "num_tokens": 380158640.0, "step": 10010 }, { "epoch": 1.2735020989695967, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5103938579559326, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8692260980606079, "num_tokens": 380196195.0, "step": 10011 }, { "epoch": 1.2736293092481872, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.716947555541992, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8684566617012024, "num_tokens": 380234777.0, "step": 10012 }, { "epoch": 1.2737565195267777, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5007667541503906, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8636795282363892, "num_tokens": 380276494.0, "step": 10013 }, { "epoch": 1.2738837298053682, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.729630708694458, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8683434724807739, "num_tokens": 380314791.0, "step": 10014 }, { "epoch": 1.2740109400839588, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5061028003692627, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8669666051864624, "num_tokens": 380354372.0, "step": 10015 }, { "epoch": 1.2741381503625493, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.742449998855591, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8807331323623657, "num_tokens": 380388022.0, "step": 10016 }, { "epoch": 1.2742653606411398, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.50776743888855, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8627126812934875, "num_tokens": 380426665.0, "step": 10017 }, { "epoch": 1.2743925709197303, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.738478899002075, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8787392377853394, "num_tokens": 380466068.0, "step": 10018 }, { "epoch": 1.2745197811983209, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5031092166900635, "learning_rate": 1e-05, "loss": 0.3902, "mean_token_accuracy": 0.8853932619094849, "num_tokens": 380500370.0, "step": 10019 }, { "epoch": 1.2746469914769114, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7554287910461426, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8710967302322388, "num_tokens": 380538886.0, "step": 10020 }, { "epoch": 1.274774201755502, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5091357231140137, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8717738389968872, "num_tokens": 380577425.0, "step": 10021 }, { "epoch": 1.2749014120340925, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.76648211479187, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8577300310134888, "num_tokens": 380616999.0, "step": 10022 }, { "epoch": 1.275028622312683, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.513761520385742, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8674284219741821, "num_tokens": 380654065.0, "step": 10023 }, { "epoch": 1.2751558325912733, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.775254487991333, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8654799461364746, "num_tokens": 380687947.0, "step": 10024 }, { "epoch": 1.2752830428698638, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5111091136932373, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8775932192802429, "num_tokens": 380728387.0, "step": 10025 }, { "epoch": 1.2754102531484544, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7925007343292236, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8634965419769287, "num_tokens": 380763906.0, "step": 10026 }, { "epoch": 1.2755374634270449, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.51892352104187, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8586196899414062, "num_tokens": 380806904.0, "step": 10027 }, { "epoch": 1.2756646737056354, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.783277750015259, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8791182041168213, "num_tokens": 380842694.0, "step": 10028 }, { "epoch": 1.275791883984226, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5213143825531006, "learning_rate": 1e-05, "loss": 0.4121, "mean_token_accuracy": 0.8797709941864014, "num_tokens": 380879624.0, "step": 10029 }, { "epoch": 1.2759190942628165, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.78416109085083, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.872153103351593, "num_tokens": 380916358.0, "step": 10030 }, { "epoch": 1.276046304541407, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5280768871307373, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8801497220993042, "num_tokens": 380957858.0, "step": 10031 }, { "epoch": 1.2761735148199975, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7688698768615723, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8775898218154907, "num_tokens": 380997301.0, "step": 10032 }, { "epoch": 1.2763007250985878, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5398006439208984, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8778514862060547, "num_tokens": 381034023.0, "step": 10033 }, { "epoch": 1.2764279353771784, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7743844985961914, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8551658391952515, "num_tokens": 381073402.0, "step": 10034 }, { "epoch": 1.2765551456557689, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.532254219055176, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.878197968006134, "num_tokens": 381111182.0, "step": 10035 }, { "epoch": 1.2766823559343594, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7536990642547607, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8712043166160583, "num_tokens": 381153527.0, "step": 10036 }, { "epoch": 1.27680956621295, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.53096866607666, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8686099052429199, "num_tokens": 381196337.0, "step": 10037 }, { "epoch": 1.2769367764915405, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.748878240585327, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8674514293670654, "num_tokens": 381239065.0, "step": 10038 }, { "epoch": 1.277063986770131, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.537929058074951, "learning_rate": 1e-05, "loss": 0.3818, "mean_token_accuracy": 0.8871031403541565, "num_tokens": 381274826.0, "step": 10039 }, { "epoch": 1.2771911970487215, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7379488945007324, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8709976673126221, "num_tokens": 381318043.0, "step": 10040 }, { "epoch": 1.277318407327312, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.554602861404419, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8689751625061035, "num_tokens": 381352583.0, "step": 10041 }, { "epoch": 1.2774456176059026, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.734515428543091, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8620306849479675, "num_tokens": 381387271.0, "step": 10042 }, { "epoch": 1.2775728278844931, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.534834146499634, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8630229830741882, "num_tokens": 381429749.0, "step": 10043 }, { "epoch": 1.2777000381630836, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.723360061645508, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8814710974693298, "num_tokens": 381468763.0, "step": 10044 }, { "epoch": 1.2778272484416742, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5339279174804688, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8745207190513611, "num_tokens": 381505789.0, "step": 10045 }, { "epoch": 1.2779544587202647, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7044453620910645, "learning_rate": 1e-05, "loss": 0.3646, "mean_token_accuracy": 0.8930641412734985, "num_tokens": 381545861.0, "step": 10046 }, { "epoch": 1.2780816689988552, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.550177574157715, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8549175262451172, "num_tokens": 381585373.0, "step": 10047 }, { "epoch": 1.2782088792774458, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6962966918945312, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8747766613960266, "num_tokens": 381623029.0, "step": 10048 }, { "epoch": 1.278336089556036, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.549509286880493, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8628662824630737, "num_tokens": 381663481.0, "step": 10049 }, { "epoch": 1.2784632998346266, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6745359897613525, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8887491822242737, "num_tokens": 381703977.0, "step": 10050 }, { "epoch": 1.2785905101132171, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.543274164199829, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8661282658576965, "num_tokens": 381741409.0, "step": 10051 }, { "epoch": 1.2787177203918076, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.681994915008545, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8673641681671143, "num_tokens": 381779013.0, "step": 10052 }, { "epoch": 1.2788449306703982, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5421953201293945, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8675343990325928, "num_tokens": 381813460.0, "step": 10053 }, { "epoch": 1.2789721409489887, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6746954917907715, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8648633360862732, "num_tokens": 381853821.0, "step": 10054 }, { "epoch": 1.2790993512275792, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.719329833984375e-05, "grad_norm": 2.5397210121154785, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8626854419708252, "num_tokens": 381895411.0, "step": 10055 }, { "epoch": 1.2792265615061698, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.683582067489624, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8632172346115112, "num_tokens": 381929067.0, "step": 10056 }, { "epoch": 1.2793537717847603, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5339081287384033, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8727483749389648, "num_tokens": 381964002.0, "step": 10057 }, { "epoch": 1.2794809820633506, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6745386123657227, "learning_rate": 1e-05, "loss": 0.3817, "mean_token_accuracy": 0.8871403932571411, "num_tokens": 382002913.0, "step": 10058 }, { "epoch": 1.2796081923419411, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5327930450439453, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8794116377830505, "num_tokens": 382039026.0, "step": 10059 }, { "epoch": 1.2797354026205316, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6806654930114746, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8711427450180054, "num_tokens": 382077413.0, "step": 10060 }, { "epoch": 1.2798626128991222, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.538464069366455, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8635165691375732, "num_tokens": 382112686.0, "step": 10061 }, { "epoch": 1.2799898231777127, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6767961978912354, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8670868277549744, "num_tokens": 382152184.0, "step": 10062 }, { "epoch": 1.2801170334563032, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.541961908340454, "learning_rate": 1e-05, "loss": 0.4929, "mean_token_accuracy": 0.8621387481689453, "num_tokens": 382190916.0, "step": 10063 }, { "epoch": 1.2802442437348938, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.697970151901245, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8580802083015442, "num_tokens": 382222986.0, "step": 10064 }, { "epoch": 1.2803714540134843, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5053703784942627, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8797698020935059, "num_tokens": 382266039.0, "step": 10065 }, { "epoch": 1.2804986642920748, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7101848125457764, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8749075531959534, "num_tokens": 382302071.0, "step": 10066 }, { "epoch": 1.2806258745706653, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.508329153060913, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8699103593826294, "num_tokens": 382340240.0, "step": 10067 }, { "epoch": 1.2807530848492559, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.745973587036133, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8738150000572205, "num_tokens": 382368308.0, "step": 10068 }, { "epoch": 1.2808802951278464, "ewc_loss": 0.060546875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.507639169692993, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8760089874267578, "num_tokens": 382406365.0, "step": 10069 }, { "epoch": 1.281007505406437, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.739367723464966, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.860579252243042, "num_tokens": 382449005.0, "step": 10070 }, { "epoch": 1.2811347156850275, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5181310176849365, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8658709526062012, "num_tokens": 382487744.0, "step": 10071 }, { "epoch": 1.281261925963618, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7486960887908936, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8684115409851074, "num_tokens": 382526954.0, "step": 10072 }, { "epoch": 1.2813891362422083, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.511441946029663, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8810739517211914, "num_tokens": 382565257.0, "step": 10073 }, { "epoch": 1.2815163465207988, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7595646381378174, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8664841651916504, "num_tokens": 382601992.0, "step": 10074 }, { "epoch": 1.2816435567993893, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.526536703109741, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8686379194259644, "num_tokens": 382637181.0, "step": 10075 }, { "epoch": 1.2817707670779799, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.748983383178711, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8590652942657471, "num_tokens": 382679886.0, "step": 10076 }, { "epoch": 1.2818979773565704, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5202391147613525, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8640270233154297, "num_tokens": 382717062.0, "step": 10077 }, { "epoch": 1.282025187635161, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.750206470489502, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8673756122589111, "num_tokens": 382752188.0, "step": 10078 }, { "epoch": 1.2821523979137515, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.000232696533203125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5354487895965576, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8695182204246521, "num_tokens": 382785229.0, "step": 10079 }, { "epoch": 1.282279608192342, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7448668479919434, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8675317168235779, "num_tokens": 382823432.0, "step": 10080 }, { "epoch": 1.2824068184709325, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5231053829193115, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8604834079742432, "num_tokens": 382862763.0, "step": 10081 }, { "epoch": 1.2825340287495228, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.752405881881714, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8710702657699585, "num_tokens": 382898312.0, "step": 10082 }, { "epoch": 1.2826612390281134, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.540041923522949, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8710184097290039, "num_tokens": 382933494.0, "step": 10083 }, { "epoch": 1.2827884493067039, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7268898487091064, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8821134567260742, "num_tokens": 382971029.0, "step": 10084 }, { "epoch": 1.2829156595852944, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5614147186279297, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.862470269203186, "num_tokens": 383003911.0, "step": 10085 }, { "epoch": 1.283042869863885, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7086615562438965, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8642646670341492, "num_tokens": 383042945.0, "step": 10086 }, { "epoch": 1.2831700801424755, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5322067737579346, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8732741475105286, "num_tokens": 383085944.0, "step": 10087 }, { "epoch": 1.283297290421066, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.704150676727295, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8651989698410034, "num_tokens": 383124854.0, "step": 10088 }, { "epoch": 1.2834245006996565, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5382657051086426, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8751204013824463, "num_tokens": 383159427.0, "step": 10089 }, { "epoch": 1.283551710978247, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6880102157592773, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8727205991744995, "num_tokens": 383199320.0, "step": 10090 }, { "epoch": 1.2836789212568376, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.541975975036621, "learning_rate": 1e-05, "loss": 0.3878, "mean_token_accuracy": 0.8829931020736694, "num_tokens": 383231367.0, "step": 10091 }, { "epoch": 1.283806131535428, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.686964750289917, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.873228132724762, "num_tokens": 383268238.0, "step": 10092 }, { "epoch": 1.2839333418140186, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5413107872009277, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8769956827163696, "num_tokens": 383303445.0, "step": 10093 }, { "epoch": 1.2840605520926092, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.691920757293701, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8529732823371887, "num_tokens": 383341106.0, "step": 10094 }, { "epoch": 1.2841877623711997, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.530844211578369, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8633863925933838, "num_tokens": 383379225.0, "step": 10095 }, { "epoch": 1.2843149726497902, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6844284534454346, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8745723962783813, "num_tokens": 383420798.0, "step": 10096 }, { "epoch": 1.2844421829283807, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.522895336151123, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8567171096801758, "num_tokens": 383467145.0, "step": 10097 }, { "epoch": 1.284569393206971, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6995246410369873, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8572296500205994, "num_tokens": 383505143.0, "step": 10098 }, { "epoch": 1.2846966034855616, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5237467288970947, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8616695404052734, "num_tokens": 383545987.0, "step": 10099 }, { "epoch": 1.2848238137641521, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7106940746307373, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.874270498752594, "num_tokens": 383578546.0, "step": 10100 }, { "epoch": 1.2849510240427426, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5131773948669434, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.873144805431366, "num_tokens": 383619603.0, "step": 10101 }, { "epoch": 1.2850782343213332, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.717395544052124, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8577479720115662, "num_tokens": 383660716.0, "step": 10102 }, { "epoch": 1.2852054445999237, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.509164571762085, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8743119239807129, "num_tokens": 383704344.0, "step": 10103 }, { "epoch": 1.2853326548785142, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.743544340133667, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8590607047080994, "num_tokens": 383738141.0, "step": 10104 }, { "epoch": 1.2854598651571048, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5156850814819336, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8600161671638489, "num_tokens": 383774682.0, "step": 10105 }, { "epoch": 1.2855870754356953, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.742100477218628, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8786827325820923, "num_tokens": 383814056.0, "step": 10106 }, { "epoch": 1.2857142857142856, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5153489112854004, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8591393232345581, "num_tokens": 383858151.0, "step": 10107 }, { "epoch": 1.2858414959928761, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.751718044281006, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8675167560577393, "num_tokens": 383895121.0, "step": 10108 }, { "epoch": 1.2859687062714666, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.528304100036621, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8559578657150269, "num_tokens": 383929273.0, "step": 10109 }, { "epoch": 1.2860959165500572, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7414867877960205, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8711884021759033, "num_tokens": 383970374.0, "step": 10110 }, { "epoch": 1.2862231268286477, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5166423320770264, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8684911131858826, "num_tokens": 384010375.0, "step": 10111 }, { "epoch": 1.2863503371072382, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7699451446533203, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8626865148544312, "num_tokens": 384040547.0, "step": 10112 }, { "epoch": 1.2864775473858288, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5378360748291016, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8645358085632324, "num_tokens": 384076005.0, "step": 10113 }, { "epoch": 1.2866047576644193, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.74685001373291, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8790132403373718, "num_tokens": 384112568.0, "step": 10114 }, { "epoch": 1.2867319679430098, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5163846015930176, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8785362243652344, "num_tokens": 384151942.0, "step": 10115 }, { "epoch": 1.2868591782216003, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7523484230041504, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8682331442832947, "num_tokens": 384192534.0, "step": 10116 }, { "epoch": 1.2869863885001909, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.526357412338257, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8762251734733582, "num_tokens": 384230600.0, "step": 10117 }, { "epoch": 1.2871135987787814, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.74613094329834, "learning_rate": 1e-05, "loss": 0.3852, "mean_token_accuracy": 0.8886257410049438, "num_tokens": 384266566.0, "step": 10118 }, { "epoch": 1.287240809057372, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.538562059402466, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8621140718460083, "num_tokens": 384304536.0, "step": 10119 }, { "epoch": 1.2873680193359625, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.738281488418579, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8683996200561523, "num_tokens": 384347865.0, "step": 10120 }, { "epoch": 1.287495229614553, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5506815910339355, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8651355504989624, "num_tokens": 384380060.0, "step": 10121 }, { "epoch": 1.2876224398931433, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 3.0014147758483887, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8541108965873718, "num_tokens": 384417574.0, "step": 10122 }, { "epoch": 1.2877496501717338, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5488781929016113, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8741241097450256, "num_tokens": 384454591.0, "step": 10123 }, { "epoch": 1.2878768604503243, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8206143379211426, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8701723217964172, "num_tokens": 384489642.0, "step": 10124 }, { "epoch": 1.2880040707289149, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5198822021484375, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8681069612503052, "num_tokens": 384528993.0, "step": 10125 }, { "epoch": 1.2881312810075054, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.82706356048584, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8677714467048645, "num_tokens": 384569321.0, "step": 10126 }, { "epoch": 1.288258491286096, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00023365020751953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5365090370178223, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8638771772384644, "num_tokens": 384614821.0, "step": 10127 }, { "epoch": 1.2883857015646865, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8042256832122803, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8807268142700195, "num_tokens": 384658557.0, "step": 10128 }, { "epoch": 1.288512911843277, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.565347909927368, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8582921028137207, "num_tokens": 384697121.0, "step": 10129 }, { "epoch": 1.2886401221218675, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7901010513305664, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8525148630142212, "num_tokens": 384735410.0, "step": 10130 }, { "epoch": 1.2887673324004578, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5873401165008545, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8675135374069214, "num_tokens": 384769394.0, "step": 10131 }, { "epoch": 1.2888945426790483, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.749981164932251, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.871403157711029, "num_tokens": 384808845.0, "step": 10132 }, { "epoch": 1.2890217529576389, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.573552131652832, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8698903918266296, "num_tokens": 384848129.0, "step": 10133 }, { "epoch": 1.2891489632362294, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.712982177734375, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8844702243804932, "num_tokens": 384885181.0, "step": 10134 }, { "epoch": 1.28927617351482, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5649847984313965, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8776689171791077, "num_tokens": 384923499.0, "step": 10135 }, { "epoch": 1.2894033837934105, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.70725154876709, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8797799348831177, "num_tokens": 384957367.0, "step": 10136 }, { "epoch": 1.289530594072001, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5530316829681396, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8747361302375793, "num_tokens": 385000409.0, "step": 10137 }, { "epoch": 1.2896578043505915, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6930079460144043, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8648777008056641, "num_tokens": 385038439.0, "step": 10138 }, { "epoch": 1.289785014629182, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.557797431945801, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.87204909324646, "num_tokens": 385073950.0, "step": 10139 }, { "epoch": 1.2899122249077726, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.687734365463257, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8625291585922241, "num_tokens": 385112394.0, "step": 10140 }, { "epoch": 1.290039435186363, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.551862955093384, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8670200109481812, "num_tokens": 385149096.0, "step": 10141 }, { "epoch": 1.2901666454649536, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.693751811981201, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.864473819732666, "num_tokens": 385181986.0, "step": 10142 }, { "epoch": 1.2902938557435442, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5612967014312744, "learning_rate": 1e-05, "loss": 0.5267, "mean_token_accuracy": 0.8425742387771606, "num_tokens": 385220730.0, "step": 10143 }, { "epoch": 1.2904210660221347, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6737518310546875, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.85074782371521, "num_tokens": 385262250.0, "step": 10144 }, { "epoch": 1.2905482763007252, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.542879581451416, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.863161027431488, "num_tokens": 385303518.0, "step": 10145 }, { "epoch": 1.2906754865793157, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6771180629730225, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8676413893699646, "num_tokens": 385342303.0, "step": 10146 }, { "epoch": 1.290802696857906, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.530425786972046, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8737195730209351, "num_tokens": 385381581.0, "step": 10147 }, { "epoch": 1.2909299071364966, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.698788642883301, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8786218762397766, "num_tokens": 385414835.0, "step": 10148 }, { "epoch": 1.291057117415087, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.529289722442627, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8696333765983582, "num_tokens": 385454022.0, "step": 10149 }, { "epoch": 1.2911843276936776, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7050938606262207, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.858285665512085, "num_tokens": 385486479.0, "step": 10150 }, { "epoch": 1.2913115379722682, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.545227527618408, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8608509302139282, "num_tokens": 385527942.0, "step": 10151 }, { "epoch": 1.2914387482508587, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6881182193756104, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8660200238227844, "num_tokens": 385567994.0, "step": 10152 }, { "epoch": 1.2915659585294492, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.518308162689209, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8746742606163025, "num_tokens": 385603908.0, "step": 10153 }, { "epoch": 1.2916931688080397, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.702413558959961, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8686950206756592, "num_tokens": 385641461.0, "step": 10154 }, { "epoch": 1.2918203790866303, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.517608642578125, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.8773177862167358, "num_tokens": 385676677.0, "step": 10155 }, { "epoch": 1.2919475893652206, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.71667742729187, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8690468072891235, "num_tokens": 385715674.0, "step": 10156 }, { "epoch": 1.2920747996438111, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5271217823028564, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8660093545913696, "num_tokens": 385747947.0, "step": 10157 }, { "epoch": 1.2922020099224016, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.714815855026245, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8625025153160095, "num_tokens": 385787395.0, "step": 10158 }, { "epoch": 1.2923292202009922, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.518562078475952, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8803116083145142, "num_tokens": 385821724.0, "step": 10159 }, { "epoch": 1.2924564304795827, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.720130205154419, "learning_rate": 1e-05, "loss": 0.3866, "mean_token_accuracy": 0.8893880248069763, "num_tokens": 385856281.0, "step": 10160 }, { "epoch": 1.2925836407581732, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.52128267288208, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8642520904541016, "num_tokens": 385895979.0, "step": 10161 }, { "epoch": 1.2927108510367638, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.733222007751465, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8784791231155396, "num_tokens": 385933134.0, "step": 10162 }, { "epoch": 1.2928380613153543, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.512315273284912, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8658448457717896, "num_tokens": 385974487.0, "step": 10163 }, { "epoch": 1.2929652715939448, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.743345260620117, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8599628210067749, "num_tokens": 386014594.0, "step": 10164 }, { "epoch": 1.2930924818725353, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.547584056854248, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8628774881362915, "num_tokens": 386044982.0, "step": 10165 }, { "epoch": 1.2932196921511259, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.734248161315918, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8772698044776917, "num_tokens": 386084517.0, "step": 10166 }, { "epoch": 1.2933469024297164, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5257933139801025, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8690563440322876, "num_tokens": 386121841.0, "step": 10167 }, { "epoch": 1.293474112708307, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7276265621185303, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8713634014129639, "num_tokens": 386157426.0, "step": 10168 }, { "epoch": 1.2936013229868975, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.528811454772949, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8649888634681702, "num_tokens": 386196739.0, "step": 10169 }, { "epoch": 1.293728533265488, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7180778980255127, "learning_rate": 1e-05, "loss": 0.3917, "mean_token_accuracy": 0.8846609592437744, "num_tokens": 386234335.0, "step": 10170 }, { "epoch": 1.2938557435440783, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.52441143989563, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8681098222732544, "num_tokens": 386274166.0, "step": 10171 }, { "epoch": 1.2939829538226688, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.725736618041992, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8603621125221252, "num_tokens": 386313709.0, "step": 10172 }, { "epoch": 1.2941101641012593, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5476913452148438, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8516374826431274, "num_tokens": 386345516.0, "step": 10173 }, { "epoch": 1.2942373743798499, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.70828914642334, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8827641606330872, "num_tokens": 386383284.0, "step": 10174 }, { "epoch": 1.2943645846584404, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.537806987762451, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.863547682762146, "num_tokens": 386423141.0, "step": 10175 }, { "epoch": 1.294491794937031, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.696075201034546, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8769415616989136, "num_tokens": 386461888.0, "step": 10176 }, { "epoch": 1.2946190052156215, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5380773544311523, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8674802780151367, "num_tokens": 386500061.0, "step": 10177 }, { "epoch": 1.294746215494212, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6926567554473877, "learning_rate": 1e-05, "loss": 0.3855, "mean_token_accuracy": 0.8881912231445312, "num_tokens": 386538843.0, "step": 10178 }, { "epoch": 1.2948734257728025, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5230636596679688, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.869787871837616, "num_tokens": 386584346.0, "step": 10179 }, { "epoch": 1.2950006360513928, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7040138244628906, "learning_rate": 1e-05, "loss": 0.4015, "mean_token_accuracy": 0.8822805881500244, "num_tokens": 386619574.0, "step": 10180 }, { "epoch": 1.2951278463299833, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5323257446289062, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8715550899505615, "num_tokens": 386655111.0, "step": 10181 }, { "epoch": 1.2952550566085739, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.712514638900757, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8708664178848267, "num_tokens": 386701102.0, "step": 10182 }, { "epoch": 1.2953822668871644, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.0002346038818359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.527815580368042, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8707457780838013, "num_tokens": 386740259.0, "step": 10183 }, { "epoch": 1.295509477165755, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7056283950805664, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8771674633026123, "num_tokens": 386780345.0, "step": 10184 }, { "epoch": 1.2956366874443455, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5317039489746094, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8735024929046631, "num_tokens": 386819401.0, "step": 10185 }, { "epoch": 1.295763897722936, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7097980976104736, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8771422505378723, "num_tokens": 386856617.0, "step": 10186 }, { "epoch": 1.2958911080015265, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5256314277648926, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8589969873428345, "num_tokens": 386895426.0, "step": 10187 }, { "epoch": 1.296018318280117, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7239773273468018, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.865248441696167, "num_tokens": 386931577.0, "step": 10188 }, { "epoch": 1.2961455285587076, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.515191078186035, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8766816854476929, "num_tokens": 386968658.0, "step": 10189 }, { "epoch": 1.296272738837298, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.778416156768799, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.873551607131958, "num_tokens": 387005364.0, "step": 10190 }, { "epoch": 1.2963999491158886, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.496933937072754, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.875145435333252, "num_tokens": 387048767.0, "step": 10191 }, { "epoch": 1.2965271593944792, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7743563652038574, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.881260871887207, "num_tokens": 387083456.0, "step": 10192 }, { "epoch": 1.2966543696730697, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.4926772117614746, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8753753900527954, "num_tokens": 387124220.0, "step": 10193 }, { "epoch": 1.2967815799516602, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7957537174224854, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8599880933761597, "num_tokens": 387163825.0, "step": 10194 }, { "epoch": 1.2969087902302507, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.519742012023926, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8711941838264465, "num_tokens": 387201230.0, "step": 10195 }, { "epoch": 1.297036000508841, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7860140800476074, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.868279218673706, "num_tokens": 387238426.0, "step": 10196 }, { "epoch": 1.2971632107874316, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5331478118896484, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8745169639587402, "num_tokens": 387275883.0, "step": 10197 }, { "epoch": 1.297290421066022, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.768667697906494, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8799580931663513, "num_tokens": 387319053.0, "step": 10198 }, { "epoch": 1.2974176313446126, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5289809703826904, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8794395923614502, "num_tokens": 387355013.0, "step": 10199 }, { "epoch": 1.2975448416232032, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.761935234069824, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8809857368469238, "num_tokens": 387392736.0, "step": 10200 }, { "epoch": 1.2976720519017937, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.546177625656128, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8573154807090759, "num_tokens": 387431938.0, "step": 10201 }, { "epoch": 1.2977992621803842, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7494940757751465, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.870836615562439, "num_tokens": 387472724.0, "step": 10202 }, { "epoch": 1.2979264724589747, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5385329723358154, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8702726364135742, "num_tokens": 387513666.0, "step": 10203 }, { "epoch": 1.2980536827375653, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.749737024307251, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.8775968551635742, "num_tokens": 387544691.0, "step": 10204 }, { "epoch": 1.2981808930161556, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.542592763900757, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8799263834953308, "num_tokens": 387580839.0, "step": 10205 }, { "epoch": 1.298308103294746, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7502923011779785, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8647235631942749, "num_tokens": 387616258.0, "step": 10206 }, { "epoch": 1.2984353135733366, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.541005849838257, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8621856570243835, "num_tokens": 387654058.0, "step": 10207 }, { "epoch": 1.2985625238519272, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.738924980163574, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8644001483917236, "num_tokens": 387690036.0, "step": 10208 }, { "epoch": 1.2986897341305177, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5360050201416016, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8698630928993225, "num_tokens": 387729114.0, "step": 10209 }, { "epoch": 1.2988169444091082, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7374770641326904, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8725364208221436, "num_tokens": 387768225.0, "step": 10210 }, { "epoch": 1.2989441546876987, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5526816844940186, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8616631627082825, "num_tokens": 387802093.0, "step": 10211 }, { "epoch": 1.2990713649662893, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7115495204925537, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8734253644943237, "num_tokens": 387838857.0, "step": 10212 }, { "epoch": 1.2991985752448798, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5434012413024902, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8713585138320923, "num_tokens": 387878974.0, "step": 10213 }, { "epoch": 1.2993257855234703, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7014219760894775, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8628404140472412, "num_tokens": 387920157.0, "step": 10214 }, { "epoch": 1.2994529958020609, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5560812950134277, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8748751282691956, "num_tokens": 387953345.0, "step": 10215 }, { "epoch": 1.2995802060806514, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.689761161804199, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8789458870887756, "num_tokens": 387990696.0, "step": 10216 }, { "epoch": 1.299707416359242, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.547096014022827, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8611724376678467, "num_tokens": 388028142.0, "step": 10217 }, { "epoch": 1.2998346266378324, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.712172031402588, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8455744981765747, "num_tokens": 388058095.0, "step": 10218 }, { "epoch": 1.299961836916423, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5272603034973145, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8691742420196533, "num_tokens": 388097141.0, "step": 10219 }, { "epoch": 1.3000890471950133, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7146337032318115, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8719644546508789, "num_tokens": 388132138.0, "step": 10220 }, { "epoch": 1.3002162574736038, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5175344944000244, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8722939491271973, "num_tokens": 388172837.0, "step": 10221 }, { "epoch": 1.3003434677521943, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.709758996963501, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8594515323638916, "num_tokens": 388214797.0, "step": 10222 }, { "epoch": 1.3004706780307849, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.509040355682373, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8632278442382812, "num_tokens": 388256374.0, "step": 10223 }, { "epoch": 1.3005978883093754, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7272450923919678, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8749892115592957, "num_tokens": 388291227.0, "step": 10224 }, { "epoch": 1.300725098587966, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5045340061187744, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8798929452896118, "num_tokens": 388332963.0, "step": 10225 }, { "epoch": 1.3008523088665565, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7303659915924072, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.8863484859466553, "num_tokens": 388372825.0, "step": 10226 }, { "epoch": 1.300979519145147, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5141191482543945, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8728014230728149, "num_tokens": 388411850.0, "step": 10227 }, { "epoch": 1.3011067294237375, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7472527027130127, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8665947914123535, "num_tokens": 388449509.0, "step": 10228 }, { "epoch": 1.3012339397023278, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.52941632270813, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8640331625938416, "num_tokens": 388483698.0, "step": 10229 }, { "epoch": 1.3013611499809183, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7494566440582275, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8685915470123291, "num_tokens": 388516832.0, "step": 10230 }, { "epoch": 1.3014883602595089, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5252342224121094, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.88094162940979, "num_tokens": 388553272.0, "step": 10231 }, { "epoch": 1.3016155705380994, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7388997077941895, "learning_rate": 1e-05, "loss": 0.3897, "mean_token_accuracy": 0.8872570991516113, "num_tokens": 388589163.0, "step": 10232 }, { "epoch": 1.30174278081669, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5252277851104736, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8613598346710205, "num_tokens": 388628305.0, "step": 10233 }, { "epoch": 1.3018699910952805, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7310831546783447, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8628183007240295, "num_tokens": 388673462.0, "step": 10234 }, { "epoch": 1.301997201373871, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5409963130950928, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8626965284347534, "num_tokens": 388709357.0, "step": 10235 }, { "epoch": 1.3021244116524615, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7157037258148193, "learning_rate": 1e-05, "loss": 0.3908, "mean_token_accuracy": 0.8879409432411194, "num_tokens": 388746550.0, "step": 10236 }, { "epoch": 1.302251621931052, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5281436443328857, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8554864525794983, "num_tokens": 388795139.0, "step": 10237 }, { "epoch": 1.3023788322096426, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.719827175140381, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8753113746643066, "num_tokens": 388832282.0, "step": 10238 }, { "epoch": 1.302506042488233, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5230109691619873, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8797286748886108, "num_tokens": 388873974.0, "step": 10239 }, { "epoch": 1.3026332527668236, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.709202289581299, "learning_rate": 1e-05, "loss": 0.3935, "mean_token_accuracy": 0.8845872282981873, "num_tokens": 388914562.0, "step": 10240 }, { "epoch": 1.3027604630454142, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5406925678253174, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8612288236618042, "num_tokens": 388950293.0, "step": 10241 }, { "epoch": 1.3028876733240047, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7069451808929443, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.870256781578064, "num_tokens": 388986649.0, "step": 10242 }, { "epoch": 1.3030148836025952, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5173418521881104, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8836278915405273, "num_tokens": 389027779.0, "step": 10243 }, { "epoch": 1.3031420938811857, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7201879024505615, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8551653623580933, "num_tokens": 389062760.0, "step": 10244 }, { "epoch": 1.303269304159776, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5291595458984375, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8580058813095093, "num_tokens": 389101892.0, "step": 10245 }, { "epoch": 1.3033965144383666, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.722458600997925, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8720035552978516, "num_tokens": 389140155.0, "step": 10246 }, { "epoch": 1.303523724716957, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.527726173400879, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8628033399581909, "num_tokens": 389181563.0, "step": 10247 }, { "epoch": 1.3036509349955476, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.723848819732666, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8757299780845642, "num_tokens": 389217415.0, "step": 10248 }, { "epoch": 1.3037781452741382, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.519181251525879, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8846702575683594, "num_tokens": 389255889.0, "step": 10249 }, { "epoch": 1.3039053555527287, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7172391414642334, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8767373561859131, "num_tokens": 389297650.0, "step": 10250 }, { "epoch": 1.3040325658313192, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.518662214279175, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.855731725692749, "num_tokens": 389339647.0, "step": 10251 }, { "epoch": 1.3041597761099097, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7382678985595703, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8722909688949585, "num_tokens": 389376135.0, "step": 10252 }, { "epoch": 1.3042869863885003, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.532735586166382, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8615300059318542, "num_tokens": 389410516.0, "step": 10253 }, { "epoch": 1.3044141966670906, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.722247838973999, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.873417317867279, "num_tokens": 389452712.0, "step": 10254 }, { "epoch": 1.304541406945681, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5329182147979736, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8735697269439697, "num_tokens": 389489257.0, "step": 10255 }, { "epoch": 1.3046686172242716, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.739851713180542, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8563629984855652, "num_tokens": 389524344.0, "step": 10256 }, { "epoch": 1.3047958275028622, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5239009857177734, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8521857261657715, "num_tokens": 389564351.0, "step": 10257 }, { "epoch": 1.3049230377814527, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.732867479324341, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8710461854934692, "num_tokens": 389604606.0, "step": 10258 }, { "epoch": 1.3050502480600432, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5327789783477783, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8717305064201355, "num_tokens": 389639600.0, "step": 10259 }, { "epoch": 1.3051774583386337, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.732327461242676, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8595683574676514, "num_tokens": 389677770.0, "step": 10260 }, { "epoch": 1.3053046686172243, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5368173122406006, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8564865589141846, "num_tokens": 389715366.0, "step": 10261 }, { "epoch": 1.3054318788958148, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.723069667816162, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8769041895866394, "num_tokens": 389754745.0, "step": 10262 }, { "epoch": 1.3055590891744053, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.528049945831299, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8612251281738281, "num_tokens": 389798599.0, "step": 10263 }, { "epoch": 1.3056862994529959, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.73272442817688, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8682463765144348, "num_tokens": 389831861.0, "step": 10264 }, { "epoch": 1.3058135097315864, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.518773078918457, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8598223924636841, "num_tokens": 389877785.0, "step": 10265 }, { "epoch": 1.305940720010177, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7256686687469482, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8845661878585815, "num_tokens": 389921969.0, "step": 10266 }, { "epoch": 1.3060679302887674, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.535421848297119, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8739666938781738, "num_tokens": 389959281.0, "step": 10267 }, { "epoch": 1.306195140567358, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.734745740890503, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8840664029121399, "num_tokens": 389992580.0, "step": 10268 }, { "epoch": 1.3063223508459483, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5410940647125244, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.861580491065979, "num_tokens": 390029377.0, "step": 10269 }, { "epoch": 1.3064495611245388, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7294094562530518, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8748936057090759, "num_tokens": 390059679.0, "step": 10270 }, { "epoch": 1.3065767714031293, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.535551071166992, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.869444727897644, "num_tokens": 390096751.0, "step": 10271 }, { "epoch": 1.3067039816817199, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7267205715179443, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8608354330062866, "num_tokens": 390138648.0, "step": 10272 }, { "epoch": 1.3068311919603104, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.538289785385132, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8708271980285645, "num_tokens": 390176008.0, "step": 10273 }, { "epoch": 1.306958402238901, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.704993963241577, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8691304922103882, "num_tokens": 390218120.0, "step": 10274 }, { "epoch": 1.3070856125174914, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5303101539611816, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8750691413879395, "num_tokens": 390264776.0, "step": 10275 }, { "epoch": 1.307212822796082, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.716705799102783, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8796641230583191, "num_tokens": 390298798.0, "step": 10276 }, { "epoch": 1.3073400330746725, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.54719877243042, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.865618884563446, "num_tokens": 390333332.0, "step": 10277 }, { "epoch": 1.3074672433532628, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7049405574798584, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8637263774871826, "num_tokens": 390374128.0, "step": 10278 }, { "epoch": 1.3075944536318533, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023555755615234375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5321593284606934, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.868771493434906, "num_tokens": 390416509.0, "step": 10279 }, { "epoch": 1.3077216639104439, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.739685297012329, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8717453479766846, "num_tokens": 390448129.0, "step": 10280 }, { "epoch": 1.3078488741890344, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5223891735076904, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8695584535598755, "num_tokens": 390489329.0, "step": 10281 }, { "epoch": 1.307976084467625, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7249488830566406, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8718239665031433, "num_tokens": 390529839.0, "step": 10282 }, { "epoch": 1.3081032947462155, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5262038707733154, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8715988397598267, "num_tokens": 390568888.0, "step": 10283 }, { "epoch": 1.308230505024806, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.734006404876709, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8762099742889404, "num_tokens": 390606709.0, "step": 10284 }, { "epoch": 1.3083577153033965, "ewc_loss": 0.061279296875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5215060710906982, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8672712445259094, "num_tokens": 390645027.0, "step": 10285 }, { "epoch": 1.308484925581987, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.73640513420105, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8698291182518005, "num_tokens": 390685923.0, "step": 10286 }, { "epoch": 1.3086121358605776, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5165677070617676, "learning_rate": 1e-05, "loss": 0.3935, "mean_token_accuracy": 0.8861905336380005, "num_tokens": 390728410.0, "step": 10287 }, { "epoch": 1.308739346139168, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7475812435150146, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.86270672082901, "num_tokens": 390764146.0, "step": 10288 }, { "epoch": 1.3088665564177586, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5264875888824463, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8685170412063599, "num_tokens": 390800363.0, "step": 10289 }, { "epoch": 1.3089937666963491, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7417781352996826, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.87185138463974, "num_tokens": 390840213.0, "step": 10290 }, { "epoch": 1.3091209769749397, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5194058418273926, "learning_rate": 1e-05, "loss": 0.3838, "mean_token_accuracy": 0.88575279712677, "num_tokens": 390879087.0, "step": 10291 }, { "epoch": 1.3092481872535302, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.747814416885376, "learning_rate": 1e-05, "loss": 0.3767, "mean_token_accuracy": 0.8879433870315552, "num_tokens": 390914074.0, "step": 10292 }, { "epoch": 1.3093753975321207, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.508415937423706, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8804829120635986, "num_tokens": 390956283.0, "step": 10293 }, { "epoch": 1.309502607810711, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7653636932373047, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8716297149658203, "num_tokens": 390996898.0, "step": 10294 }, { "epoch": 1.3096298180893016, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5189454555511475, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.882576584815979, "num_tokens": 391033280.0, "step": 10295 }, { "epoch": 1.309757028367892, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7532424926757812, "learning_rate": 1e-05, "loss": 0.3863, "mean_token_accuracy": 0.88778156042099, "num_tokens": 391074159.0, "step": 10296 }, { "epoch": 1.3098842386464826, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5189669132232666, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8744888305664062, "num_tokens": 391115649.0, "step": 10297 }, { "epoch": 1.3100114489250732, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7655653953552246, "learning_rate": 1e-05, "loss": 0.3775, "mean_token_accuracy": 0.8925344944000244, "num_tokens": 391147601.0, "step": 10298 }, { "epoch": 1.3101386592036637, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5382165908813477, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.872491717338562, "num_tokens": 391183483.0, "step": 10299 }, { "epoch": 1.3102658694822542, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.762927532196045, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.887313723564148, "num_tokens": 391217346.0, "step": 10300 }, { "epoch": 1.3103930797608447, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.535801887512207, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.879593551158905, "num_tokens": 391255575.0, "step": 10301 }, { "epoch": 1.3105202900394353, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7663915157318115, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8777565360069275, "num_tokens": 391291680.0, "step": 10302 }, { "epoch": 1.3106475003180256, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.535222291946411, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8744113445281982, "num_tokens": 391330034.0, "step": 10303 }, { "epoch": 1.310774710596616, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.754348039627075, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8814252614974976, "num_tokens": 391370558.0, "step": 10304 }, { "epoch": 1.3109019208752066, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5285232067108154, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8692187666893005, "num_tokens": 391414091.0, "step": 10305 }, { "epoch": 1.3110291311537972, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7503418922424316, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8610578775405884, "num_tokens": 391458057.0, "step": 10306 }, { "epoch": 1.3111563414323877, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.544602394104004, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8742762804031372, "num_tokens": 391495419.0, "step": 10307 }, { "epoch": 1.3112835517109782, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.763591766357422, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8696602582931519, "num_tokens": 391529353.0, "step": 10308 }, { "epoch": 1.3114107619895687, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.547219753265381, "learning_rate": 1e-05, "loss": 0.5188, "mean_token_accuracy": 0.8471882343292236, "num_tokens": 391571111.0, "step": 10309 }, { "epoch": 1.3115379722681593, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7399089336395264, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8679301142692566, "num_tokens": 391612284.0, "step": 10310 }, { "epoch": 1.3116651825467498, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.557222604751587, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8638010025024414, "num_tokens": 391646934.0, "step": 10311 }, { "epoch": 1.3117923928253403, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7184298038482666, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8797954320907593, "num_tokens": 391686871.0, "step": 10312 }, { "epoch": 1.3119196031039309, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.544450283050537, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.871683657169342, "num_tokens": 391728188.0, "step": 10313 }, { "epoch": 1.3120468133825214, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.708782434463501, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8644330501556396, "num_tokens": 391767434.0, "step": 10314 }, { "epoch": 1.312174023661112, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023651123046875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5315029621124268, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.877318799495697, "num_tokens": 391809581.0, "step": 10315 }, { "epoch": 1.3123012339397024, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7073814868927, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8661414384841919, "num_tokens": 391849877.0, "step": 10316 }, { "epoch": 1.312428444218293, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5499725341796875, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8607025146484375, "num_tokens": 391885995.0, "step": 10317 }, { "epoch": 1.3125556544968833, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.713986396789551, "learning_rate": 1e-05, "loss": 0.3964, "mean_token_accuracy": 0.8820724487304688, "num_tokens": 391919100.0, "step": 10318 }, { "epoch": 1.3126828647754738, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.551480293273926, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8685991764068604, "num_tokens": 391953246.0, "step": 10319 }, { "epoch": 1.3128100750540643, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.719064235687256, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8767207860946655, "num_tokens": 391986618.0, "step": 10320 }, { "epoch": 1.3129372853326549, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.526381254196167, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.866932213306427, "num_tokens": 392029663.0, "step": 10321 }, { "epoch": 1.3130644956112454, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7268636226654053, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8704701662063599, "num_tokens": 392063329.0, "step": 10322 }, { "epoch": 1.313191705889836, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.519383668899536, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8761113882064819, "num_tokens": 392101400.0, "step": 10323 }, { "epoch": 1.3133189161684264, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.734685182571411, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8673563003540039, "num_tokens": 392140146.0, "step": 10324 }, { "epoch": 1.313446126447017, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5226736068725586, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8805094957351685, "num_tokens": 392176427.0, "step": 10325 }, { "epoch": 1.3135733367256075, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.727740526199341, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8665639758110046, "num_tokens": 392221105.0, "step": 10326 }, { "epoch": 1.3137005470041978, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5362226963043213, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8558971285820007, "num_tokens": 392257301.0, "step": 10327 }, { "epoch": 1.3138277572827883, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7256276607513428, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8741954565048218, "num_tokens": 392295849.0, "step": 10328 }, { "epoch": 1.3139549675613789, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5260956287384033, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.861325740814209, "num_tokens": 392336480.0, "step": 10329 }, { "epoch": 1.3140821778399694, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.8125009536743164, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8638883829116821, "num_tokens": 392372562.0, "step": 10330 }, { "epoch": 1.31420938811856, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.516155958175659, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8825594782829285, "num_tokens": 392408843.0, "step": 10331 }, { "epoch": 1.3143365983971504, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7736494541168213, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8738439083099365, "num_tokens": 392440725.0, "step": 10332 }, { "epoch": 1.314463808675741, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.509371757507324, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8587228655815125, "num_tokens": 392484643.0, "step": 10333 }, { "epoch": 1.3145910189543315, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.783766746520996, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8746858835220337, "num_tokens": 392523849.0, "step": 10334 }, { "epoch": 1.314718229232922, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5295023918151855, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8724745512008667, "num_tokens": 392559598.0, "step": 10335 }, { "epoch": 1.3148454395115126, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7649288177490234, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8528436422348022, "num_tokens": 392603856.0, "step": 10336 }, { "epoch": 1.314972649790103, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5295987129211426, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8802664279937744, "num_tokens": 392645317.0, "step": 10337 }, { "epoch": 1.3150998600686936, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7779080867767334, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8690669536590576, "num_tokens": 392681650.0, "step": 10338 }, { "epoch": 1.3152270703472841, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.553342819213867, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8624714016914368, "num_tokens": 392711588.0, "step": 10339 }, { "epoch": 1.3153542806258747, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.744346857070923, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8760940432548523, "num_tokens": 392758504.0, "step": 10340 }, { "epoch": 1.3154814909044652, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.55623459815979, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8759008646011353, "num_tokens": 392792153.0, "step": 10341 }, { "epoch": 1.3156087011830557, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7054293155670166, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8833032250404358, "num_tokens": 392840971.0, "step": 10342 }, { "epoch": 1.315735911461646, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5584428310394287, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.865692138671875, "num_tokens": 392882760.0, "step": 10343 }, { "epoch": 1.3158631217402366, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7010934352874756, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8587027192115784, "num_tokens": 392922165.0, "step": 10344 }, { "epoch": 1.315990332018827, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5729377269744873, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.862395167350769, "num_tokens": 392957181.0, "step": 10345 }, { "epoch": 1.3161175422974176, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.679954767227173, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8647462725639343, "num_tokens": 392998405.0, "step": 10346 }, { "epoch": 1.3162447525760081, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.555849075317383, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8759445548057556, "num_tokens": 393036437.0, "step": 10347 }, { "epoch": 1.3163719628545987, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6747071743011475, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8774445056915283, "num_tokens": 393070711.0, "step": 10348 }, { "epoch": 1.3164991731331892, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5621368885040283, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8554106950759888, "num_tokens": 393109052.0, "step": 10349 }, { "epoch": 1.3166263834117797, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.673814296722412, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8619983196258545, "num_tokens": 393146244.0, "step": 10350 }, { "epoch": 1.3167535936903703, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.556974411010742, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8749688267707825, "num_tokens": 393180630.0, "step": 10351 }, { "epoch": 1.3168808039689606, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.651066541671753, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8705827593803406, "num_tokens": 393222687.0, "step": 10352 }, { "epoch": 1.317008014247551, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.542102336883545, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8664741516113281, "num_tokens": 393267383.0, "step": 10353 }, { "epoch": 1.3171352245261416, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.669985294342041, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8785606622695923, "num_tokens": 393302055.0, "step": 10354 }, { "epoch": 1.3172624348047322, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5352375507354736, "learning_rate": 1e-05, "loss": 0.3923, "mean_token_accuracy": 0.882635772228241, "num_tokens": 393336863.0, "step": 10355 }, { "epoch": 1.3173896450833227, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6899988651275635, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8499392867088318, "num_tokens": 393373602.0, "step": 10356 }, { "epoch": 1.3175168553619132, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5474350452423096, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8608964681625366, "num_tokens": 393409858.0, "step": 10357 }, { "epoch": 1.3176440656405037, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6694869995117188, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8761703968048096, "num_tokens": 393451198.0, "step": 10358 }, { "epoch": 1.3177712759190943, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5391247272491455, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8812108039855957, "num_tokens": 393486916.0, "step": 10359 }, { "epoch": 1.3178984861976848, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6835267543792725, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8844119310379028, "num_tokens": 393520039.0, "step": 10360 }, { "epoch": 1.3180256964762753, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5392417907714844, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8606494665145874, "num_tokens": 393557960.0, "step": 10361 }, { "epoch": 1.3181529067548658, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.687007427215576, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8598216772079468, "num_tokens": 393592697.0, "step": 10362 }, { "epoch": 1.3182801170334564, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.526170253753662, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8616455793380737, "num_tokens": 393635423.0, "step": 10363 }, { "epoch": 1.318407327312047, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.693960189819336, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.877650797367096, "num_tokens": 393673508.0, "step": 10364 }, { "epoch": 1.3185345375906374, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.519213914871216, "learning_rate": 1e-05, "loss": 0.3898, "mean_token_accuracy": 0.8867753744125366, "num_tokens": 393711974.0, "step": 10365 }, { "epoch": 1.318661747869228, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.723127603530884, "learning_rate": 1e-05, "loss": 0.5071, "mean_token_accuracy": 0.8493648767471313, "num_tokens": 393752670.0, "step": 10366 }, { "epoch": 1.3187889581478183, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.526810646057129, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8670693635940552, "num_tokens": 393794312.0, "step": 10367 }, { "epoch": 1.3189161684264088, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.721848249435425, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8587061166763306, "num_tokens": 393833830.0, "step": 10368 }, { "epoch": 1.3190433787049993, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5152173042297363, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8742413520812988, "num_tokens": 393874844.0, "step": 10369 }, { "epoch": 1.3191705889835899, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7389261722564697, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8552000522613525, "num_tokens": 393913861.0, "step": 10370 }, { "epoch": 1.3192977992621804, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5256428718566895, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8652017116546631, "num_tokens": 393953014.0, "step": 10371 }, { "epoch": 1.319425009540771, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.732696533203125, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8717682957649231, "num_tokens": 393995603.0, "step": 10372 }, { "epoch": 1.3195522198193614, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5162131786346436, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.87761390209198, "num_tokens": 394032127.0, "step": 10373 }, { "epoch": 1.319679430097952, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.734287977218628, "learning_rate": 1e-05, "loss": 0.3862, "mean_token_accuracy": 0.8865610361099243, "num_tokens": 394067153.0, "step": 10374 }, { "epoch": 1.3198066403765425, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5229642391204834, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8721080422401428, "num_tokens": 394103484.0, "step": 10375 }, { "epoch": 1.3199338506551328, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7311251163482666, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8763318061828613, "num_tokens": 394144302.0, "step": 10376 }, { "epoch": 1.3200610609337233, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5337212085723877, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8671935796737671, "num_tokens": 394179258.0, "step": 10377 }, { "epoch": 1.3201882712123139, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7318758964538574, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.860910177230835, "num_tokens": 394221498.0, "step": 10378 }, { "epoch": 1.3203154814909044, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5344631671905518, "learning_rate": 1e-05, "loss": 0.4121, "mean_token_accuracy": 0.8831295967102051, "num_tokens": 394255985.0, "step": 10379 }, { "epoch": 1.320442691769495, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7466838359832764, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8578370809555054, "num_tokens": 394291020.0, "step": 10380 }, { "epoch": 1.3205699020480854, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.525967836380005, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8665058612823486, "num_tokens": 394329415.0, "step": 10381 }, { "epoch": 1.320697112326676, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.746366024017334, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8694953918457031, "num_tokens": 394365287.0, "step": 10382 }, { "epoch": 1.3208243226052665, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5253424644470215, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8747094869613647, "num_tokens": 394403749.0, "step": 10383 }, { "epoch": 1.320951532883857, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7295100688934326, "learning_rate": 1e-05, "loss": 0.3804, "mean_token_accuracy": 0.8884813189506531, "num_tokens": 394445443.0, "step": 10384 }, { "epoch": 1.3210787431624476, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5121161937713623, "learning_rate": 1e-05, "loss": 0.3812, "mean_token_accuracy": 0.8872357606887817, "num_tokens": 394487650.0, "step": 10385 }, { "epoch": 1.321205953441038, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.749767780303955, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8804755210876465, "num_tokens": 394524606.0, "step": 10386 }, { "epoch": 1.3213331637196286, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5335609912872314, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8629391193389893, "num_tokens": 394561177.0, "step": 10387 }, { "epoch": 1.3214603739982191, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.732624053955078, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8810852766036987, "num_tokens": 394605212.0, "step": 10388 }, { "epoch": 1.3215875842768097, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5466747283935547, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8731603622436523, "num_tokens": 394637662.0, "step": 10389 }, { "epoch": 1.3217147945554002, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.728726387023926, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.878091037273407, "num_tokens": 394675551.0, "step": 10390 }, { "epoch": 1.3218420048339907, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.540606737136841, "learning_rate": 1e-05, "loss": 0.4838, "mean_token_accuracy": 0.856061577796936, "num_tokens": 394715225.0, "step": 10391 }, { "epoch": 1.321969215112581, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7196364402770996, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.8813369274139404, "num_tokens": 394749872.0, "step": 10392 }, { "epoch": 1.3220964253911716, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.536329746246338, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8575659990310669, "num_tokens": 394791674.0, "step": 10393 }, { "epoch": 1.322223635669762, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.722907304763794, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8600936532020569, "num_tokens": 394829556.0, "step": 10394 }, { "epoch": 1.3223508459483526, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5346481800079346, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.875093936920166, "num_tokens": 394864421.0, "step": 10395 }, { "epoch": 1.3224780562269431, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7192208766937256, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8746232986450195, "num_tokens": 394901008.0, "step": 10396 }, { "epoch": 1.3226052665055337, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.534794807434082, "learning_rate": 1e-05, "loss": 0.5352, "mean_token_accuracy": 0.8447486758232117, "num_tokens": 394947172.0, "step": 10397 }, { "epoch": 1.3227324767841242, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7196054458618164, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8629043698310852, "num_tokens": 394984515.0, "step": 10398 }, { "epoch": 1.3228596870627147, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5445914268493652, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8594611883163452, "num_tokens": 395019819.0, "step": 10399 }, { "epoch": 1.3229868973413053, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7291319370269775, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8680703639984131, "num_tokens": 395058606.0, "step": 10400 }, { "epoch": 1.3231141076198956, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5229039192199707, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8800569772720337, "num_tokens": 395094354.0, "step": 10401 }, { "epoch": 1.323241317898486, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7152018547058105, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8702666163444519, "num_tokens": 395136034.0, "step": 10402 }, { "epoch": 1.3233685281770766, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5187954902648926, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.8872441053390503, "num_tokens": 395176335.0, "step": 10403 }, { "epoch": 1.3234957384556671, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.722433567047119, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8816879987716675, "num_tokens": 395216297.0, "step": 10404 }, { "epoch": 1.3236229487342577, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.519676923751831, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8731992244720459, "num_tokens": 395255685.0, "step": 10405 }, { "epoch": 1.3237501590128482, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7283051013946533, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8631407618522644, "num_tokens": 395300484.0, "step": 10406 }, { "epoch": 1.3238773692914387, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.543121576309204, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8656256198883057, "num_tokens": 395333760.0, "step": 10407 }, { "epoch": 1.3240045795700293, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.720381259918213, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.868484616279602, "num_tokens": 395376264.0, "step": 10408 }, { "epoch": 1.3241317898486198, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5313572883605957, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8566985130310059, "num_tokens": 395417758.0, "step": 10409 }, { "epoch": 1.3242590001272103, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7164039611816406, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8657886981964111, "num_tokens": 395458657.0, "step": 10410 }, { "epoch": 1.3243862104058008, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5359439849853516, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8582801222801208, "num_tokens": 395500838.0, "step": 10411 }, { "epoch": 1.3245134206843914, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7121903896331787, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.868048369884491, "num_tokens": 395544912.0, "step": 10412 }, { "epoch": 1.324640630962982, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5280003547668457, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8726868629455566, "num_tokens": 395584661.0, "step": 10413 }, { "epoch": 1.3247678412415724, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7096192836761475, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8712071180343628, "num_tokens": 395623374.0, "step": 10414 }, { "epoch": 1.324895051520163, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.533687114715576, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.8811911344528198, "num_tokens": 395658854.0, "step": 10415 }, { "epoch": 1.3250222617987533, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7077698707580566, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8775268793106079, "num_tokens": 395695298.0, "step": 10416 }, { "epoch": 1.3251494720773438, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5444440841674805, "learning_rate": 1e-05, "loss": 0.5036, "mean_token_accuracy": 0.8470101356506348, "num_tokens": 395731800.0, "step": 10417 }, { "epoch": 1.3252766823559343, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.693641185760498, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.861386775970459, "num_tokens": 395773811.0, "step": 10418 }, { "epoch": 1.3254038926345248, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.528879404067993, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8745543360710144, "num_tokens": 395817280.0, "step": 10419 }, { "epoch": 1.3255311029131154, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.700458288192749, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8629122972488403, "num_tokens": 395857950.0, "step": 10420 }, { "epoch": 1.325658313191706, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5340137481689453, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8610881567001343, "num_tokens": 395893048.0, "step": 10421 }, { "epoch": 1.3257855234702964, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6973788738250732, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8750134706497192, "num_tokens": 395932678.0, "step": 10422 }, { "epoch": 1.325912733748887, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5375969409942627, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8760906457901001, "num_tokens": 395967253.0, "step": 10423 }, { "epoch": 1.3260399440274775, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6969587802886963, "learning_rate": 1e-05, "loss": 0.3862, "mean_token_accuracy": 0.8875962495803833, "num_tokens": 396000935.0, "step": 10424 }, { "epoch": 1.3261671543060678, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.533123254776001, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8558192253112793, "num_tokens": 396039148.0, "step": 10425 }, { "epoch": 1.3262943645846583, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.704791784286499, "learning_rate": 1e-05, "loss": 0.3936, "mean_token_accuracy": 0.8837833404541016, "num_tokens": 396072035.0, "step": 10426 }, { "epoch": 1.3264215748632489, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5265793800354004, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8656886219978333, "num_tokens": 396112047.0, "step": 10427 }, { "epoch": 1.3265487851418394, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.716994524002075, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8699705600738525, "num_tokens": 396149844.0, "step": 10428 }, { "epoch": 1.32667599542043, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.512317657470703, "learning_rate": 1e-05, "loss": 0.3687, "mean_token_accuracy": 0.8922326564788818, "num_tokens": 396187493.0, "step": 10429 }, { "epoch": 1.3268032056990204, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7385106086730957, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8719926476478577, "num_tokens": 396220590.0, "step": 10430 }, { "epoch": 1.326930415977611, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.517472982406616, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8663040399551392, "num_tokens": 396262590.0, "step": 10431 }, { "epoch": 1.3270576262562015, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7455270290374756, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8735198378562927, "num_tokens": 396296491.0, "step": 10432 }, { "epoch": 1.327184836534792, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5085630416870117, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8724120855331421, "num_tokens": 396334798.0, "step": 10433 }, { "epoch": 1.3273120468133826, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.775311231613159, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8687325716018677, "num_tokens": 396368278.0, "step": 10434 }, { "epoch": 1.327439257091973, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5075061321258545, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8749832510948181, "num_tokens": 396405034.0, "step": 10435 }, { "epoch": 1.3275664673705636, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7802927494049072, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8563896417617798, "num_tokens": 396445503.0, "step": 10436 }, { "epoch": 1.3276936776491541, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5152153968811035, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8761680722236633, "num_tokens": 396482462.0, "step": 10437 }, { "epoch": 1.3278208879277447, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00023746490478515625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7769248485565186, "learning_rate": 1e-05, "loss": 0.3778, "mean_token_accuracy": 0.8890496492385864, "num_tokens": 396519186.0, "step": 10438 }, { "epoch": 1.3279480982063352, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.51798152923584, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8699593544006348, "num_tokens": 396558996.0, "step": 10439 }, { "epoch": 1.3280753084849257, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.8042829036712646, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8526850938796997, "num_tokens": 396593158.0, "step": 10440 }, { "epoch": 1.328202518763516, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5362088680267334, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.861685037612915, "num_tokens": 396629189.0, "step": 10441 }, { "epoch": 1.3283297290421066, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.798184871673584, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8524823784828186, "num_tokens": 396665932.0, "step": 10442 }, { "epoch": 1.328456939320697, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.546738624572754, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8571517467498779, "num_tokens": 396706638.0, "step": 10443 }, { "epoch": 1.3285841495992876, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.778710126876831, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8748074173927307, "num_tokens": 396744354.0, "step": 10444 }, { "epoch": 1.3287113598778781, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5655248165130615, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8741727471351624, "num_tokens": 396783806.0, "step": 10445 }, { "epoch": 1.3288385701564687, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.748838186264038, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8711053133010864, "num_tokens": 396822337.0, "step": 10446 }, { "epoch": 1.3289657804350592, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5625576972961426, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8561047315597534, "num_tokens": 396861674.0, "step": 10447 }, { "epoch": 1.3290929907136497, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7359554767608643, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8644749522209167, "num_tokens": 396897686.0, "step": 10448 }, { "epoch": 1.3292202009922403, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5554463863372803, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8631308674812317, "num_tokens": 396936261.0, "step": 10449 }, { "epoch": 1.3293474112708306, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.734264373779297, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8641714453697205, "num_tokens": 396967452.0, "step": 10450 }, { "epoch": 1.329474621549421, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5417771339416504, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.8854204416275024, "num_tokens": 397004672.0, "step": 10451 }, { "epoch": 1.3296018318280116, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7245607376098633, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8679512739181519, "num_tokens": 397039310.0, "step": 10452 }, { "epoch": 1.3297290421066021, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.548917531967163, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8777991533279419, "num_tokens": 397077465.0, "step": 10453 }, { "epoch": 1.3298562523851927, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.727285146713257, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8775047063827515, "num_tokens": 397108606.0, "step": 10454 }, { "epoch": 1.3299834626637832, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.545426845550537, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8669307231903076, "num_tokens": 397146628.0, "step": 10455 }, { "epoch": 1.3301106729423737, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7156128883361816, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8705685138702393, "num_tokens": 397182345.0, "step": 10456 }, { "epoch": 1.3302378832209643, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5457818508148193, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8711493015289307, "num_tokens": 397218673.0, "step": 10457 }, { "epoch": 1.3303650934995548, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.721912384033203, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8795835971832275, "num_tokens": 397250186.0, "step": 10458 }, { "epoch": 1.3304923037781453, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.535945177078247, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8691303730010986, "num_tokens": 397287947.0, "step": 10459 }, { "epoch": 1.3306195140567358, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7092695236206055, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8671444654464722, "num_tokens": 397328523.0, "step": 10460 }, { "epoch": 1.3307467243353264, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.532804489135742, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8724469542503357, "num_tokens": 397366238.0, "step": 10461 }, { "epoch": 1.330873934613917, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.703028678894043, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8844107389450073, "num_tokens": 397404324.0, "step": 10462 }, { "epoch": 1.3310011448925074, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.53157377243042, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8708266615867615, "num_tokens": 397441074.0, "step": 10463 }, { "epoch": 1.331128355171098, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7104203701019287, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8595048189163208, "num_tokens": 397476913.0, "step": 10464 }, { "epoch": 1.3312555654496883, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5424089431762695, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8750250339508057, "num_tokens": 397507569.0, "step": 10465 }, { "epoch": 1.3313827757282788, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6996538639068604, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.877139151096344, "num_tokens": 397544131.0, "step": 10466 }, { "epoch": 1.3315099860068693, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5215399265289307, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8864284753799438, "num_tokens": 397585042.0, "step": 10467 }, { "epoch": 1.3316371962854598, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7297818660736084, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8659987449645996, "num_tokens": 397620169.0, "step": 10468 }, { "epoch": 1.3317644065640504, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.530142307281494, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8668018579483032, "num_tokens": 397657349.0, "step": 10469 }, { "epoch": 1.331891616842641, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.743849754333496, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8539924621582031, "num_tokens": 397696309.0, "step": 10470 }, { "epoch": 1.3320188271212314, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5177297592163086, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8758487701416016, "num_tokens": 397734520.0, "step": 10471 }, { "epoch": 1.332146037399822, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7550933361053467, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8618741631507874, "num_tokens": 397773078.0, "step": 10472 }, { "epoch": 1.3322732476784125, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5149295330047607, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8809046745300293, "num_tokens": 397814180.0, "step": 10473 }, { "epoch": 1.3324004579570028, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.753176689147949, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.8799763917922974, "num_tokens": 397850618.0, "step": 10474 }, { "epoch": 1.3325276682355933, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.538632392883301, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.853767991065979, "num_tokens": 397888538.0, "step": 10475 }, { "epoch": 1.3326548785141838, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7698285579681396, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8795557022094727, "num_tokens": 397917861.0, "step": 10476 }, { "epoch": 1.3327820887927744, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.524268388748169, "learning_rate": 1e-05, "loss": 0.3835, "mean_token_accuracy": 0.8872326612472534, "num_tokens": 397954052.0, "step": 10477 }, { "epoch": 1.332909299071365, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7492876052856445, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8716515302658081, "num_tokens": 397996458.0, "step": 10478 }, { "epoch": 1.3330365093499554, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5338408946990967, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8807697296142578, "num_tokens": 398034164.0, "step": 10479 }, { "epoch": 1.333163719628546, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7421457767486572, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8803961277008057, "num_tokens": 398075828.0, "step": 10480 }, { "epoch": 1.3332909299071365, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5573744773864746, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8715839982032776, "num_tokens": 398112941.0, "step": 10481 }, { "epoch": 1.333418140185727, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.737060546875, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8806875944137573, "num_tokens": 398146255.0, "step": 10482 }, { "epoch": 1.3335453504643175, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5498671531677246, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.881844162940979, "num_tokens": 398178552.0, "step": 10483 }, { "epoch": 1.333672560742908, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7153398990631104, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.882335901260376, "num_tokens": 398216310.0, "step": 10484 }, { "epoch": 1.3337997710214986, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5568244457244873, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8710932731628418, "num_tokens": 398249756.0, "step": 10485 }, { "epoch": 1.3339269813000891, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7189242839813232, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8695000410079956, "num_tokens": 398285264.0, "step": 10486 }, { "epoch": 1.3340541915786797, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5577454566955566, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8537662029266357, "num_tokens": 398320964.0, "step": 10487 }, { "epoch": 1.3341814018572702, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7142508029937744, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8522747159004211, "num_tokens": 398355392.0, "step": 10488 }, { "epoch": 1.3343086121358605, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5468955039978027, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8682916164398193, "num_tokens": 398393551.0, "step": 10489 }, { "epoch": 1.334435822414451, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6980104446411133, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8548944592475891, "num_tokens": 398433812.0, "step": 10490 }, { "epoch": 1.3345630326930416, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5461442470550537, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.86949622631073, "num_tokens": 398469790.0, "step": 10491 }, { "epoch": 1.334690242971632, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.684515953063965, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8771675229072571, "num_tokens": 398512253.0, "step": 10492 }, { "epoch": 1.3348174532502226, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5499303340911865, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8508316278457642, "num_tokens": 398549790.0, "step": 10493 }, { "epoch": 1.3349446635288131, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.68234920501709, "learning_rate": 1e-05, "loss": 0.3855, "mean_token_accuracy": 0.8854024410247803, "num_tokens": 398586844.0, "step": 10494 }, { "epoch": 1.3350718738074037, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.55077862739563, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8679680824279785, "num_tokens": 398624177.0, "step": 10495 }, { "epoch": 1.3351990840859942, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6752822399139404, "learning_rate": 1e-05, "loss": 0.3987, "mean_token_accuracy": 0.8824517130851746, "num_tokens": 398661840.0, "step": 10496 }, { "epoch": 1.3353262943645847, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.542999267578125, "learning_rate": 1e-05, "loss": 0.3988, "mean_token_accuracy": 0.8833751678466797, "num_tokens": 398699535.0, "step": 10497 }, { "epoch": 1.3354535046431752, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.686811685562134, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8600976467132568, "num_tokens": 398736294.0, "step": 10498 }, { "epoch": 1.3355807149217656, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.541102886199951, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8778307437896729, "num_tokens": 398776372.0, "step": 10499 }, { "epoch": 1.335707925200356, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6896145343780518, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8660767674446106, "num_tokens": 398816000.0, "step": 10500 }, { "epoch": 1.3358351354789466, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5315089225769043, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8809279203414917, "num_tokens": 398851960.0, "step": 10501 }, { "epoch": 1.3359623457575371, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7123677730560303, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8689540028572083, "num_tokens": 398886485.0, "step": 10502 }, { "epoch": 1.3360895560361277, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5317299365997314, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8629697561264038, "num_tokens": 398923800.0, "step": 10503 }, { "epoch": 1.3362167663147182, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.708730697631836, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.87406325340271, "num_tokens": 398963578.0, "step": 10504 }, { "epoch": 1.3363439765933087, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.509573221206665, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8737263679504395, "num_tokens": 399009791.0, "step": 10505 }, { "epoch": 1.3364711868718993, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.721653461456299, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8800467848777771, "num_tokens": 399048701.0, "step": 10506 }, { "epoch": 1.3365983971504898, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5121853351593018, "learning_rate": 1e-05, "loss": 0.4003, "mean_token_accuracy": 0.8800290822982788, "num_tokens": 399083441.0, "step": 10507 }, { "epoch": 1.3367256074290803, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7525157928466797, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8773658871650696, "num_tokens": 399113762.0, "step": 10508 }, { "epoch": 1.3368528177076708, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5128087997436523, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8773849010467529, "num_tokens": 399149966.0, "step": 10509 }, { "epoch": 1.3369800279862614, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.768566846847534, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8687261939048767, "num_tokens": 399188777.0, "step": 10510 }, { "epoch": 1.337107238264852, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.497758626937866, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8798093795776367, "num_tokens": 399229618.0, "step": 10511 }, { "epoch": 1.3372344485434424, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7774763107299805, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.877822756767273, "num_tokens": 399270558.0, "step": 10512 }, { "epoch": 1.337361658822033, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.521881580352783, "learning_rate": 1e-05, "loss": 0.5367, "mean_token_accuracy": 0.8406683206558228, "num_tokens": 399310660.0, "step": 10513 }, { "epoch": 1.3374888691006233, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7817094326019287, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8684754967689514, "num_tokens": 399347838.0, "step": 10514 }, { "epoch": 1.3376160793792138, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.527296304702759, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8770129680633545, "num_tokens": 399384790.0, "step": 10515 }, { "epoch": 1.3377432896578043, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7779452800750732, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8793502449989319, "num_tokens": 399425479.0, "step": 10516 }, { "epoch": 1.3378704999363948, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5257632732391357, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.87236088514328, "num_tokens": 399467161.0, "step": 10517 }, { "epoch": 1.3379977102149854, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8027420043945312, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8608710765838623, "num_tokens": 399500204.0, "step": 10518 }, { "epoch": 1.338124920493576, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5303096771240234, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8568558692932129, "num_tokens": 399546142.0, "step": 10519 }, { "epoch": 1.3382521307721664, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7961742877960205, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.853955090045929, "num_tokens": 399582446.0, "step": 10520 }, { "epoch": 1.338379341050757, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5609018802642822, "learning_rate": 1e-05, "loss": 0.533, "mean_token_accuracy": 0.8426505327224731, "num_tokens": 399616998.0, "step": 10521 }, { "epoch": 1.3385065513293475, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7836170196533203, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8746538162231445, "num_tokens": 399648649.0, "step": 10522 }, { "epoch": 1.3386337616079378, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5282363891601562, "learning_rate": 1e-05, "loss": 0.3902, "mean_token_accuracy": 0.8851484060287476, "num_tokens": 399690417.0, "step": 10523 }, { "epoch": 1.3387609718865283, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7716801166534424, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8639112114906311, "num_tokens": 399732346.0, "step": 10524 }, { "epoch": 1.3388881821651188, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5462212562561035, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.861088752746582, "num_tokens": 399770516.0, "step": 10525 }, { "epoch": 1.3390153924437094, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.772707939147949, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.868492841720581, "num_tokens": 399803266.0, "step": 10526 }, { "epoch": 1.3391426027223, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5538697242736816, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8677060008049011, "num_tokens": 399841940.0, "step": 10527 }, { "epoch": 1.3392698130008904, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7620303630828857, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8594437837600708, "num_tokens": 399880207.0, "step": 10528 }, { "epoch": 1.339397023279481, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5616724491119385, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8631467223167419, "num_tokens": 399915731.0, "step": 10529 }, { "epoch": 1.3395242335580715, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7400808334350586, "learning_rate": 1e-05, "loss": 0.3674, "mean_token_accuracy": 0.8944344520568848, "num_tokens": 399953398.0, "step": 10530 }, { "epoch": 1.339651443836662, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.563493490219116, "learning_rate": 1e-05, "loss": 0.4885, "mean_token_accuracy": 0.8538447618484497, "num_tokens": 399991132.0, "step": 10531 }, { "epoch": 1.3397786541152525, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7168664932250977, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8644778728485107, "num_tokens": 400031615.0, "step": 10532 }, { "epoch": 1.339905864393843, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5614473819732666, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8518801331520081, "num_tokens": 400070804.0, "step": 10533 }, { "epoch": 1.3400330746724336, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.70460844039917, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8640906810760498, "num_tokens": 400114021.0, "step": 10534 }, { "epoch": 1.3401602849510241, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5442609786987305, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8750059604644775, "num_tokens": 400156526.0, "step": 10535 }, { "epoch": 1.3402874952296147, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7104856967926025, "learning_rate": 1e-05, "loss": 0.3784, "mean_token_accuracy": 0.890904426574707, "num_tokens": 400188487.0, "step": 10536 }, { "epoch": 1.3404147055082052, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5663981437683105, "learning_rate": 1e-05, "loss": 0.5127, "mean_token_accuracy": 0.8482557535171509, "num_tokens": 400222223.0, "step": 10537 }, { "epoch": 1.3405419157867955, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6857850551605225, "learning_rate": 1e-05, "loss": 0.39, "mean_token_accuracy": 0.8861018419265747, "num_tokens": 400259181.0, "step": 10538 }, { "epoch": 1.340669126065386, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.538198232650757, "learning_rate": 1e-05, "loss": 0.3924, "mean_token_accuracy": 0.8813918828964233, "num_tokens": 400295202.0, "step": 10539 }, { "epoch": 1.3407963363439765, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.684385299682617, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8814072608947754, "num_tokens": 400330936.0, "step": 10540 }, { "epoch": 1.340923546622567, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5510542392730713, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8721211552619934, "num_tokens": 400367971.0, "step": 10541 }, { "epoch": 1.3410507569011576, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.681640386581421, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8658876419067383, "num_tokens": 400406868.0, "step": 10542 }, { "epoch": 1.3411779671797481, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5555357933044434, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8556690216064453, "num_tokens": 400442907.0, "step": 10543 }, { "epoch": 1.3413051774583387, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.695932626724243, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8574658036231995, "num_tokens": 400479485.0, "step": 10544 }, { "epoch": 1.3414323877369292, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5436010360717773, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8668618202209473, "num_tokens": 400513790.0, "step": 10545 }, { "epoch": 1.3415595980155197, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6791257858276367, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8704686164855957, "num_tokens": 400557638.0, "step": 10546 }, { "epoch": 1.3416868082941102, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5360360145568848, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.8809613585472107, "num_tokens": 400594121.0, "step": 10547 }, { "epoch": 1.3418140185727006, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6858034133911133, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8693510293960571, "num_tokens": 400632427.0, "step": 10548 }, { "epoch": 1.341941228851291, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5441644191741943, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8746522665023804, "num_tokens": 400666938.0, "step": 10549 }, { "epoch": 1.3420684391298816, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6861090660095215, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8830640316009521, "num_tokens": 400703510.0, "step": 10550 }, { "epoch": 1.3421956494084721, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5443129539489746, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8744347095489502, "num_tokens": 400737082.0, "step": 10551 }, { "epoch": 1.3423228596870627, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6655354499816895, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8704203367233276, "num_tokens": 400787254.0, "step": 10552 }, { "epoch": 1.3424500699656532, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5450901985168457, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8732388019561768, "num_tokens": 400824384.0, "step": 10553 }, { "epoch": 1.3425772802442437, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.680263042449951, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8646758794784546, "num_tokens": 400859536.0, "step": 10554 }, { "epoch": 1.3427044905228342, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5404341220855713, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8817628026008606, "num_tokens": 400896022.0, "step": 10555 }, { "epoch": 1.3428317008014248, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6767683029174805, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.8789438009262085, "num_tokens": 400936291.0, "step": 10556 }, { "epoch": 1.3429589110800153, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.519557237625122, "learning_rate": 1e-05, "loss": 0.3964, "mean_token_accuracy": 0.8801674842834473, "num_tokens": 400980749.0, "step": 10557 }, { "epoch": 1.3430861213586058, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7051331996917725, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8539087772369385, "num_tokens": 401016591.0, "step": 10558 }, { "epoch": 1.3432133316371964, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5123484134674072, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8687102794647217, "num_tokens": 401062520.0, "step": 10559 }, { "epoch": 1.343340541915787, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.71152400970459, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.875354528427124, "num_tokens": 401105968.0, "step": 10560 }, { "epoch": 1.3434677521943774, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.528350353240967, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8844523429870605, "num_tokens": 401145290.0, "step": 10561 }, { "epoch": 1.343594962472968, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7216005325317383, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.879568338394165, "num_tokens": 401181253.0, "step": 10562 }, { "epoch": 1.3437221727515583, "ewc_loss": 0.06103515625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5082921981811523, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8720041513442993, "num_tokens": 401218191.0, "step": 10563 }, { "epoch": 1.3438493830301488, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7309467792510986, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8699480891227722, "num_tokens": 401252049.0, "step": 10564 }, { "epoch": 1.3439765933087393, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.514493942260742, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.849022626876831, "num_tokens": 401294819.0, "step": 10565 }, { "epoch": 1.3441038035873298, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7368180751800537, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8755525350570679, "num_tokens": 401332007.0, "step": 10566 }, { "epoch": 1.3442310138659204, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5053036212921143, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8809182047843933, "num_tokens": 401371442.0, "step": 10567 }, { "epoch": 1.344358224144511, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.754106044769287, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8775482773780823, "num_tokens": 401409956.0, "step": 10568 }, { "epoch": 1.3444854344231014, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5138752460479736, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8806614875793457, "num_tokens": 401446445.0, "step": 10569 }, { "epoch": 1.344612644701692, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7623205184936523, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8614929914474487, "num_tokens": 401489366.0, "step": 10570 }, { "epoch": 1.3447398549802825, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.510486602783203, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8653572797775269, "num_tokens": 401533326.0, "step": 10571 }, { "epoch": 1.3448670652588728, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7769174575805664, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8751040697097778, "num_tokens": 401571607.0, "step": 10572 }, { "epoch": 1.3449942755374633, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.538444757461548, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8518984317779541, "num_tokens": 401607930.0, "step": 10573 }, { "epoch": 1.3451214858160538, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7608211040496826, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.884118914604187, "num_tokens": 401645841.0, "step": 10574 }, { "epoch": 1.3452486960946444, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.546962022781372, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8622527718544006, "num_tokens": 401680795.0, "step": 10575 }, { "epoch": 1.345375906373235, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.751739025115967, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8717164993286133, "num_tokens": 401719710.0, "step": 10576 }, { "epoch": 1.3455031166518254, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.535951614379883, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8743386268615723, "num_tokens": 401758956.0, "step": 10577 }, { "epoch": 1.345630326930416, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7477962970733643, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8607022762298584, "num_tokens": 401798108.0, "step": 10578 }, { "epoch": 1.3457575372090065, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5393567085266113, "learning_rate": 1e-05, "loss": 0.4098, "mean_token_accuracy": 0.878002405166626, "num_tokens": 401833495.0, "step": 10579 }, { "epoch": 1.345884747487597, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002384185791015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.731856346130371, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8688762187957764, "num_tokens": 401875399.0, "step": 10580 }, { "epoch": 1.3460119577661875, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5366129875183105, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.872441828250885, "num_tokens": 401915540.0, "step": 10581 }, { "epoch": 1.346139168044778, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.732456922531128, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8572946190834045, "num_tokens": 401960013.0, "step": 10582 }, { "epoch": 1.3462663783233686, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5653653144836426, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.859847366809845, "num_tokens": 401990709.0, "step": 10583 }, { "epoch": 1.3463935886019591, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7066690921783447, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8771947026252747, "num_tokens": 402031277.0, "step": 10584 }, { "epoch": 1.3465207988805497, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.547121047973633, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8742209672927856, "num_tokens": 402073248.0, "step": 10585 }, { "epoch": 1.3466480091591402, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.702302932739258, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8665655851364136, "num_tokens": 402116072.0, "step": 10586 }, { "epoch": 1.3467752194377305, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.545292377471924, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.871685802936554, "num_tokens": 402153092.0, "step": 10587 }, { "epoch": 1.346902429716321, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7031357288360596, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8770503997802734, "num_tokens": 402192959.0, "step": 10588 }, { "epoch": 1.3470296399949115, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.564911365509033, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8700859546661377, "num_tokens": 402232245.0, "step": 10589 }, { "epoch": 1.347156850273502, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.703599691390991, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8829096555709839, "num_tokens": 402265862.0, "step": 10590 }, { "epoch": 1.3472840605520926, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5312674045562744, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8772341012954712, "num_tokens": 402303747.0, "step": 10591 }, { "epoch": 1.3474112708306831, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6950953006744385, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8739526271820068, "num_tokens": 402342952.0, "step": 10592 }, { "epoch": 1.3475384811092737, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.543398380279541, "learning_rate": 1e-05, "loss": 0.5165, "mean_token_accuracy": 0.8454689979553223, "num_tokens": 402382388.0, "step": 10593 }, { "epoch": 1.3476656913878642, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6955161094665527, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8765985369682312, "num_tokens": 402418937.0, "step": 10594 }, { "epoch": 1.3477929016664547, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5466434955596924, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8611083626747131, "num_tokens": 402454347.0, "step": 10595 }, { "epoch": 1.3479201119450452, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7120649814605713, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8543699383735657, "num_tokens": 402490887.0, "step": 10596 }, { "epoch": 1.3480473222236355, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5419657230377197, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8628126382827759, "num_tokens": 402531293.0, "step": 10597 }, { "epoch": 1.348174532502226, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.700251340866089, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8582409620285034, "num_tokens": 402574396.0, "step": 10598 }, { "epoch": 1.3483017427808166, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5297152996063232, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8703957796096802, "num_tokens": 402615609.0, "step": 10599 }, { "epoch": 1.3484289530594071, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7066903114318848, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8674225211143494, "num_tokens": 402650627.0, "step": 10600 }, { "epoch": 1.3485561633379977, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5272958278656006, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8664495944976807, "num_tokens": 402693036.0, "step": 10601 }, { "epoch": 1.3486833736165882, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7086691856384277, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8808331489562988, "num_tokens": 402730748.0, "step": 10602 }, { "epoch": 1.3488105838951787, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5511744022369385, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8610296249389648, "num_tokens": 402762642.0, "step": 10603 }, { "epoch": 1.3489377941737692, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.724912166595459, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8642603158950806, "num_tokens": 402794265.0, "step": 10604 }, { "epoch": 1.3490650044523598, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.544530153274536, "learning_rate": 1e-05, "loss": 0.5085, "mean_token_accuracy": 0.8506759405136108, "num_tokens": 402830609.0, "step": 10605 }, { "epoch": 1.3491922147309503, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.706521511077881, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8772773146629333, "num_tokens": 402866608.0, "step": 10606 }, { "epoch": 1.3493194250095408, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5322368144989014, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8637079000473022, "num_tokens": 402904030.0, "step": 10607 }, { "epoch": 1.3494466352881314, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.725112199783325, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8704234957695007, "num_tokens": 402939552.0, "step": 10608 }, { "epoch": 1.3495738455667219, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.523739814758301, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8775672912597656, "num_tokens": 402981683.0, "step": 10609 }, { "epoch": 1.3497010558453124, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7309303283691406, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8573564887046814, "num_tokens": 403019497.0, "step": 10610 }, { "epoch": 1.349828266123903, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5371205806732178, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8681919574737549, "num_tokens": 403053415.0, "step": 10611 }, { "epoch": 1.3499554764024932, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.728680372238159, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8708805441856384, "num_tokens": 403089676.0, "step": 10612 }, { "epoch": 1.3500826866810838, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5239274501800537, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8669820427894592, "num_tokens": 403130429.0, "step": 10613 }, { "epoch": 1.3502098969596743, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.724472999572754, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8833113312721252, "num_tokens": 403171420.0, "step": 10614 }, { "epoch": 1.3503371072382648, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5274744033813477, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8686551451683044, "num_tokens": 403210700.0, "step": 10615 }, { "epoch": 1.3504643175168554, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7421653270721436, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8714009523391724, "num_tokens": 403249572.0, "step": 10616 }, { "epoch": 1.350591527795446, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5295042991638184, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.864352822303772, "num_tokens": 403288176.0, "step": 10617 }, { "epoch": 1.3507187380740364, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7215044498443604, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8826522827148438, "num_tokens": 403329403.0, "step": 10618 }, { "epoch": 1.350845948352627, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5479519367218018, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8685859441757202, "num_tokens": 403362276.0, "step": 10619 }, { "epoch": 1.3509731586312175, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7242119312286377, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8705143332481384, "num_tokens": 403397797.0, "step": 10620 }, { "epoch": 1.3511003689098078, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5330569744110107, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8696939945220947, "num_tokens": 403437943.0, "step": 10621 }, { "epoch": 1.3512275791883983, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.741572856903076, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8770781755447388, "num_tokens": 403473575.0, "step": 10622 }, { "epoch": 1.3513547894669888, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.552081823348999, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8656896948814392, "num_tokens": 403504720.0, "step": 10623 }, { "epoch": 1.3514819997455794, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7212119102478027, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.860148549079895, "num_tokens": 403549244.0, "step": 10624 }, { "epoch": 1.35160921002417, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00023937225341796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.546947956085205, "learning_rate": 1e-05, "loss": 0.4965, "mean_token_accuracy": 0.8542768955230713, "num_tokens": 403582881.0, "step": 10625 }, { "epoch": 1.3517364203027604, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7074646949768066, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8809713125228882, "num_tokens": 403623477.0, "step": 10626 }, { "epoch": 1.351863630581351, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5412514209747314, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8572953939437866, "num_tokens": 403662399.0, "step": 10627 }, { "epoch": 1.3519908408599415, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.700207233428955, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8713893294334412, "num_tokens": 403703908.0, "step": 10628 }, { "epoch": 1.352118051138532, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5318546295166016, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.873120129108429, "num_tokens": 403746385.0, "step": 10629 }, { "epoch": 1.3522452614171225, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7044520378112793, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8751090168952942, "num_tokens": 403785728.0, "step": 10630 }, { "epoch": 1.352372471695713, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.525238037109375, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8762147426605225, "num_tokens": 403827991.0, "step": 10631 }, { "epoch": 1.3524996819743036, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7278261184692383, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8664425611495972, "num_tokens": 403861963.0, "step": 10632 }, { "epoch": 1.3526268922528941, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5186502933502197, "learning_rate": 1e-05, "loss": 0.3829, "mean_token_accuracy": 0.8846271634101868, "num_tokens": 403899744.0, "step": 10633 }, { "epoch": 1.3527541025314846, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7298166751861572, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.8872628211975098, "num_tokens": 403939482.0, "step": 10634 }, { "epoch": 1.3528813128100752, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5162105560302734, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8836900591850281, "num_tokens": 403976233.0, "step": 10635 }, { "epoch": 1.3530085230886655, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7522287368774414, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8800787925720215, "num_tokens": 404011181.0, "step": 10636 }, { "epoch": 1.353135733367256, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.503899097442627, "learning_rate": 1e-05, "loss": 0.3954, "mean_token_accuracy": 0.8824983239173889, "num_tokens": 404050519.0, "step": 10637 }, { "epoch": 1.3532629436458465, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7619802951812744, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8788438439369202, "num_tokens": 404089407.0, "step": 10638 }, { "epoch": 1.353390153924437, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5067479610443115, "learning_rate": 1e-05, "loss": 0.3652, "mean_token_accuracy": 0.8933929800987244, "num_tokens": 404126564.0, "step": 10639 }, { "epoch": 1.3535173642030276, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.772491455078125, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8687629699707031, "num_tokens": 404166113.0, "step": 10640 }, { "epoch": 1.3536445744816181, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5215282440185547, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8697354793548584, "num_tokens": 404204513.0, "step": 10641 }, { "epoch": 1.3537717847602087, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7855477333068848, "learning_rate": 1e-05, "loss": 0.5232, "mean_token_accuracy": 0.8533766865730286, "num_tokens": 404244475.0, "step": 10642 }, { "epoch": 1.3538989950387992, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5391364097595215, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8626006841659546, "num_tokens": 404278974.0, "step": 10643 }, { "epoch": 1.3540262053173897, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7814133167266846, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8624087572097778, "num_tokens": 404313123.0, "step": 10644 }, { "epoch": 1.3541534155959802, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5270986557006836, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.875333309173584, "num_tokens": 404350195.0, "step": 10645 }, { "epoch": 1.3542806258745705, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7791547775268555, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8736605048179626, "num_tokens": 404387044.0, "step": 10646 }, { "epoch": 1.354407836153161, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5365054607391357, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8724567890167236, "num_tokens": 404423877.0, "step": 10647 }, { "epoch": 1.3545350464317516, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7691256999969482, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.877612292766571, "num_tokens": 404466428.0, "step": 10648 }, { "epoch": 1.3546622567103421, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.545196533203125, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8684112429618835, "num_tokens": 404503341.0, "step": 10649 }, { "epoch": 1.3547894669889327, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7591724395751953, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8688709735870361, "num_tokens": 404543650.0, "step": 10650 }, { "epoch": 1.3549166772675232, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5365445613861084, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8648566603660583, "num_tokens": 404588834.0, "step": 10651 }, { "epoch": 1.3550438875461137, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7751169204711914, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8664655089378357, "num_tokens": 404625903.0, "step": 10652 }, { "epoch": 1.3551710978247042, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5364794731140137, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8746978044509888, "num_tokens": 404666299.0, "step": 10653 }, { "epoch": 1.3552983081032948, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.747999906539917, "learning_rate": 1e-05, "loss": 0.3756, "mean_token_accuracy": 0.8900595903396606, "num_tokens": 404702743.0, "step": 10654 }, { "epoch": 1.3554255183818853, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5415234565734863, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8520972728729248, "num_tokens": 404748010.0, "step": 10655 }, { "epoch": 1.3555527286604758, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.748387575149536, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.870789647102356, "num_tokens": 404785886.0, "step": 10656 }, { "epoch": 1.3556799389390664, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5469536781311035, "learning_rate": 1e-05, "loss": 0.3755, "mean_token_accuracy": 0.8896223306655884, "num_tokens": 404820982.0, "step": 10657 }, { "epoch": 1.3558071492176569, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7328598499298096, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8681666851043701, "num_tokens": 404863943.0, "step": 10658 }, { "epoch": 1.3559343594962474, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5469577312469482, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8652013540267944, "num_tokens": 404905168.0, "step": 10659 }, { "epoch": 1.356061569774838, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.736950635910034, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8699761033058167, "num_tokens": 404941863.0, "step": 10660 }, { "epoch": 1.3561887800534282, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5491437911987305, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8665710687637329, "num_tokens": 404979780.0, "step": 10661 }, { "epoch": 1.3563159903320188, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7231030464172363, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8631942272186279, "num_tokens": 405019194.0, "step": 10662 }, { "epoch": 1.3564432006106093, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.541375160217285, "learning_rate": 1e-05, "loss": 0.3813, "mean_token_accuracy": 0.8888336420059204, "num_tokens": 405056021.0, "step": 10663 }, { "epoch": 1.3565704108891998, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.715280771255493, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8607826232910156, "num_tokens": 405099195.0, "step": 10664 }, { "epoch": 1.3566976211677904, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.535069227218628, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8676237463951111, "num_tokens": 405142821.0, "step": 10665 }, { "epoch": 1.3568248314463809, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7321503162384033, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8666061758995056, "num_tokens": 405178434.0, "step": 10666 }, { "epoch": 1.3569520417249714, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5516390800476074, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8566892147064209, "num_tokens": 405211197.0, "step": 10667 }, { "epoch": 1.357079252003562, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.700958728790283, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8759469389915466, "num_tokens": 405250330.0, "step": 10668 }, { "epoch": 1.3572064622821525, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5374252796173096, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8725422620773315, "num_tokens": 405291694.0, "step": 10669 }, { "epoch": 1.3573336725607428, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.701202392578125, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8812459707260132, "num_tokens": 405326162.0, "step": 10670 }, { "epoch": 1.3574608828393333, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.54805064201355, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8797858953475952, "num_tokens": 405361862.0, "step": 10671 }, { "epoch": 1.3575880931179238, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6946022510528564, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8717886209487915, "num_tokens": 405402934.0, "step": 10672 }, { "epoch": 1.3577153033965144, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5357956886291504, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8637837171554565, "num_tokens": 405442206.0, "step": 10673 }, { "epoch": 1.357842513675105, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7079505920410156, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8439565896987915, "num_tokens": 405483351.0, "step": 10674 }, { "epoch": 1.3579697239536954, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5350472927093506, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8642557859420776, "num_tokens": 405523763.0, "step": 10675 }, { "epoch": 1.358096934232286, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.70510196685791, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.883772611618042, "num_tokens": 405559640.0, "step": 10676 }, { "epoch": 1.3582241445108765, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.540966510772705, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8752285242080688, "num_tokens": 405594907.0, "step": 10677 }, { "epoch": 1.358351354789467, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6992478370666504, "learning_rate": 1e-05, "loss": 0.3916, "mean_token_accuracy": 0.8860756754875183, "num_tokens": 405630346.0, "step": 10678 }, { "epoch": 1.3584785650680575, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5558207035064697, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8636608123779297, "num_tokens": 405661603.0, "step": 10679 }, { "epoch": 1.358605775346648, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.690634250640869, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.870820164680481, "num_tokens": 405698907.0, "step": 10680 }, { "epoch": 1.3587329856252386, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5320839881896973, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8614028096199036, "num_tokens": 405744780.0, "step": 10681 }, { "epoch": 1.3588601959038291, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.694636583328247, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8787569403648376, "num_tokens": 405777883.0, "step": 10682 }, { "epoch": 1.3589874061824196, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5406951904296875, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8692588210105896, "num_tokens": 405814680.0, "step": 10683 }, { "epoch": 1.3591146164610102, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7105610370635986, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8540558815002441, "num_tokens": 405847734.0, "step": 10684 }, { "epoch": 1.3592418267396005, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5390307903289795, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8675971031188965, "num_tokens": 405882365.0, "step": 10685 }, { "epoch": 1.359369037018191, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.701620101928711, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8740261197090149, "num_tokens": 405920641.0, "step": 10686 }, { "epoch": 1.3594962472967815, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.00024127960205078125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.538360595703125, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8646795749664307, "num_tokens": 405958523.0, "step": 10687 }, { "epoch": 1.359623457575372, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.699291944503784, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8758742809295654, "num_tokens": 405994338.0, "step": 10688 }, { "epoch": 1.3597506678539626, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.543226480484009, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.856048047542572, "num_tokens": 406033285.0, "step": 10689 }, { "epoch": 1.3598778781325531, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6767001152038574, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.879631757736206, "num_tokens": 406076933.0, "step": 10690 }, { "epoch": 1.3600050884111436, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5357024669647217, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.872880220413208, "num_tokens": 406118258.0, "step": 10691 }, { "epoch": 1.3601322986897342, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6872215270996094, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8785583972930908, "num_tokens": 406155445.0, "step": 10692 }, { "epoch": 1.3602595089683247, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5302109718322754, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8686262965202332, "num_tokens": 406195128.0, "step": 10693 }, { "epoch": 1.3603867192469152, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.699622869491577, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.8768885135650635, "num_tokens": 406229311.0, "step": 10694 }, { "epoch": 1.3605139295255055, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.532033920288086, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8701741695404053, "num_tokens": 406266252.0, "step": 10695 }, { "epoch": 1.360641139804096, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6898810863494873, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.877799391746521, "num_tokens": 406310555.0, "step": 10696 }, { "epoch": 1.3607683500826866, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.52667236328125, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.8819994926452637, "num_tokens": 406349620.0, "step": 10697 }, { "epoch": 1.3608955603612771, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.723050832748413, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8725125193595886, "num_tokens": 406380956.0, "step": 10698 }, { "epoch": 1.3610227706398677, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5136396884918213, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.884967565536499, "num_tokens": 406417274.0, "step": 10699 }, { "epoch": 1.3611499809184582, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.741440773010254, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8576800227165222, "num_tokens": 406453809.0, "step": 10700 }, { "epoch": 1.3612771911970487, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5147619247436523, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.875507116317749, "num_tokens": 406489227.0, "step": 10701 }, { "epoch": 1.3614044014756392, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.742002010345459, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8754401206970215, "num_tokens": 406532736.0, "step": 10702 }, { "epoch": 1.3615316117542298, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.518554210662842, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8704074621200562, "num_tokens": 406574409.0, "step": 10703 }, { "epoch": 1.3616588220328203, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7602531909942627, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.858773410320282, "num_tokens": 406612519.0, "step": 10704 }, { "epoch": 1.3617860323114108, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.520289182662964, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8634561896324158, "num_tokens": 406652578.0, "step": 10705 }, { "epoch": 1.3619132425900013, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.770768165588379, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8577337265014648, "num_tokens": 406688739.0, "step": 10706 }, { "epoch": 1.3620404528685919, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.523881196975708, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8773568868637085, "num_tokens": 406727721.0, "step": 10707 }, { "epoch": 1.3621676631471824, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7800633907318115, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8654212951660156, "num_tokens": 406762864.0, "step": 10708 }, { "epoch": 1.362294873425773, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5118045806884766, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.879980206489563, "num_tokens": 406803333.0, "step": 10709 }, { "epoch": 1.3624220837043632, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7893857955932617, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8765908479690552, "num_tokens": 406840942.0, "step": 10710 }, { "epoch": 1.3625492939829538, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.529747724533081, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8730595707893372, "num_tokens": 406877293.0, "step": 10711 }, { "epoch": 1.3626765042615443, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8058764934539795, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8727981448173523, "num_tokens": 406915622.0, "step": 10712 }, { "epoch": 1.3628037145401348, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.526002883911133, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8827662467956543, "num_tokens": 406955378.0, "step": 10713 }, { "epoch": 1.3629309248187254, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8039467334747314, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.862762987613678, "num_tokens": 406997305.0, "step": 10714 }, { "epoch": 1.3630581350973159, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.521416187286377, "learning_rate": 1e-05, "loss": 0.3911, "mean_token_accuracy": 0.8862482905387878, "num_tokens": 407039121.0, "step": 10715 }, { "epoch": 1.3631853453759064, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.79764723777771, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8629752397537231, "num_tokens": 407081007.0, "step": 10716 }, { "epoch": 1.363312555654497, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.54258131980896, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8727870583534241, "num_tokens": 407122498.0, "step": 10717 }, { "epoch": 1.3634397659330875, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7744061946868896, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.880670428276062, "num_tokens": 407169304.0, "step": 10718 }, { "epoch": 1.3635669762116778, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.569700002670288, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8585518598556519, "num_tokens": 407204364.0, "step": 10719 }, { "epoch": 1.3636941864902683, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.771531581878662, "learning_rate": 1e-05, "loss": 0.5111, "mean_token_accuracy": 0.8476625680923462, "num_tokens": 407242467.0, "step": 10720 }, { "epoch": 1.3638213967688588, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5626368522644043, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8738427758216858, "num_tokens": 407279965.0, "step": 10721 }, { "epoch": 1.3639486070474494, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.74996018409729, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8809509873390198, "num_tokens": 407316086.0, "step": 10722 }, { "epoch": 1.3640758173260399, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5600318908691406, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8612055778503418, "num_tokens": 407355433.0, "step": 10723 }, { "epoch": 1.3642030276046304, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.727536201477051, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8769475221633911, "num_tokens": 407392919.0, "step": 10724 }, { "epoch": 1.364330237883221, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002422332763671875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.570711374282837, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8668718934059143, "num_tokens": 407428089.0, "step": 10725 }, { "epoch": 1.3644574481618115, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.706914186477661, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8782769441604614, "num_tokens": 407467232.0, "step": 10726 }, { "epoch": 1.364584658440402, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.566871166229248, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8633284568786621, "num_tokens": 407504975.0, "step": 10727 }, { "epoch": 1.3647118687189925, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7040557861328125, "learning_rate": 1e-05, "loss": 0.3702, "mean_token_accuracy": 0.8916758298873901, "num_tokens": 407535308.0, "step": 10728 }, { "epoch": 1.364839078997583, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.554326295852661, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.873183012008667, "num_tokens": 407578956.0, "step": 10729 }, { "epoch": 1.3649662892761736, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6928584575653076, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8744373321533203, "num_tokens": 407617633.0, "step": 10730 }, { "epoch": 1.3650934995547641, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5776336193084717, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.873927652835846, "num_tokens": 407646322.0, "step": 10731 }, { "epoch": 1.3652207098333546, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.676602602005005, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8620667457580566, "num_tokens": 407684957.0, "step": 10732 }, { "epoch": 1.3653479201119452, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.554023027420044, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8752550482749939, "num_tokens": 407722917.0, "step": 10733 }, { "epoch": 1.3654751303905355, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.67331862449646, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8719279766082764, "num_tokens": 407759599.0, "step": 10734 }, { "epoch": 1.365602340669126, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5528063774108887, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8677961826324463, "num_tokens": 407796416.0, "step": 10735 }, { "epoch": 1.3657295509477165, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.672609567642212, "learning_rate": 1e-05, "loss": 0.5214, "mean_token_accuracy": 0.8486751317977905, "num_tokens": 407841211.0, "step": 10736 }, { "epoch": 1.365856761226307, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5479180812835693, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.869365930557251, "num_tokens": 407879409.0, "step": 10737 }, { "epoch": 1.3659839715048976, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6730337142944336, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8675104975700378, "num_tokens": 407915275.0, "step": 10738 }, { "epoch": 1.3661111817834881, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.54476261138916, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8736027479171753, "num_tokens": 407953641.0, "step": 10739 }, { "epoch": 1.3662383920620786, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6691195964813232, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8772133588790894, "num_tokens": 407992662.0, "step": 10740 }, { "epoch": 1.3663656023406692, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.552642583847046, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8789225816726685, "num_tokens": 408026005.0, "step": 10741 }, { "epoch": 1.3664928126192597, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6676931381225586, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8712537288665771, "num_tokens": 408065250.0, "step": 10742 }, { "epoch": 1.3666200228978502, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5357584953308105, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8706807494163513, "num_tokens": 408109333.0, "step": 10743 }, { "epoch": 1.3667472331764405, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6801183223724365, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8755767345428467, "num_tokens": 408142416.0, "step": 10744 }, { "epoch": 1.366874443455031, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.52947735786438, "learning_rate": 1e-05, "loss": 0.3842, "mean_token_accuracy": 0.8880504369735718, "num_tokens": 408184335.0, "step": 10745 }, { "epoch": 1.3670016537336216, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6936566829681396, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8705244064331055, "num_tokens": 408221585.0, "step": 10746 }, { "epoch": 1.3671288640122121, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5312087535858154, "learning_rate": 1e-05, "loss": 0.4935, "mean_token_accuracy": 0.8483410477638245, "num_tokens": 408262105.0, "step": 10747 }, { "epoch": 1.3672560742908026, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6992621421813965, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8742756843566895, "num_tokens": 408302365.0, "step": 10748 }, { "epoch": 1.3673832845693932, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5260519981384277, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8651269674301147, "num_tokens": 408342033.0, "step": 10749 }, { "epoch": 1.3675104948479837, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7066547870635986, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8750072121620178, "num_tokens": 408381512.0, "step": 10750 }, { "epoch": 1.3676377051265742, "ewc_loss": 0.0615234375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5122408866882324, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8697412014007568, "num_tokens": 408423867.0, "step": 10751 }, { "epoch": 1.3677649154051648, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7161154747009277, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8737496733665466, "num_tokens": 408468417.0, "step": 10752 }, { "epoch": 1.3678921256837553, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5160391330718994, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8506761789321899, "num_tokens": 408514628.0, "step": 10753 }, { "epoch": 1.3680193359623458, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.736924409866333, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8588780760765076, "num_tokens": 408551936.0, "step": 10754 }, { "epoch": 1.3681465462409363, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.538264274597168, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8673454523086548, "num_tokens": 408581713.0, "step": 10755 }, { "epoch": 1.3682737565195269, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.72485613822937, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.879821240901947, "num_tokens": 408618265.0, "step": 10756 }, { "epoch": 1.3684009667981174, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.526228189468384, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8709703683853149, "num_tokens": 408655701.0, "step": 10757 }, { "epoch": 1.368528177076708, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7380409240722656, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8617058992385864, "num_tokens": 408690281.0, "step": 10758 }, { "epoch": 1.3686553873552982, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5185341835021973, "learning_rate": 1e-05, "loss": 0.5028, "mean_token_accuracy": 0.8526864051818848, "num_tokens": 408733483.0, "step": 10759 }, { "epoch": 1.3687825976338888, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7376182079315186, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.87200927734375, "num_tokens": 408770596.0, "step": 10760 }, { "epoch": 1.3689098079124793, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.523165464401245, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8811508417129517, "num_tokens": 408808265.0, "step": 10761 }, { "epoch": 1.3690370181910698, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7400288581848145, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8805869221687317, "num_tokens": 408847205.0, "step": 10762 }, { "epoch": 1.3691642284696603, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.515751600265503, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8669153451919556, "num_tokens": 408885744.0, "step": 10763 }, { "epoch": 1.3692914387482509, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7547874450683594, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8782341480255127, "num_tokens": 408925393.0, "step": 10764 }, { "epoch": 1.3694186490268414, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.530492067337036, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8677208423614502, "num_tokens": 408964995.0, "step": 10765 }, { "epoch": 1.369545859305432, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.769864797592163, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8794152736663818, "num_tokens": 408996336.0, "step": 10766 }, { "epoch": 1.3696730695840225, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5345630645751953, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.853676438331604, "num_tokens": 409034972.0, "step": 10767 }, { "epoch": 1.3698002798626128, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.767719268798828, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8841051459312439, "num_tokens": 409069421.0, "step": 10768 }, { "epoch": 1.3699274901412033, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.528618574142456, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8643742799758911, "num_tokens": 409104763.0, "step": 10769 }, { "epoch": 1.3700547004197938, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7666962146759033, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8752866387367249, "num_tokens": 409142799.0, "step": 10770 }, { "epoch": 1.3701819106983844, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.538280725479126, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8626072406768799, "num_tokens": 409178861.0, "step": 10771 }, { "epoch": 1.3703091209769749, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7633590698242188, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8736280202865601, "num_tokens": 409218628.0, "step": 10772 }, { "epoch": 1.3704363312555654, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.550262212753296, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8528352975845337, "num_tokens": 409255209.0, "step": 10773 }, { "epoch": 1.370563541534156, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7437856197357178, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8685221076011658, "num_tokens": 409295807.0, "step": 10774 }, { "epoch": 1.3706907518127465, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5546600818634033, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8742122650146484, "num_tokens": 409329263.0, "step": 10775 }, { "epoch": 1.370817962091337, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7295501232147217, "learning_rate": 1e-05, "loss": 0.3877, "mean_token_accuracy": 0.8897836208343506, "num_tokens": 409365571.0, "step": 10776 }, { "epoch": 1.3709451723699275, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5358517169952393, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8679326772689819, "num_tokens": 409406096.0, "step": 10777 }, { "epoch": 1.371072382648518, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7256667613983154, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8595647215843201, "num_tokens": 409450436.0, "step": 10778 }, { "epoch": 1.3711995929271086, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.550348997116089, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8697179555892944, "num_tokens": 409485134.0, "step": 10779 }, { "epoch": 1.371326803205699, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.728745460510254, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8724550604820251, "num_tokens": 409522890.0, "step": 10780 }, { "epoch": 1.3714540134842896, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.541938543319702, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8670598864555359, "num_tokens": 409561537.0, "step": 10781 }, { "epoch": 1.3715812237628802, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7224833965301514, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8817312717437744, "num_tokens": 409595259.0, "step": 10782 }, { "epoch": 1.3717084340414705, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5360805988311768, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8786033391952515, "num_tokens": 409633978.0, "step": 10783 }, { "epoch": 1.371835644320061, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.731308937072754, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8713732361793518, "num_tokens": 409671320.0, "step": 10784 }, { "epoch": 1.3719628545986515, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.528852939605713, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.8747965097427368, "num_tokens": 409712816.0, "step": 10785 }, { "epoch": 1.372090064877242, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.739525556564331, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8687194585800171, "num_tokens": 409751918.0, "step": 10786 }, { "epoch": 1.3722172751558326, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5378594398498535, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8862036466598511, "num_tokens": 409786905.0, "step": 10787 }, { "epoch": 1.3723444854344231, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.729675054550171, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8707488775253296, "num_tokens": 409827634.0, "step": 10788 }, { "epoch": 1.3724716957130136, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5436179637908936, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8753051161766052, "num_tokens": 409863060.0, "step": 10789 }, { "epoch": 1.3725989059916042, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7252466678619385, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8776102066040039, "num_tokens": 409896723.0, "step": 10790 }, { "epoch": 1.3727261162701947, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5362489223480225, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8710711598396301, "num_tokens": 409935912.0, "step": 10791 }, { "epoch": 1.3728533265487852, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.769972562789917, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8746140003204346, "num_tokens": 409973381.0, "step": 10792 }, { "epoch": 1.3729805368273755, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5297322273254395, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8636739253997803, "num_tokens": 410010548.0, "step": 10793 }, { "epoch": 1.373107747105966, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.743711233139038, "learning_rate": 1e-05, "loss": 0.3983, "mean_token_accuracy": 0.8851135969161987, "num_tokens": 410046456.0, "step": 10794 }, { "epoch": 1.3732349573845566, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.52531361579895, "learning_rate": 1e-05, "loss": 0.3828, "mean_token_accuracy": 0.8871871829032898, "num_tokens": 410080568.0, "step": 10795 }, { "epoch": 1.3733621676631471, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7365972995758057, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8776799440383911, "num_tokens": 410121008.0, "step": 10796 }, { "epoch": 1.3734893779417376, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5191144943237305, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8750016689300537, "num_tokens": 410165872.0, "step": 10797 }, { "epoch": 1.3736165882203282, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.743476152420044, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8760548830032349, "num_tokens": 410207379.0, "step": 10798 }, { "epoch": 1.3737437984989187, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5374014377593994, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8602806329727173, "num_tokens": 410246729.0, "step": 10799 }, { "epoch": 1.3738710087775092, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7439863681793213, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8763266205787659, "num_tokens": 410281237.0, "step": 10800 }, { "epoch": 1.3739982190560998, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5395452976226807, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8662306070327759, "num_tokens": 410318485.0, "step": 10801 }, { "epoch": 1.3741254293346903, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7405731678009033, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8707296252250671, "num_tokens": 410357995.0, "step": 10802 }, { "epoch": 1.3742526396132808, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.546292304992676, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8694882392883301, "num_tokens": 410395672.0, "step": 10803 }, { "epoch": 1.3743798498918713, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.746652841567993, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8494971990585327, "num_tokens": 410433566.0, "step": 10804 }, { "epoch": 1.3745070601704619, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5326695442199707, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8700153231620789, "num_tokens": 410473622.0, "step": 10805 }, { "epoch": 1.3746342704490524, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.742791175842285, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8770116567611694, "num_tokens": 410507155.0, "step": 10806 }, { "epoch": 1.374761480727643, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5396742820739746, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8673853278160095, "num_tokens": 410543352.0, "step": 10807 }, { "epoch": 1.3748886910062332, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7335402965545654, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8777626156806946, "num_tokens": 410584560.0, "step": 10808 }, { "epoch": 1.3750159012848238, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.544569492340088, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8803322911262512, "num_tokens": 410615948.0, "step": 10809 }, { "epoch": 1.3751431115634143, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7318408489227295, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8729376196861267, "num_tokens": 410649513.0, "step": 10810 }, { "epoch": 1.3752703218420048, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.534104824066162, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8691490292549133, "num_tokens": 410693087.0, "step": 10811 }, { "epoch": 1.3753975321205953, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.723586320877075, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8831980228424072, "num_tokens": 410729200.0, "step": 10812 }, { "epoch": 1.3755247423991859, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5366997718811035, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8656455278396606, "num_tokens": 410769830.0, "step": 10813 }, { "epoch": 1.3756519526777764, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.721672534942627, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8677207827568054, "num_tokens": 410815570.0, "step": 10814 }, { "epoch": 1.375779162956367, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.528315544128418, "learning_rate": 1e-05, "loss": 0.3655, "mean_token_accuracy": 0.8949906826019287, "num_tokens": 410852777.0, "step": 10815 }, { "epoch": 1.3759063732349575, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7419137954711914, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8753010630607605, "num_tokens": 410886809.0, "step": 10816 }, { "epoch": 1.3760335835135478, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5316081047058105, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8746559619903564, "num_tokens": 410924631.0, "step": 10817 }, { "epoch": 1.3761607937921383, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.738201856613159, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.872453510761261, "num_tokens": 410965212.0, "step": 10818 }, { "epoch": 1.3762880040707288, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.540949583053589, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8589921593666077, "num_tokens": 411003429.0, "step": 10819 }, { "epoch": 1.3764152143493193, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.729832410812378, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.857077956199646, "num_tokens": 411043121.0, "step": 10820 }, { "epoch": 1.3765424246279099, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5443880558013916, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8607975244522095, "num_tokens": 411082146.0, "step": 10821 }, { "epoch": 1.3766696349065004, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7363688945770264, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.866877555847168, "num_tokens": 411116070.0, "step": 10822 }, { "epoch": 1.376796845185091, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.527686595916748, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8663722276687622, "num_tokens": 411162066.0, "step": 10823 }, { "epoch": 1.3769240554636815, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.736401319503784, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8753902316093445, "num_tokens": 411200939.0, "step": 10824 }, { "epoch": 1.377051265742272, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5353245735168457, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8659551739692688, "num_tokens": 411241161.0, "step": 10825 }, { "epoch": 1.3771784760208625, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7281136512756348, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8689534068107605, "num_tokens": 411284316.0, "step": 10826 }, { "epoch": 1.377305686299453, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5579137802124023, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8542238473892212, "num_tokens": 411320795.0, "step": 10827 }, { "epoch": 1.3774328965780436, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.722014904022217, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8750341534614563, "num_tokens": 411357513.0, "step": 10828 }, { "epoch": 1.377560106856634, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5265750885009766, "learning_rate": 1e-05, "loss": 0.375, "mean_token_accuracy": 0.8916463851928711, "num_tokens": 411395310.0, "step": 10829 }, { "epoch": 1.3776873171352246, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7264909744262695, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8700328469276428, "num_tokens": 411437774.0, "step": 10830 }, { "epoch": 1.3778145274138152, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5406928062438965, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8803648948669434, "num_tokens": 411471399.0, "step": 10831 }, { "epoch": 1.3779417376924055, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7387843132019043, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8596712946891785, "num_tokens": 411505289.0, "step": 10832 }, { "epoch": 1.378068947970996, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5338356494903564, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8815127015113831, "num_tokens": 411541463.0, "step": 10833 }, { "epoch": 1.3781961582495865, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7340025901794434, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8687957525253296, "num_tokens": 411577276.0, "step": 10834 }, { "epoch": 1.378323368528177, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.567885160446167, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8430905342102051, "num_tokens": 411610417.0, "step": 10835 }, { "epoch": 1.3784505788067676, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7250781059265137, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8522546291351318, "num_tokens": 411646539.0, "step": 10836 }, { "epoch": 1.378577789085358, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5321924686431885, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8648938536643982, "num_tokens": 411685890.0, "step": 10837 }, { "epoch": 1.3787049993639486, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7069625854492188, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8801639080047607, "num_tokens": 411725566.0, "step": 10838 }, { "epoch": 1.3788322096425392, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.534360408782959, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8771871328353882, "num_tokens": 411767256.0, "step": 10839 }, { "epoch": 1.3789594199211297, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7212679386138916, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8700466752052307, "num_tokens": 411802210.0, "step": 10840 }, { "epoch": 1.3790866301997202, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5240771770477295, "learning_rate": 1e-05, "loss": 0.3733, "mean_token_accuracy": 0.8902276158332825, "num_tokens": 411842011.0, "step": 10841 }, { "epoch": 1.3792138404783105, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7319717407226562, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8768970966339111, "num_tokens": 411877137.0, "step": 10842 }, { "epoch": 1.379341050756901, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5380983352661133, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8787840008735657, "num_tokens": 411909130.0, "step": 10843 }, { "epoch": 1.3794682610354916, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7406511306762695, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8663708567619324, "num_tokens": 411942542.0, "step": 10844 }, { "epoch": 1.3795954713140821, "ewc_loss": 0.0625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5412847995758057, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8626375198364258, "num_tokens": 411974272.0, "step": 10845 }, { "epoch": 1.3797226815926726, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7171382904052734, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8801747560501099, "num_tokens": 412012748.0, "step": 10846 }, { "epoch": 1.3798498918712632, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5337705612182617, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8662694692611694, "num_tokens": 412051171.0, "step": 10847 }, { "epoch": 1.3799771021498537, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.732670783996582, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.8585677146911621, "num_tokens": 412090267.0, "step": 10848 }, { "epoch": 1.3801043124284442, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5597829818725586, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8633215427398682, "num_tokens": 412118471.0, "step": 10849 }, { "epoch": 1.3802315227070348, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00024318695068359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.698779582977295, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8762328624725342, "num_tokens": 412163639.0, "step": 10850 }, { "epoch": 1.3803587329856253, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.539297580718994, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8593716621398926, "num_tokens": 412203685.0, "step": 10851 }, { "epoch": 1.3804859432642158, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7099432945251465, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8722864985466003, "num_tokens": 412238810.0, "step": 10852 }, { "epoch": 1.3806131535428063, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5695672035217285, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8685911893844604, "num_tokens": 412273840.0, "step": 10853 }, { "epoch": 1.3807403638213969, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.715482234954834, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8691949844360352, "num_tokens": 412309931.0, "step": 10854 }, { "epoch": 1.3808675740999874, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5570273399353027, "learning_rate": 1e-05, "loss": 0.5283, "mean_token_accuracy": 0.8399158120155334, "num_tokens": 412346934.0, "step": 10855 }, { "epoch": 1.380994784378578, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6928679943084717, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8682578802108765, "num_tokens": 412387942.0, "step": 10856 }, { "epoch": 1.3811219946571682, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.534175157546997, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8718950152397156, "num_tokens": 412429726.0, "step": 10857 }, { "epoch": 1.3812492049357588, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6869847774505615, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.874714732170105, "num_tokens": 412472711.0, "step": 10858 }, { "epoch": 1.3813764152143493, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5378177165985107, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8797861337661743, "num_tokens": 412511384.0, "step": 10859 }, { "epoch": 1.3815036254929398, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.701267719268799, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8562263250350952, "num_tokens": 412552991.0, "step": 10860 }, { "epoch": 1.3816308357715303, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.554370641708374, "learning_rate": 1e-05, "loss": 0.5415, "mean_token_accuracy": 0.8363035321235657, "num_tokens": 412593794.0, "step": 10861 }, { "epoch": 1.3817580460501209, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6966094970703125, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8797604441642761, "num_tokens": 412625569.0, "step": 10862 }, { "epoch": 1.3818852563287114, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.542083263397217, "learning_rate": 1e-05, "loss": 0.3862, "mean_token_accuracy": 0.8857494592666626, "num_tokens": 412658086.0, "step": 10863 }, { "epoch": 1.382012466607302, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6958141326904297, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8711574673652649, "num_tokens": 412698642.0, "step": 10864 }, { "epoch": 1.3821396768858925, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5397815704345703, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8740218877792358, "num_tokens": 412738327.0, "step": 10865 }, { "epoch": 1.3822668871644828, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6996452808380127, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8712986707687378, "num_tokens": 412773420.0, "step": 10866 }, { "epoch": 1.3823940974430733, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5457849502563477, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8770040273666382, "num_tokens": 412805383.0, "step": 10867 }, { "epoch": 1.3825213077216638, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.683706283569336, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8773066997528076, "num_tokens": 412846924.0, "step": 10868 }, { "epoch": 1.3826485180002543, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.546388626098633, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8620025515556335, "num_tokens": 412883232.0, "step": 10869 }, { "epoch": 1.3827757282788449, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.689438581466675, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8788729906082153, "num_tokens": 412922110.0, "step": 10870 }, { "epoch": 1.3829029385574354, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5405309200286865, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8664277791976929, "num_tokens": 412963661.0, "step": 10871 }, { "epoch": 1.383030148836026, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6817963123321533, "learning_rate": 1e-05, "loss": 0.3617, "mean_token_accuracy": 0.896178662776947, "num_tokens": 413002468.0, "step": 10872 }, { "epoch": 1.3831573591146165, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.544252634048462, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8565651774406433, "num_tokens": 413048135.0, "step": 10873 }, { "epoch": 1.383284569393207, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6971607208251953, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8715223670005798, "num_tokens": 413082552.0, "step": 10874 }, { "epoch": 1.3834117796717975, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5337584018707275, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8643146753311157, "num_tokens": 413129230.0, "step": 10875 }, { "epoch": 1.383538989950388, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.700059175491333, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.88456130027771, "num_tokens": 413168830.0, "step": 10876 }, { "epoch": 1.3836662002289786, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.523066520690918, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8719300031661987, "num_tokens": 413208423.0, "step": 10877 }, { "epoch": 1.383793410507569, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7357654571533203, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8705747127532959, "num_tokens": 413238133.0, "step": 10878 }, { "epoch": 1.3839206207861596, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5255203247070312, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8773438334465027, "num_tokens": 413273653.0, "step": 10879 }, { "epoch": 1.3840478310647502, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7228171825408936, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8725889325141907, "num_tokens": 413316492.0, "step": 10880 }, { "epoch": 1.3841750413433405, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.509941577911377, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8811562061309814, "num_tokens": 413359101.0, "step": 10881 }, { "epoch": 1.384302251621931, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7503862380981445, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8820958137512207, "num_tokens": 413396618.0, "step": 10882 }, { "epoch": 1.3844294619005215, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.511789560317993, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8824936151504517, "num_tokens": 413431863.0, "step": 10883 }, { "epoch": 1.384556672179112, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7635180950164795, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.867501974105835, "num_tokens": 413470137.0, "step": 10884 }, { "epoch": 1.3846838824577026, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.505753517150879, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8821471929550171, "num_tokens": 413508475.0, "step": 10885 }, { "epoch": 1.384811092736293, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7817559242248535, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8695167303085327, "num_tokens": 413548846.0, "step": 10886 }, { "epoch": 1.3849383030148836, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.512424945831299, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8687698841094971, "num_tokens": 413587375.0, "step": 10887 }, { "epoch": 1.3850655132934742, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.8055663108825684, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8735260963439941, "num_tokens": 413626701.0, "step": 10888 }, { "epoch": 1.3851927235720647, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.52646541595459, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.8561415076255798, "num_tokens": 413665207.0, "step": 10889 }, { "epoch": 1.385319933850655, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.8041934967041016, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8856896162033081, "num_tokens": 413700549.0, "step": 10890 }, { "epoch": 1.3854471441292455, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5443170070648193, "learning_rate": 1e-05, "loss": 0.5277, "mean_token_accuracy": 0.8421101570129395, "num_tokens": 413738308.0, "step": 10891 }, { "epoch": 1.385574354407836, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7879645824432373, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.873728334903717, "num_tokens": 413779540.0, "step": 10892 }, { "epoch": 1.3857015646864266, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.531726837158203, "learning_rate": 1e-05, "loss": 0.3785, "mean_token_accuracy": 0.8886016607284546, "num_tokens": 413818677.0, "step": 10893 }, { "epoch": 1.385828774965017, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.779398202896118, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8676170706748962, "num_tokens": 413861397.0, "step": 10894 }, { "epoch": 1.3859559852436076, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.545023202896118, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8769149780273438, "num_tokens": 413897629.0, "step": 10895 }, { "epoch": 1.3860831955221982, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7913360595703125, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8542767763137817, "num_tokens": 413935711.0, "step": 10896 }, { "epoch": 1.3862104058007887, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.540654182434082, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8799435496330261, "num_tokens": 413977340.0, "step": 10897 }, { "epoch": 1.3863376160793792, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.779667854309082, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8809354305267334, "num_tokens": 414014429.0, "step": 10898 }, { "epoch": 1.3864648263579697, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5495803356170654, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8595651388168335, "num_tokens": 414055426.0, "step": 10899 }, { "epoch": 1.3865920366365603, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.784712791442871, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8635512590408325, "num_tokens": 414086549.0, "step": 10900 }, { "epoch": 1.3867192469151508, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5590577125549316, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.877142071723938, "num_tokens": 414121767.0, "step": 10901 }, { "epoch": 1.3868464571937413, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7755379676818848, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8649195432662964, "num_tokens": 414154201.0, "step": 10902 }, { "epoch": 1.3869736674723319, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.573280096054077, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8589361310005188, "num_tokens": 414189898.0, "step": 10903 }, { "epoch": 1.3871008777509224, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7382431030273438, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8744750022888184, "num_tokens": 414226507.0, "step": 10904 }, { "epoch": 1.387228088029513, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5583393573760986, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8664370775222778, "num_tokens": 414261909.0, "step": 10905 }, { "epoch": 1.3873552983081032, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7256133556365967, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8640928864479065, "num_tokens": 414303296.0, "step": 10906 }, { "epoch": 1.3874825085866938, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5683510303497314, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8689453601837158, "num_tokens": 414340638.0, "step": 10907 }, { "epoch": 1.3876097188652843, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7109856605529785, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8693143725395203, "num_tokens": 414376900.0, "step": 10908 }, { "epoch": 1.3877369291438748, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5660593509674072, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8740754127502441, "num_tokens": 414410657.0, "step": 10909 }, { "epoch": 1.3878641394224653, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.706641912460327, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8554236888885498, "num_tokens": 414443165.0, "step": 10910 }, { "epoch": 1.3879913497010559, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.557806968688965, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8572840690612793, "num_tokens": 414482185.0, "step": 10911 }, { "epoch": 1.3881185599796464, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7072718143463135, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8578919172286987, "num_tokens": 414517509.0, "step": 10912 }, { "epoch": 1.388245770258237, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5366265773773193, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8728695511817932, "num_tokens": 414558650.0, "step": 10913 }, { "epoch": 1.3883729805368275, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7086868286132812, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8753650784492493, "num_tokens": 414594301.0, "step": 10914 }, { "epoch": 1.3885001908154178, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5467464923858643, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8777689933776855, "num_tokens": 414626895.0, "step": 10915 }, { "epoch": 1.3886274010940083, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.702141284942627, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8717725872993469, "num_tokens": 414664708.0, "step": 10916 }, { "epoch": 1.3887546113725988, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.539048433303833, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8521149158477783, "num_tokens": 414706243.0, "step": 10917 }, { "epoch": 1.3888818216511893, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.712655782699585, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8709605932235718, "num_tokens": 414740681.0, "step": 10918 }, { "epoch": 1.3890090319297799, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.542710065841675, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8737160563468933, "num_tokens": 414774942.0, "step": 10919 }, { "epoch": 1.3891362422083704, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.690584659576416, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8809508085250854, "num_tokens": 414817369.0, "step": 10920 }, { "epoch": 1.389263452486961, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5483624935150146, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8668080568313599, "num_tokens": 414855503.0, "step": 10921 }, { "epoch": 1.3893906627655515, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7021331787109375, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.86539626121521, "num_tokens": 414891974.0, "step": 10922 }, { "epoch": 1.389517873044142, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.536269187927246, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.881605863571167, "num_tokens": 414926773.0, "step": 10923 }, { "epoch": 1.3896450833227325, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.702697992324829, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8796706795692444, "num_tokens": 414962848.0, "step": 10924 }, { "epoch": 1.389772293601323, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5296263694763184, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8755829930305481, "num_tokens": 415003341.0, "step": 10925 }, { "epoch": 1.3898995038799136, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7069547176361084, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8814303874969482, "num_tokens": 415036836.0, "step": 10926 }, { "epoch": 1.390026714158504, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.521522283554077, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.87229323387146, "num_tokens": 415078589.0, "step": 10927 }, { "epoch": 1.3901539244370946, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7121834754943848, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8851937055587769, "num_tokens": 415114916.0, "step": 10928 }, { "epoch": 1.3902811347156852, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5265634059906006, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8688274621963501, "num_tokens": 415153774.0, "step": 10929 }, { "epoch": 1.3904083449942755, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7396955490112305, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8574832677841187, "num_tokens": 415188359.0, "step": 10930 }, { "epoch": 1.390535555272866, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.522414445877075, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.876886785030365, "num_tokens": 415224423.0, "step": 10931 }, { "epoch": 1.3906627655514565, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.735630989074707, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8643794655799866, "num_tokens": 415263219.0, "step": 10932 }, { "epoch": 1.390789975830047, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5322041511535645, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8689932823181152, "num_tokens": 415301043.0, "step": 10933 }, { "epoch": 1.3909171861086376, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.74025821685791, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8695360422134399, "num_tokens": 415339128.0, "step": 10934 }, { "epoch": 1.391044396387228, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5304830074310303, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8702025413513184, "num_tokens": 415374574.0, "step": 10935 }, { "epoch": 1.3911716066658186, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.759601354598999, "learning_rate": 1e-05, "loss": 0.5213, "mean_token_accuracy": 0.8443282842636108, "num_tokens": 415410133.0, "step": 10936 }, { "epoch": 1.3912988169444092, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5282692909240723, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.859548032283783, "num_tokens": 415451329.0, "step": 10937 }, { "epoch": 1.3914260272229997, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7489287853240967, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8698279857635498, "num_tokens": 415491448.0, "step": 10938 }, { "epoch": 1.39155323750159, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.51725697517395, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8710606694221497, "num_tokens": 415533583.0, "step": 10939 }, { "epoch": 1.3916804477801805, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7761337757110596, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8599876165390015, "num_tokens": 415567985.0, "step": 10940 }, { "epoch": 1.391807658058771, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5313167572021484, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8724936246871948, "num_tokens": 415603130.0, "step": 10941 }, { "epoch": 1.3919348683373616, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.77710223197937, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8622403144836426, "num_tokens": 415641284.0, "step": 10942 }, { "epoch": 1.392062078615952, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5384693145751953, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8566861152648926, "num_tokens": 415682483.0, "step": 10943 }, { "epoch": 1.3921892888945426, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.769578695297241, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8542555570602417, "num_tokens": 415723364.0, "step": 10944 }, { "epoch": 1.3923164991731332, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5380373001098633, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8608336448669434, "num_tokens": 415762750.0, "step": 10945 }, { "epoch": 1.3924437094517237, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7641234397888184, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8603508472442627, "num_tokens": 415801363.0, "step": 10946 }, { "epoch": 1.3925709197303142, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5443568229675293, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.873578667640686, "num_tokens": 415836121.0, "step": 10947 }, { "epoch": 1.3926981300089047, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.749265432357788, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8589271903038025, "num_tokens": 415880145.0, "step": 10948 }, { "epoch": 1.3928253402874953, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5399038791656494, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.869681715965271, "num_tokens": 415921244.0, "step": 10949 }, { "epoch": 1.3929525505660858, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7559397220611572, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8707579970359802, "num_tokens": 415957166.0, "step": 10950 }, { "epoch": 1.3930797608446763, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.537888526916504, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8618801236152649, "num_tokens": 415999462.0, "step": 10951 }, { "epoch": 1.3932069711232669, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7493553161621094, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8704149127006531, "num_tokens": 416038619.0, "step": 10952 }, { "epoch": 1.3933341814018574, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5330722332000732, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8668485283851624, "num_tokens": 416084885.0, "step": 10953 }, { "epoch": 1.393461391680448, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7488656044006348, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8592095375061035, "num_tokens": 416126863.0, "step": 10954 }, { "epoch": 1.3935886019590382, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5380783081054688, "learning_rate": 1e-05, "loss": 0.3647, "mean_token_accuracy": 0.8942996263504028, "num_tokens": 416160888.0, "step": 10955 }, { "epoch": 1.3937158122376287, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7502660751342773, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8761701583862305, "num_tokens": 416199012.0, "step": 10956 }, { "epoch": 1.3938430225162193, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5454559326171875, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8799980878829956, "num_tokens": 416234078.0, "step": 10957 }, { "epoch": 1.3939702327948098, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.73628830909729, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8769413232803345, "num_tokens": 416272912.0, "step": 10958 }, { "epoch": 1.3940974430734003, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.544484853744507, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.879790723323822, "num_tokens": 416312829.0, "step": 10959 }, { "epoch": 1.3942246533519909, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7251367568969727, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8710463047027588, "num_tokens": 416353337.0, "step": 10960 }, { "epoch": 1.3943518636305814, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5532662868499756, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.866515040397644, "num_tokens": 416392425.0, "step": 10961 }, { "epoch": 1.394479073909172, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7157585620880127, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8596940636634827, "num_tokens": 416434641.0, "step": 10962 }, { "epoch": 1.3946062841877624, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.544116973876953, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8674373626708984, "num_tokens": 416479107.0, "step": 10963 }, { "epoch": 1.3947334944663528, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7285568714141846, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8653596043586731, "num_tokens": 416512523.0, "step": 10964 }, { "epoch": 1.3948607047449433, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.547449827194214, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8700448870658875, "num_tokens": 416550176.0, "step": 10965 }, { "epoch": 1.3949879150235338, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7166388034820557, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8703141212463379, "num_tokens": 416583878.0, "step": 10966 }, { "epoch": 1.3951151253021243, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.569121837615967, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8666725158691406, "num_tokens": 416612863.0, "step": 10967 }, { "epoch": 1.3952423355807149, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7064011096954346, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8574730157852173, "num_tokens": 416650691.0, "step": 10968 }, { "epoch": 1.3953695458593054, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.541001558303833, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8746559619903564, "num_tokens": 416687227.0, "step": 10969 }, { "epoch": 1.395496756137896, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7857768535614014, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8626523017883301, "num_tokens": 416725712.0, "step": 10970 }, { "epoch": 1.3956239664164865, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5447869300842285, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8788770437240601, "num_tokens": 416756169.0, "step": 10971 }, { "epoch": 1.395751176695077, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7378480434417725, "learning_rate": 1e-05, "loss": 0.5016, "mean_token_accuracy": 0.8492303490638733, "num_tokens": 416794351.0, "step": 10972 }, { "epoch": 1.3958783869736675, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5331180095672607, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8841407299041748, "num_tokens": 416829201.0, "step": 10973 }, { "epoch": 1.396005597252258, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7461180686950684, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.877082109451294, "num_tokens": 416862171.0, "step": 10974 }, { "epoch": 1.3961328075308486, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.55051851272583, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8596127033233643, "num_tokens": 416895923.0, "step": 10975 }, { "epoch": 1.396260017809439, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7427923679351807, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8646187782287598, "num_tokens": 416927441.0, "step": 10976 }, { "epoch": 1.3963872280880296, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.538949966430664, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8731300234794617, "num_tokens": 416962911.0, "step": 10977 }, { "epoch": 1.3965144383666201, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.74416446685791, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8639788627624512, "num_tokens": 417001403.0, "step": 10978 }, { "epoch": 1.3966416486452105, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.535766839981079, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8608720302581787, "num_tokens": 417038043.0, "step": 10979 }, { "epoch": 1.396768858923801, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7517943382263184, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8637946844100952, "num_tokens": 417073516.0, "step": 10980 }, { "epoch": 1.3968960692023915, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.538801908493042, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8680300116539001, "num_tokens": 417112334.0, "step": 10981 }, { "epoch": 1.397023279480982, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7348506450653076, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8713696002960205, "num_tokens": 417155273.0, "step": 10982 }, { "epoch": 1.3971504897595726, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.543734312057495, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8767279386520386, "num_tokens": 417190145.0, "step": 10983 }, { "epoch": 1.397277700038163, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.734222650527954, "learning_rate": 1e-05, "loss": 0.3927, "mean_token_accuracy": 0.883454442024231, "num_tokens": 417228184.0, "step": 10984 }, { "epoch": 1.3974049103167536, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5504567623138428, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8826677799224854, "num_tokens": 417258590.0, "step": 10985 }, { "epoch": 1.3975321205953442, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.718757390975952, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8809606432914734, "num_tokens": 417295969.0, "step": 10986 }, { "epoch": 1.3976593308739347, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5546305179595947, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8544212579727173, "num_tokens": 417334428.0, "step": 10987 }, { "epoch": 1.397786541152525, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.718823194503784, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8739942908287048, "num_tokens": 417370074.0, "step": 10988 }, { "epoch": 1.3979137514311155, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5409634113311768, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8755292892456055, "num_tokens": 417413544.0, "step": 10989 }, { "epoch": 1.398040961709706, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.731660842895508, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8663667440414429, "num_tokens": 417445893.0, "step": 10990 }, { "epoch": 1.3981681719882966, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.537278175354004, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8809500336647034, "num_tokens": 417481469.0, "step": 10991 }, { "epoch": 1.398295382266887, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7378275394439697, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8640635013580322, "num_tokens": 417517070.0, "step": 10992 }, { "epoch": 1.3984225925454776, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5358541011810303, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8662436604499817, "num_tokens": 417558265.0, "step": 10993 }, { "epoch": 1.3985498028240682, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.729388475418091, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8739758729934692, "num_tokens": 417599083.0, "step": 10994 }, { "epoch": 1.3986770131026587, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5414037704467773, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8812544941902161, "num_tokens": 417632425.0, "step": 10995 }, { "epoch": 1.3988042233812492, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7370455265045166, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8771794438362122, "num_tokens": 417667209.0, "step": 10996 }, { "epoch": 1.3989314336598397, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.533128023147583, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.874863862991333, "num_tokens": 417706168.0, "step": 10997 }, { "epoch": 1.3990586439384303, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7380049228668213, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8788348436355591, "num_tokens": 417745225.0, "step": 10998 }, { "epoch": 1.3991858542170208, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5593576431274414, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8763928413391113, "num_tokens": 417782688.0, "step": 10999 }, { "epoch": 1.3993130644956113, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7352311611175537, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8690692186355591, "num_tokens": 417814157.0, "step": 11000 }, { "epoch": 1.3994402747742019, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5576021671295166, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8627417683601379, "num_tokens": 417845460.0, "step": 11001 }, { "epoch": 1.3995674850527924, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7185781002044678, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8743194341659546, "num_tokens": 417881460.0, "step": 11002 }, { "epoch": 1.399694695331383, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5341739654541016, "learning_rate": 1e-05, "loss": 0.3843, "mean_token_accuracy": 0.8867799043655396, "num_tokens": 417919690.0, "step": 11003 }, { "epoch": 1.3998219056099732, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7170422077178955, "learning_rate": 1e-05, "loss": 0.3828, "mean_token_accuracy": 0.888352632522583, "num_tokens": 417959474.0, "step": 11004 }, { "epoch": 1.3999491158885637, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5522067546844482, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.869499146938324, "num_tokens": 417992342.0, "step": 11005 }, { "epoch": 1.4000763261671543, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.718858003616333, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8534817099571228, "num_tokens": 418032006.0, "step": 11006 }, { "epoch": 1.4002035364457448, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5464019775390625, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8587284088134766, "num_tokens": 418069318.0, "step": 11007 }, { "epoch": 1.4003307467243353, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7074151039123535, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8850809931755066, "num_tokens": 418103605.0, "step": 11008 }, { "epoch": 1.4004579570029259, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5503461360931396, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8652082085609436, "num_tokens": 418136363.0, "step": 11009 }, { "epoch": 1.4005851672815164, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6980502605438232, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8712693452835083, "num_tokens": 418178096.0, "step": 11010 }, { "epoch": 1.400712377560107, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.546656608581543, "learning_rate": 1e-05, "loss": 0.3954, "mean_token_accuracy": 0.8827239274978638, "num_tokens": 418211025.0, "step": 11011 }, { "epoch": 1.4008395878386974, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7073493003845215, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.873493492603302, "num_tokens": 418251210.0, "step": 11012 }, { "epoch": 1.4009667981172877, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5436880588531494, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8790237307548523, "num_tokens": 418286593.0, "step": 11013 }, { "epoch": 1.4010940083958783, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7033228874206543, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8824253082275391, "num_tokens": 418324794.0, "step": 11014 }, { "epoch": 1.4012212186744688, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.531036853790283, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8777123689651489, "num_tokens": 418366067.0, "step": 11015 }, { "epoch": 1.4013484289530593, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7010457515716553, "learning_rate": 1e-05, "loss": 0.3832, "mean_token_accuracy": 0.8894615769386292, "num_tokens": 418406600.0, "step": 11016 }, { "epoch": 1.4014756392316499, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.547451972961426, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8744690418243408, "num_tokens": 418440168.0, "step": 11017 }, { "epoch": 1.4016028495102404, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7034642696380615, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.853054404258728, "num_tokens": 418482247.0, "step": 11018 }, { "epoch": 1.401730059788831, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5514063835144043, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.881817102432251, "num_tokens": 418517376.0, "step": 11019 }, { "epoch": 1.4018572700674214, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.716874361038208, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8655311465263367, "num_tokens": 418555841.0, "step": 11020 }, { "epoch": 1.401984480346012, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.534296989440918, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8715313076972961, "num_tokens": 418594232.0, "step": 11021 }, { "epoch": 1.4021116906246025, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7082595825195312, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8797222971916199, "num_tokens": 418631726.0, "step": 11022 }, { "epoch": 1.402238900903193, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.521476984024048, "learning_rate": 1e-05, "loss": 0.3732, "mean_token_accuracy": 0.8939220905303955, "num_tokens": 418671172.0, "step": 11023 }, { "epoch": 1.4023661111817836, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.726207971572876, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8775213956832886, "num_tokens": 418706541.0, "step": 11024 }, { "epoch": 1.402493321460374, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5294158458709717, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8766779899597168, "num_tokens": 418743557.0, "step": 11025 }, { "epoch": 1.4026205317389646, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.726483106613159, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8698639869689941, "num_tokens": 418783741.0, "step": 11026 }, { "epoch": 1.4027477420175551, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5253143310546875, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8788097500801086, "num_tokens": 418822208.0, "step": 11027 }, { "epoch": 1.4028749522961454, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7338156700134277, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8762729167938232, "num_tokens": 418862717.0, "step": 11028 }, { "epoch": 1.403002162574736, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.545712471008301, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8751112818717957, "num_tokens": 418892961.0, "step": 11029 }, { "epoch": 1.4031293728533265, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7376160621643066, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8548591136932373, "num_tokens": 418933384.0, "step": 11030 }, { "epoch": 1.403256583131917, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.525268077850342, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.877116322517395, "num_tokens": 418971042.0, "step": 11031 }, { "epoch": 1.4033837934105076, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.738593578338623, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8752244710922241, "num_tokens": 419009949.0, "step": 11032 }, { "epoch": 1.403511003689098, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5310792922973633, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8777951002120972, "num_tokens": 419044607.0, "step": 11033 }, { "epoch": 1.4036382139676886, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7378461360931396, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8748955726623535, "num_tokens": 419081407.0, "step": 11034 }, { "epoch": 1.4037654242462791, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5578062534332275, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8800963163375854, "num_tokens": 419125007.0, "step": 11035 }, { "epoch": 1.4038926345248697, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.729952812194824, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8760295510292053, "num_tokens": 419166009.0, "step": 11036 }, { "epoch": 1.40401984480346, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5614755153656006, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8693761825561523, "num_tokens": 419194185.0, "step": 11037 }, { "epoch": 1.4041470550820505, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.704152822494507, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8763261437416077, "num_tokens": 419236742.0, "step": 11038 }, { "epoch": 1.404274265360641, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.548898696899414, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.872002363204956, "num_tokens": 419272477.0, "step": 11039 }, { "epoch": 1.4044014756392316, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7098937034606934, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8638666272163391, "num_tokens": 419309198.0, "step": 11040 }, { "epoch": 1.404528685917822, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5528409481048584, "learning_rate": 1e-05, "loss": 0.3983, "mean_token_accuracy": 0.878961443901062, "num_tokens": 419343065.0, "step": 11041 }, { "epoch": 1.4046558961964126, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7032856941223145, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8626219034194946, "num_tokens": 419382826.0, "step": 11042 }, { "epoch": 1.4047831064750032, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.556948184967041, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8635283708572388, "num_tokens": 419416839.0, "step": 11043 }, { "epoch": 1.4049103167535937, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.7102105617523193, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8728784322738647, "num_tokens": 419448041.0, "step": 11044 }, { "epoch": 1.4050375270321842, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5709848403930664, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8469477891921997, "num_tokens": 419484917.0, "step": 11045 }, { "epoch": 1.4051647373107747, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6879403591156006, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8711061477661133, "num_tokens": 419521329.0, "step": 11046 }, { "epoch": 1.4052919475893653, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.565480947494507, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8696386218070984, "num_tokens": 419556141.0, "step": 11047 }, { "epoch": 1.4054191578679558, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6748404502868652, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8738160133361816, "num_tokens": 419593027.0, "step": 11048 }, { "epoch": 1.4055463681465463, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.552726984024048, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.881370484828949, "num_tokens": 419632419.0, "step": 11049 }, { "epoch": 1.4056735784251368, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6672816276550293, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8676021099090576, "num_tokens": 419677625.0, "step": 11050 }, { "epoch": 1.4058007887037274, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5489401817321777, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8692796230316162, "num_tokens": 419721546.0, "step": 11051 }, { "epoch": 1.405927998982318, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.690669298171997, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8544533252716064, "num_tokens": 419759345.0, "step": 11052 }, { "epoch": 1.4060552092609082, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5647969245910645, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.868439793586731, "num_tokens": 419789819.0, "step": 11053 }, { "epoch": 1.4061824195394987, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6764841079711914, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8681211471557617, "num_tokens": 419830525.0, "step": 11054 }, { "epoch": 1.4063096298180893, "ewc_loss": 0.0625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.533766984939575, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8706499934196472, "num_tokens": 419875592.0, "step": 11055 }, { "epoch": 1.4064368400966798, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6820316314697266, "learning_rate": 1e-05, "loss": 0.3909, "mean_token_accuracy": 0.8882667422294617, "num_tokens": 419913530.0, "step": 11056 }, { "epoch": 1.4065640503752703, "ewc_loss": 0.062255859375, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5272207260131836, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8705945611000061, "num_tokens": 419955947.0, "step": 11057 }, { "epoch": 1.4066912606538609, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.709510087966919, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8822635412216187, "num_tokens": 419986931.0, "step": 11058 }, { "epoch": 1.4068184709324514, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5256614685058594, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8722240924835205, "num_tokens": 420024037.0, "step": 11059 }, { "epoch": 1.406945681211042, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7142066955566406, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.866603672504425, "num_tokens": 420062843.0, "step": 11060 }, { "epoch": 1.4070728914896324, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.526301145553589, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.85567307472229, "num_tokens": 420102311.0, "step": 11061 }, { "epoch": 1.4072001017682227, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7172482013702393, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8648569583892822, "num_tokens": 420143364.0, "step": 11062 }, { "epoch": 1.4073273120468133, "ewc_loss": 0.061767578125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.743171691894531e-05, "grad_norm": 2.5133445262908936, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8770706653594971, "num_tokens": 420180930.0, "step": 11063 }, { "epoch": 1.4074545223254038, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7318201065063477, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8721215724945068, "num_tokens": 420216048.0, "step": 11064 }, { "epoch": 1.4075817326039943, "ewc_loss": 0.06201171875, "ewc_loss_diag": 0.000244140625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5221633911132812, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8715753555297852, "num_tokens": 420250792.0, "step": 11065 }, { "epoch": 1.4077089428825849, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7592172622680664, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8764638900756836, "num_tokens": 420280518.0, "step": 11066 }, { "epoch": 1.4078361531611754, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5241639614105225, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8548742532730103, "num_tokens": 420318274.0, "step": 11067 }, { "epoch": 1.407963363439766, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.763171672821045, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8737752437591553, "num_tokens": 420353277.0, "step": 11068 }, { "epoch": 1.4080905737183564, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5340304374694824, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8781436681747437, "num_tokens": 420384457.0, "step": 11069 }, { "epoch": 1.408217783996947, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7666006088256836, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8609167337417603, "num_tokens": 420419229.0, "step": 11070 }, { "epoch": 1.4083449942755375, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.532541275024414, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8622366189956665, "num_tokens": 420457002.0, "step": 11071 }, { "epoch": 1.408472204554128, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.764364242553711, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8803858160972595, "num_tokens": 420490802.0, "step": 11072 }, { "epoch": 1.4085994148327186, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5181884765625, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8790889978408813, "num_tokens": 420531782.0, "step": 11073 }, { "epoch": 1.408726625111309, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7690610885620117, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8604336977005005, "num_tokens": 420575577.0, "step": 11074 }, { "epoch": 1.4088538353898996, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5337836742401123, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8741155862808228, "num_tokens": 420612963.0, "step": 11075 }, { "epoch": 1.4089810456684901, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.769925832748413, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8805783987045288, "num_tokens": 420649518.0, "step": 11076 }, { "epoch": 1.4091082559470804, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5601820945739746, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.862206757068634, "num_tokens": 420684377.0, "step": 11077 }, { "epoch": 1.409235466225671, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7577946186065674, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8626956343650818, "num_tokens": 420721566.0, "step": 11078 }, { "epoch": 1.4093626765042615, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5514004230499268, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8730093240737915, "num_tokens": 420758591.0, "step": 11079 }, { "epoch": 1.409489886782852, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7524189949035645, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8624643087387085, "num_tokens": 420796864.0, "step": 11080 }, { "epoch": 1.4096170970614426, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5409960746765137, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8811071515083313, "num_tokens": 420834689.0, "step": 11081 }, { "epoch": 1.409744307340033, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7388546466827393, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8714205026626587, "num_tokens": 420877795.0, "step": 11082 }, { "epoch": 1.4098715176186236, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5435914993286133, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.874604344367981, "num_tokens": 420917217.0, "step": 11083 }, { "epoch": 1.4099987278972141, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.75154447555542, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.869456946849823, "num_tokens": 420952607.0, "step": 11084 }, { "epoch": 1.4101259381758047, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5548288822174072, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8632394671440125, "num_tokens": 420986570.0, "step": 11085 }, { "epoch": 1.410253148454395, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.741852283477783, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8717837333679199, "num_tokens": 421024482.0, "step": 11086 }, { "epoch": 1.4103803587329855, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5376288890838623, "learning_rate": 1e-05, "loss": 0.3644, "mean_token_accuracy": 0.8934671878814697, "num_tokens": 421058185.0, "step": 11087 }, { "epoch": 1.410507569011576, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7587389945983887, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8714250326156616, "num_tokens": 421089796.0, "step": 11088 }, { "epoch": 1.4106347792901666, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5407960414886475, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8663238286972046, "num_tokens": 421132644.0, "step": 11089 }, { "epoch": 1.410761989568757, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.755291700363159, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8686075210571289, "num_tokens": 421167790.0, "step": 11090 }, { "epoch": 1.4108891998473476, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5391111373901367, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8649032115936279, "num_tokens": 421206765.0, "step": 11091 }, { "epoch": 1.4110164101259381, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7405076026916504, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8796406984329224, "num_tokens": 421249997.0, "step": 11092 }, { "epoch": 1.4111436204045287, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.537297487258911, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8740487098693848, "num_tokens": 421287674.0, "step": 11093 }, { "epoch": 1.4112708306831192, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.741032123565674, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8767591714859009, "num_tokens": 421325494.0, "step": 11094 }, { "epoch": 1.4113980409617097, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5396299362182617, "learning_rate": 1e-05, "loss": 0.4004, "mean_token_accuracy": 0.8811209201812744, "num_tokens": 421359934.0, "step": 11095 }, { "epoch": 1.4115252512403003, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002460479736328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7528438568115234, "learning_rate": 1e-05, "loss": 0.4978, "mean_token_accuracy": 0.857538104057312, "num_tokens": 421399195.0, "step": 11096 }, { "epoch": 1.4116524615188908, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.549541711807251, "learning_rate": 1e-05, "loss": 0.3747, "mean_token_accuracy": 0.889892041683197, "num_tokens": 421428822.0, "step": 11097 }, { "epoch": 1.4117796717974813, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7408900260925293, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8676575422286987, "num_tokens": 421468409.0, "step": 11098 }, { "epoch": 1.4119068820760718, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5662035942077637, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8507897853851318, "num_tokens": 421506905.0, "step": 11099 }, { "epoch": 1.4120340923546624, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7275776863098145, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.857338547706604, "num_tokens": 421545347.0, "step": 11100 }, { "epoch": 1.412161302633253, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.592331886291504, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8722212314605713, "num_tokens": 421580819.0, "step": 11101 }, { "epoch": 1.4122885129118432, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.679316282272339, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8728131651878357, "num_tokens": 421628474.0, "step": 11102 }, { "epoch": 1.4124157231904337, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5666663646698, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8638677000999451, "num_tokens": 421670437.0, "step": 11103 }, { "epoch": 1.4125429334690243, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.684292793273926, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8534250855445862, "num_tokens": 421709157.0, "step": 11104 }, { "epoch": 1.4126701437476148, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.564462184906006, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8779268860816956, "num_tokens": 421743783.0, "step": 11105 }, { "epoch": 1.4127973540262053, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.667236566543579, "learning_rate": 1e-05, "loss": 0.3909, "mean_token_accuracy": 0.8858668208122253, "num_tokens": 421781376.0, "step": 11106 }, { "epoch": 1.4129245643047958, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.559420108795166, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8826684951782227, "num_tokens": 421819367.0, "step": 11107 }, { "epoch": 1.4130517745833864, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6581592559814453, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.882786750793457, "num_tokens": 421862830.0, "step": 11108 }, { "epoch": 1.413178984861977, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5590646266937256, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8806226253509521, "num_tokens": 421898353.0, "step": 11109 }, { "epoch": 1.4133061951405674, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6701571941375732, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8702547550201416, "num_tokens": 421932911.0, "step": 11110 }, { "epoch": 1.4134334054191577, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.555176019668579, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8749361038208008, "num_tokens": 421974409.0, "step": 11111 }, { "epoch": 1.4135606156977483, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.676102638244629, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8528566360473633, "num_tokens": 422011913.0, "step": 11112 }, { "epoch": 1.4136878259763388, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.547076463699341, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8719730377197266, "num_tokens": 422054487.0, "step": 11113 }, { "epoch": 1.4138150362549293, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6800739765167236, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.871385931968689, "num_tokens": 422092980.0, "step": 11114 }, { "epoch": 1.4139422465335199, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5268325805664062, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8622958064079285, "num_tokens": 422139368.0, "step": 11115 }, { "epoch": 1.4140694568121104, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6883976459503174, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8832458853721619, "num_tokens": 422177342.0, "step": 11116 }, { "epoch": 1.414196667090701, "ewc_loss": 0.0625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.526529312133789, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8769736289978027, "num_tokens": 422221378.0, "step": 11117 }, { "epoch": 1.4143238773692914, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7033658027648926, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8700256943702698, "num_tokens": 422260503.0, "step": 11118 }, { "epoch": 1.414451087647882, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.520991802215576, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8645399212837219, "num_tokens": 422301218.0, "step": 11119 }, { "epoch": 1.4145782979264725, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.761767625808716, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8800792694091797, "num_tokens": 422337366.0, "step": 11120 }, { "epoch": 1.414705508205063, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5100460052490234, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8841315507888794, "num_tokens": 422376087.0, "step": 11121 }, { "epoch": 1.4148327184836536, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.758910894393921, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8722692131996155, "num_tokens": 422413871.0, "step": 11122 }, { "epoch": 1.414959928762244, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.520158529281616, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8725881576538086, "num_tokens": 422450834.0, "step": 11123 }, { "epoch": 1.4150871390408346, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7608838081359863, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8803274631500244, "num_tokens": 422494566.0, "step": 11124 }, { "epoch": 1.4152143493194251, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5446090698242188, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8646251559257507, "num_tokens": 422530107.0, "step": 11125 }, { "epoch": 1.4153415595980154, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7739813327789307, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8514052629470825, "num_tokens": 422567598.0, "step": 11126 }, { "epoch": 1.415468769876606, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.517503499984741, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8721709847450256, "num_tokens": 422611705.0, "step": 11127 }, { "epoch": 1.4155959801551965, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.784111261367798, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8657414317131042, "num_tokens": 422648082.0, "step": 11128 }, { "epoch": 1.415723190433787, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5178635120391846, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.873142659664154, "num_tokens": 422691974.0, "step": 11129 }, { "epoch": 1.4158504007123776, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8005692958831787, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8633451461791992, "num_tokens": 422728245.0, "step": 11130 }, { "epoch": 1.415977610990968, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.545483350753784, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8736675977706909, "num_tokens": 422760920.0, "step": 11131 }, { "epoch": 1.4161048212695586, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7912933826446533, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8693079948425293, "num_tokens": 422796188.0, "step": 11132 }, { "epoch": 1.4162320315481491, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.528445243835449, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8826704025268555, "num_tokens": 422836232.0, "step": 11133 }, { "epoch": 1.4163592418267397, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7973809242248535, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8778244256973267, "num_tokens": 422872855.0, "step": 11134 }, { "epoch": 1.41648645210533, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5396251678466797, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8688647150993347, "num_tokens": 422912826.0, "step": 11135 }, { "epoch": 1.4166136623839205, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.808485507965088, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.862026572227478, "num_tokens": 422950034.0, "step": 11136 }, { "epoch": 1.416740872662511, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5463171005249023, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8754293918609619, "num_tokens": 422987764.0, "step": 11137 }, { "epoch": 1.4168680829411016, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.785024404525757, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.8887124061584473, "num_tokens": 423024274.0, "step": 11138 }, { "epoch": 1.416995293219692, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5727956295013428, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8605223894119263, "num_tokens": 423058630.0, "step": 11139 }, { "epoch": 1.4171225034982826, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.758045196533203, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8812392950057983, "num_tokens": 423099829.0, "step": 11140 }, { "epoch": 1.4172497137768731, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.584650754928589, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8721817135810852, "num_tokens": 423130369.0, "step": 11141 }, { "epoch": 1.4173769240554637, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7368266582489014, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8738061785697937, "num_tokens": 423166437.0, "step": 11142 }, { "epoch": 1.4175041343340542, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5580320358276367, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8862712383270264, "num_tokens": 423202220.0, "step": 11143 }, { "epoch": 1.4176313446126447, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7300057411193848, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8687263131141663, "num_tokens": 423243417.0, "step": 11144 }, { "epoch": 1.4177585548912353, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.558328151702881, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8682860136032104, "num_tokens": 423282337.0, "step": 11145 }, { "epoch": 1.4178857651698258, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.738264799118042, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8655303120613098, "num_tokens": 423318666.0, "step": 11146 }, { "epoch": 1.4180129754484163, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.55112361907959, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8651009202003479, "num_tokens": 423357895.0, "step": 11147 }, { "epoch": 1.4181401857270068, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.725734233856201, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8710213899612427, "num_tokens": 423395330.0, "step": 11148 }, { "epoch": 1.4182673960055974, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.541724920272827, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8712462186813354, "num_tokens": 423436824.0, "step": 11149 }, { "epoch": 1.418394606284188, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7455508708953857, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8693453669548035, "num_tokens": 423471906.0, "step": 11150 }, { "epoch": 1.4185218165627782, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.53525447845459, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8734664916992188, "num_tokens": 423510774.0, "step": 11151 }, { "epoch": 1.4186490268413687, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.743734836578369, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8716846704483032, "num_tokens": 423549825.0, "step": 11152 }, { "epoch": 1.4187762371199593, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5717787742614746, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8763840198516846, "num_tokens": 423577546.0, "step": 11153 }, { "epoch": 1.4189034473985498, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7212650775909424, "learning_rate": 1e-05, "loss": 0.4061, "mean_token_accuracy": 0.8816831111907959, "num_tokens": 423613167.0, "step": 11154 }, { "epoch": 1.4190306576771403, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5559322834014893, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8597724437713623, "num_tokens": 423649601.0, "step": 11155 }, { "epoch": 1.4191578679557308, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.722337245941162, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8625333905220032, "num_tokens": 423685429.0, "step": 11156 }, { "epoch": 1.4192850782343214, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.543475389480591, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8795225620269775, "num_tokens": 423724270.0, "step": 11157 }, { "epoch": 1.419412288512912, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7161459922790527, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8818165063858032, "num_tokens": 423760935.0, "step": 11158 }, { "epoch": 1.4195394987915024, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5459179878234863, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8696295619010925, "num_tokens": 423801949.0, "step": 11159 }, { "epoch": 1.4196667090700927, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7098288536071777, "learning_rate": 1e-05, "loss": 0.4008, "mean_token_accuracy": 0.8820667862892151, "num_tokens": 423840532.0, "step": 11160 }, { "epoch": 1.4197939193486833, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5508034229278564, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8624318838119507, "num_tokens": 423877912.0, "step": 11161 }, { "epoch": 1.4199211296272738, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7003557682037354, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8786641955375671, "num_tokens": 423917836.0, "step": 11162 }, { "epoch": 1.4200483399058643, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002498626708984375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5480353832244873, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8644765615463257, "num_tokens": 423957307.0, "step": 11163 }, { "epoch": 1.4201755501844548, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7157680988311768, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8705477118492126, "num_tokens": 423994280.0, "step": 11164 }, { "epoch": 1.4203027604630454, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.542081594467163, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8708452582359314, "num_tokens": 424033682.0, "step": 11165 }, { "epoch": 1.420429970741636, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7094242572784424, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.866675615310669, "num_tokens": 424072432.0, "step": 11166 }, { "epoch": 1.4205571810202264, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5467824935913086, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8621363639831543, "num_tokens": 424110074.0, "step": 11167 }, { "epoch": 1.420684391298817, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7120447158813477, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8663294911384583, "num_tokens": 424148263.0, "step": 11168 }, { "epoch": 1.4208116015774075, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.532729387283325, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8612672686576843, "num_tokens": 424190342.0, "step": 11169 }, { "epoch": 1.420938811855998, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7153687477111816, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8629344701766968, "num_tokens": 424236136.0, "step": 11170 }, { "epoch": 1.4210660221345885, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5408053398132324, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8590680956840515, "num_tokens": 424274436.0, "step": 11171 }, { "epoch": 1.421193232413179, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.727611780166626, "learning_rate": 1e-05, "loss": 0.3954, "mean_token_accuracy": 0.8887698650360107, "num_tokens": 424309661.0, "step": 11172 }, { "epoch": 1.4213204426917696, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.541987419128418, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8615843057632446, "num_tokens": 424347127.0, "step": 11173 }, { "epoch": 1.4214476529703601, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.705467462539673, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8787535429000854, "num_tokens": 424391287.0, "step": 11174 }, { "epoch": 1.4215748632489504, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5328290462493896, "learning_rate": 1e-05, "loss": 0.3887, "mean_token_accuracy": 0.8805967569351196, "num_tokens": 424428091.0, "step": 11175 }, { "epoch": 1.421702073527541, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.71242094039917, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8682243824005127, "num_tokens": 424469270.0, "step": 11176 }, { "epoch": 1.4218292838061315, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.537229299545288, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8584539890289307, "num_tokens": 424509116.0, "step": 11177 }, { "epoch": 1.421956494084722, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.72688889503479, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.874147891998291, "num_tokens": 424542777.0, "step": 11178 }, { "epoch": 1.4220837043633126, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5225889682769775, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8733689188957214, "num_tokens": 424582299.0, "step": 11179 }, { "epoch": 1.422210914641903, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7344439029693604, "learning_rate": 1e-05, "loss": 0.3936, "mean_token_accuracy": 0.8859379291534424, "num_tokens": 424617543.0, "step": 11180 }, { "epoch": 1.4223381249204936, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.518928289413452, "learning_rate": 1e-05, "loss": 0.3988, "mean_token_accuracy": 0.8820346593856812, "num_tokens": 424657512.0, "step": 11181 }, { "epoch": 1.4224653351990841, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7517948150634766, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8624107837677002, "num_tokens": 424695954.0, "step": 11182 }, { "epoch": 1.4225925454776747, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.523928165435791, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8594083786010742, "num_tokens": 424736174.0, "step": 11183 }, { "epoch": 1.422719755756265, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.756817102432251, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8711809515953064, "num_tokens": 424776371.0, "step": 11184 }, { "epoch": 1.4228469660348555, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5249223709106445, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8617371320724487, "num_tokens": 424816874.0, "step": 11185 }, { "epoch": 1.422974176313446, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.767423152923584, "learning_rate": 1e-05, "loss": 0.3852, "mean_token_accuracy": 0.8868391513824463, "num_tokens": 424847924.0, "step": 11186 }, { "epoch": 1.4231013865920366, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5186564922332764, "learning_rate": 1e-05, "loss": 0.3877, "mean_token_accuracy": 0.8864376544952393, "num_tokens": 424887950.0, "step": 11187 }, { "epoch": 1.423228596870627, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.786604881286621, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8634872436523438, "num_tokens": 424922608.0, "step": 11188 }, { "epoch": 1.4233558071492176, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.536180019378662, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8789260387420654, "num_tokens": 424958564.0, "step": 11189 }, { "epoch": 1.4234830174278081, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7823522090911865, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8683174848556519, "num_tokens": 424993351.0, "step": 11190 }, { "epoch": 1.4236102277063987, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.541234254837036, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8711832761764526, "num_tokens": 425028018.0, "step": 11191 }, { "epoch": 1.4237374379849892, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.77626633644104, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8736422061920166, "num_tokens": 425065023.0, "step": 11192 }, { "epoch": 1.4238646482635797, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5437567234039307, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8765819668769836, "num_tokens": 425101152.0, "step": 11193 }, { "epoch": 1.4239918585421703, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7697694301605225, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.864761471748352, "num_tokens": 425137756.0, "step": 11194 }, { "epoch": 1.4241190688207608, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.553992509841919, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8757473230361938, "num_tokens": 425171947.0, "step": 11195 }, { "epoch": 1.4242462790993513, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.761519193649292, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8712899684906006, "num_tokens": 425207219.0, "step": 11196 }, { "epoch": 1.4243734893779418, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5611889362335205, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8633184432983398, "num_tokens": 425242515.0, "step": 11197 }, { "epoch": 1.4245006996565324, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7414143085479736, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8586891889572144, "num_tokens": 425283409.0, "step": 11198 }, { "epoch": 1.424627909935123, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5570614337921143, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8651895523071289, "num_tokens": 425321862.0, "step": 11199 }, { "epoch": 1.4247551202137132, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7294204235076904, "learning_rate": 1e-05, "loss": 0.4121, "mean_token_accuracy": 0.8803588151931763, "num_tokens": 425362034.0, "step": 11200 }, { "epoch": 1.4248823304923037, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.567934274673462, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8547677397727966, "num_tokens": 425397900.0, "step": 11201 }, { "epoch": 1.4250095407708943, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.714885950088501, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8721989393234253, "num_tokens": 425435458.0, "step": 11202 }, { "epoch": 1.4251367510494848, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5683321952819824, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.880782961845398, "num_tokens": 425465687.0, "step": 11203 }, { "epoch": 1.4252639613280753, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6969196796417236, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8687636852264404, "num_tokens": 425500702.0, "step": 11204 }, { "epoch": 1.4253911716066658, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.554896831512451, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8725504875183105, "num_tokens": 425542827.0, "step": 11205 }, { "epoch": 1.4255183818852564, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6972522735595703, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8618265390396118, "num_tokens": 425585146.0, "step": 11206 }, { "epoch": 1.425645592163847, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.554290771484375, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8788087368011475, "num_tokens": 425623790.0, "step": 11207 }, { "epoch": 1.4257728024424374, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6973559856414795, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8709932565689087, "num_tokens": 425663290.0, "step": 11208 }, { "epoch": 1.4259000127210277, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5436105728149414, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8757354021072388, "num_tokens": 425701226.0, "step": 11209 }, { "epoch": 1.4260272229996183, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.707772731781006, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8717631101608276, "num_tokens": 425737524.0, "step": 11210 }, { "epoch": 1.4261544332782088, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5603857040405273, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8782861828804016, "num_tokens": 425771010.0, "step": 11211 }, { "epoch": 1.4262816435567993, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6932930946350098, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8646464943885803, "num_tokens": 425813077.0, "step": 11212 }, { "epoch": 1.4264088538353898, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5344390869140625, "learning_rate": 1e-05, "loss": 0.399, "mean_token_accuracy": 0.885351300239563, "num_tokens": 425852718.0, "step": 11213 }, { "epoch": 1.4265360641139804, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.724928140640259, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8709661960601807, "num_tokens": 425888196.0, "step": 11214 }, { "epoch": 1.426663274392571, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.519299268722534, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8800400495529175, "num_tokens": 425926774.0, "step": 11215 }, { "epoch": 1.4267904846711614, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7254154682159424, "learning_rate": 1e-05, "loss": 0.3905, "mean_token_accuracy": 0.8868019580841064, "num_tokens": 425966207.0, "step": 11216 }, { "epoch": 1.426917694949752, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5224907398223877, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.86214280128479, "num_tokens": 426008792.0, "step": 11217 }, { "epoch": 1.4270449052283425, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7376136779785156, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8710811138153076, "num_tokens": 426049650.0, "step": 11218 }, { "epoch": 1.427172115506933, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.516038656234741, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8813507556915283, "num_tokens": 426092143.0, "step": 11219 }, { "epoch": 1.4272993257855235, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7444164752960205, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8593415021896362, "num_tokens": 426135393.0, "step": 11220 }, { "epoch": 1.427426536064114, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.53159761428833, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8721128702163696, "num_tokens": 426171858.0, "step": 11221 }, { "epoch": 1.4275537463427046, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.758234977722168, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8815393447875977, "num_tokens": 426206635.0, "step": 11222 }, { "epoch": 1.4276809566212951, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.533029794692993, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8747473955154419, "num_tokens": 426240942.0, "step": 11223 }, { "epoch": 1.4278081668998854, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7531299591064453, "learning_rate": 1e-05, "loss": 0.3746, "mean_token_accuracy": 0.8919834494590759, "num_tokens": 426279592.0, "step": 11224 }, { "epoch": 1.427935377178476, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5314457416534424, "learning_rate": 1e-05, "loss": 0.3945, "mean_token_accuracy": 0.8839045763015747, "num_tokens": 426314218.0, "step": 11225 }, { "epoch": 1.4280625874570665, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.757892370223999, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8717825412750244, "num_tokens": 426351514.0, "step": 11226 }, { "epoch": 1.428189797735657, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.539675712585449, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8603534698486328, "num_tokens": 426389553.0, "step": 11227 }, { "epoch": 1.4283170080142475, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.752887487411499, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8805391192436218, "num_tokens": 426424026.0, "step": 11228 }, { "epoch": 1.428444218292838, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5358378887176514, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8722970485687256, "num_tokens": 426466711.0, "step": 11229 }, { "epoch": 1.4285714285714286, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.753159999847412, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8677050471305847, "num_tokens": 426507636.0, "step": 11230 }, { "epoch": 1.4286986388500191, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.53310489654541, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8679995536804199, "num_tokens": 426551811.0, "step": 11231 }, { "epoch": 1.4288258491286097, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7672290802001953, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8704848885536194, "num_tokens": 426585814.0, "step": 11232 }, { "epoch": 1.4289530594072, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.535586357116699, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8788707256317139, "num_tokens": 426619327.0, "step": 11233 }, { "epoch": 1.4290802696857905, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7732479572296143, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8635272979736328, "num_tokens": 426656976.0, "step": 11234 }, { "epoch": 1.429207479964381, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.54024076461792, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8796067237854004, "num_tokens": 426693930.0, "step": 11235 }, { "epoch": 1.4293346902429716, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7598001956939697, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.8890878558158875, "num_tokens": 426731655.0, "step": 11236 }, { "epoch": 1.429461900521562, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5515525341033936, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8678125739097595, "num_tokens": 426768384.0, "step": 11237 }, { "epoch": 1.4295891108001526, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7531044483184814, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8692557215690613, "num_tokens": 426806319.0, "step": 11238 }, { "epoch": 1.4297163210787431, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5649309158325195, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8483892679214478, "num_tokens": 426841076.0, "step": 11239 }, { "epoch": 1.4298435313573337, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.737560987472534, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8786120414733887, "num_tokens": 426877647.0, "step": 11240 }, { "epoch": 1.4299707416359242, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.54353404045105, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8731245398521423, "num_tokens": 426917975.0, "step": 11241 }, { "epoch": 1.4300979519145147, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7321829795837402, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8715060949325562, "num_tokens": 426958975.0, "step": 11242 }, { "epoch": 1.4302251621931052, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.547424554824829, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8713558316230774, "num_tokens": 426997205.0, "step": 11243 }, { "epoch": 1.4303523724716958, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.723768472671509, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8715640306472778, "num_tokens": 427037455.0, "step": 11244 }, { "epoch": 1.4304795827502863, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5576486587524414, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8682962656021118, "num_tokens": 427071443.0, "step": 11245 }, { "epoch": 1.4306067930288768, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7085015773773193, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8741910457611084, "num_tokens": 427113258.0, "step": 11246 }, { "epoch": 1.4307340033074674, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5414161682128906, "learning_rate": 1e-05, "loss": 0.4121, "mean_token_accuracy": 0.8760156631469727, "num_tokens": 427153523.0, "step": 11247 }, { "epoch": 1.430861213586058, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7194766998291016, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8730446100234985, "num_tokens": 427186818.0, "step": 11248 }, { "epoch": 1.4309884238646482, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.547650098800659, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8575893640518188, "num_tokens": 427226817.0, "step": 11249 }, { "epoch": 1.4311156341432387, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.709728479385376, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8693302869796753, "num_tokens": 427268471.0, "step": 11250 }, { "epoch": 1.4312428444218293, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5453531742095947, "learning_rate": 1e-05, "loss": 0.3922, "mean_token_accuracy": 0.8849495053291321, "num_tokens": 427301242.0, "step": 11251 }, { "epoch": 1.4313700547004198, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7386839389801025, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8629350662231445, "num_tokens": 427335230.0, "step": 11252 }, { "epoch": 1.4314972649790103, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.54720401763916, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8626286387443542, "num_tokens": 427371993.0, "step": 11253 }, { "epoch": 1.4316244752576008, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7295069694519043, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.863476037979126, "num_tokens": 427408810.0, "step": 11254 }, { "epoch": 1.4317516855361914, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.559821605682373, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8572702407836914, "num_tokens": 427448740.0, "step": 11255 }, { "epoch": 1.431878895814782, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7027599811553955, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8736473321914673, "num_tokens": 427488234.0, "step": 11256 }, { "epoch": 1.4320061060933724, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.54472279548645, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8702126145362854, "num_tokens": 427528053.0, "step": 11257 }, { "epoch": 1.4321333163719627, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00025177001953125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.697028398513794, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8784160614013672, "num_tokens": 427566882.0, "step": 11258 }, { "epoch": 1.4322605266505533, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.557234287261963, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8748565912246704, "num_tokens": 427600464.0, "step": 11259 }, { "epoch": 1.4323877369291438, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.701591968536377, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8774665594100952, "num_tokens": 427633394.0, "step": 11260 }, { "epoch": 1.4325149472077343, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.547398567199707, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8713657855987549, "num_tokens": 427670591.0, "step": 11261 }, { "epoch": 1.4326421574863248, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.686753511428833, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.883337140083313, "num_tokens": 427711752.0, "step": 11262 }, { "epoch": 1.4327693677649154, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.565145492553711, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8731166124343872, "num_tokens": 427744655.0, "step": 11263 }, { "epoch": 1.432896578043506, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6860568523406982, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8778388500213623, "num_tokens": 427784388.0, "step": 11264 }, { "epoch": 1.4330237883220964, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5768916606903076, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8617146015167236, "num_tokens": 427823498.0, "step": 11265 }, { "epoch": 1.433150998600687, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.656303644180298, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8724182844161987, "num_tokens": 427868185.0, "step": 11266 }, { "epoch": 1.4332782088792775, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.578519105911255, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8640285134315491, "num_tokens": 427903315.0, "step": 11267 }, { "epoch": 1.433405419157868, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.643369197845459, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8591601848602295, "num_tokens": 427950206.0, "step": 11268 }, { "epoch": 1.4335326294364585, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.613792657852173, "learning_rate": 1e-05, "loss": 0.5345, "mean_token_accuracy": 0.8510143756866455, "num_tokens": 427983038.0, "step": 11269 }, { "epoch": 1.433659839715049, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.635157346725464, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8791717886924744, "num_tokens": 428015148.0, "step": 11270 }, { "epoch": 1.4337870499936396, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5863354206085205, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8720694780349731, "num_tokens": 428054300.0, "step": 11271 }, { "epoch": 1.4339142602722301, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.624619722366333, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8594672679901123, "num_tokens": 428093802.0, "step": 11272 }, { "epoch": 1.4340414705508204, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5789976119995117, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8814202547073364, "num_tokens": 428133283.0, "step": 11273 }, { "epoch": 1.434168680829411, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.635676860809326, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8641208410263062, "num_tokens": 428171125.0, "step": 11274 }, { "epoch": 1.4342958911080015, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5691792964935303, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8543590307235718, "num_tokens": 428211731.0, "step": 11275 }, { "epoch": 1.434423101386592, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.65484881401062, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.867214560508728, "num_tokens": 428243737.0, "step": 11276 }, { "epoch": 1.4345503116651825, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5653510093688965, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8799682855606079, "num_tokens": 428278761.0, "step": 11277 }, { "epoch": 1.434677521943773, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.647426128387451, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.8892761468887329, "num_tokens": 428316158.0, "step": 11278 }, { "epoch": 1.4348047322223636, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.581334114074707, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8675874471664429, "num_tokens": 428347180.0, "step": 11279 }, { "epoch": 1.4349319425009541, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.639914035797119, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8789228200912476, "num_tokens": 428389962.0, "step": 11280 }, { "epoch": 1.4350591527795447, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.564732789993286, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8498916625976562, "num_tokens": 428432950.0, "step": 11281 }, { "epoch": 1.435186363058135, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.652519464492798, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8554198741912842, "num_tokens": 428470745.0, "step": 11282 }, { "epoch": 1.4353135733367255, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.7003495693206787, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.866658627986908, "num_tokens": 428508836.0, "step": 11283 }, { "epoch": 1.435440783615316, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.587707281112671, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8730971813201904, "num_tokens": 428546371.0, "step": 11284 }, { "epoch": 1.4355679938939065, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.6328494548797607, "learning_rate": 1e-05, "loss": 0.3908, "mean_token_accuracy": 0.8841685652732849, "num_tokens": 428578878.0, "step": 11285 }, { "epoch": 1.435695204172497, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5811843872070312, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8650633096694946, "num_tokens": 428611218.0, "step": 11286 }, { "epoch": 1.4358224144510876, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.644926071166992, "learning_rate": 1e-05, "loss": 0.5271, "mean_token_accuracy": 0.8440399169921875, "num_tokens": 428649381.0, "step": 11287 }, { "epoch": 1.4359496247296781, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.574953556060791, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8694337606430054, "num_tokens": 428685221.0, "step": 11288 }, { "epoch": 1.4360768350082687, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.635049343109131, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8724305629730225, "num_tokens": 428724194.0, "step": 11289 }, { "epoch": 1.4362040452868592, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5877156257629395, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.868853747844696, "num_tokens": 428760501.0, "step": 11290 }, { "epoch": 1.4363312555654497, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.643580913543701, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8597442507743835, "num_tokens": 428793917.0, "step": 11291 }, { "epoch": 1.4364584658440402, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.569283962249756, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8789433240890503, "num_tokens": 428828998.0, "step": 11292 }, { "epoch": 1.4365856761226308, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.635720729827881, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8864459991455078, "num_tokens": 428864739.0, "step": 11293 }, { "epoch": 1.4367128864012213, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.568162202835083, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8593307733535767, "num_tokens": 428901026.0, "step": 11294 }, { "epoch": 1.4368400966798118, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.638395309448242, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8790119886398315, "num_tokens": 428937698.0, "step": 11295 }, { "epoch": 1.4369673069584024, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.567491054534912, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8538883924484253, "num_tokens": 428978288.0, "step": 11296 }, { "epoch": 1.4370945172369929, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6451454162597656, "learning_rate": 1e-05, "loss": 0.3795, "mean_token_accuracy": 0.8872057199478149, "num_tokens": 429014299.0, "step": 11297 }, { "epoch": 1.4372217275155832, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5788309574127197, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.863761305809021, "num_tokens": 429052455.0, "step": 11298 }, { "epoch": 1.4373489377941737, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6284878253936768, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8704244494438171, "num_tokens": 429094168.0, "step": 11299 }, { "epoch": 1.4374761480727642, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5601754188537598, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8694218397140503, "num_tokens": 429137147.0, "step": 11300 }, { "epoch": 1.4376033583513548, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6288864612579346, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.877224326133728, "num_tokens": 429181679.0, "step": 11301 }, { "epoch": 1.4377305686299453, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5659799575805664, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8714070320129395, "num_tokens": 429222378.0, "step": 11302 }, { "epoch": 1.4378577789085358, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6424992084503174, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8685256838798523, "num_tokens": 429262044.0, "step": 11303 }, { "epoch": 1.4379849891871264, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5580291748046875, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8710639476776123, "num_tokens": 429301643.0, "step": 11304 }, { "epoch": 1.438112199465717, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6647987365722656, "learning_rate": 1e-05, "loss": 0.5171, "mean_token_accuracy": 0.8473372459411621, "num_tokens": 429339190.0, "step": 11305 }, { "epoch": 1.4382394097443074, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5504939556121826, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8679381012916565, "num_tokens": 429378569.0, "step": 11306 }, { "epoch": 1.4383666200228977, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.661648750305176, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8697512149810791, "num_tokens": 429416012.0, "step": 11307 }, { "epoch": 1.4384938303014883, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.529528856277466, "learning_rate": 1e-05, "loss": 0.3689, "mean_token_accuracy": 0.8924987316131592, "num_tokens": 429453242.0, "step": 11308 }, { "epoch": 1.4386210405800788, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.677804708480835, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8763353824615479, "num_tokens": 429493908.0, "step": 11309 }, { "epoch": 1.4387482508586693, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.523813009262085, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.8744051456451416, "num_tokens": 429530335.0, "step": 11310 }, { "epoch": 1.4388754611372598, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7027318477630615, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8666179776191711, "num_tokens": 429567523.0, "step": 11311 }, { "epoch": 1.4390026714158504, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5167837142944336, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8794645667076111, "num_tokens": 429605887.0, "step": 11312 }, { "epoch": 1.439129881694441, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7079951763153076, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.8825663328170776, "num_tokens": 429644834.0, "step": 11313 }, { "epoch": 1.4392570919730314, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.5209648609161377, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.8795318603515625, "num_tokens": 429681151.0, "step": 11314 }, { "epoch": 1.439384302251622, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7314367294311523, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8608511686325073, "num_tokens": 429719135.0, "step": 11315 }, { "epoch": 1.4395115125302125, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.512904167175293, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8799941539764404, "num_tokens": 429753443.0, "step": 11316 }, { "epoch": 1.439638722808803, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.753180742263794, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8749837875366211, "num_tokens": 429788762.0, "step": 11317 }, { "epoch": 1.4397659330873935, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.7670135498046875e-05, "grad_norm": 2.512003183364868, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8604543209075928, "num_tokens": 429824054.0, "step": 11318 }, { "epoch": 1.439893143365984, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7859556674957275, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8691419363021851, "num_tokens": 429857772.0, "step": 11319 }, { "epoch": 1.4400203536445746, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.521256685256958, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8771969676017761, "num_tokens": 429891184.0, "step": 11320 }, { "epoch": 1.4401475639231651, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.77820086479187, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8785892724990845, "num_tokens": 429930322.0, "step": 11321 }, { "epoch": 1.4402747742017554, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.523087978363037, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.875454843044281, "num_tokens": 429965708.0, "step": 11322 }, { "epoch": 1.440401984480346, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.777569055557251, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8710713982582092, "num_tokens": 430007266.0, "step": 11323 }, { "epoch": 1.4405291947589365, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.534517526626587, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8638314008712769, "num_tokens": 430044470.0, "step": 11324 }, { "epoch": 1.440656405037527, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.785670518875122, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8728172183036804, "num_tokens": 430079776.0, "step": 11325 }, { "epoch": 1.4407836153161175, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.542503595352173, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.880477786064148, "num_tokens": 430110697.0, "step": 11326 }, { "epoch": 1.440910825594708, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7815542221069336, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8711423277854919, "num_tokens": 430150350.0, "step": 11327 }, { "epoch": 1.4410380358732986, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.563206672668457, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8536620140075684, "num_tokens": 430185415.0, "step": 11328 }, { "epoch": 1.4411652461518891, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7545483112335205, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8663240671157837, "num_tokens": 430227865.0, "step": 11329 }, { "epoch": 1.4412924564304797, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5452680587768555, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8799736499786377, "num_tokens": 430267821.0, "step": 11330 }, { "epoch": 1.44141966670907, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.744403839111328, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8751046657562256, "num_tokens": 430307251.0, "step": 11331 }, { "epoch": 1.4415468769876605, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.548074245452881, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8582648038864136, "num_tokens": 430347842.0, "step": 11332 }, { "epoch": 1.441674087266251, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7529919147491455, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8644011616706848, "num_tokens": 430383175.0, "step": 11333 }, { "epoch": 1.4418012975448415, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5637683868408203, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8669422268867493, "num_tokens": 430414319.0, "step": 11334 }, { "epoch": 1.441928507823432, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7257657051086426, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8831927180290222, "num_tokens": 430454142.0, "step": 11335 }, { "epoch": 1.4420557181020226, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5490622520446777, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8602343797683716, "num_tokens": 430496980.0, "step": 11336 }, { "epoch": 1.4421829283806131, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.734877109527588, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8620021343231201, "num_tokens": 430533645.0, "step": 11337 }, { "epoch": 1.4423101386592037, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5560660362243652, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8732323050498962, "num_tokens": 430567744.0, "step": 11338 }, { "epoch": 1.4424373489377942, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.740792989730835, "learning_rate": 1e-05, "loss": 0.5234, "mean_token_accuracy": 0.8456417322158813, "num_tokens": 430603887.0, "step": 11339 }, { "epoch": 1.4425645592163847, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.549989938735962, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8714522123336792, "num_tokens": 430638748.0, "step": 11340 }, { "epoch": 1.4426917694949752, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.733011245727539, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8711341619491577, "num_tokens": 430677799.0, "step": 11341 }, { "epoch": 1.4428189797735658, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.535731077194214, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.865955650806427, "num_tokens": 430720816.0, "step": 11342 }, { "epoch": 1.4429461900521563, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.739885091781616, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8710315823554993, "num_tokens": 430755220.0, "step": 11343 }, { "epoch": 1.4430734003307468, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5568177700042725, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8718195557594299, "num_tokens": 430787737.0, "step": 11344 }, { "epoch": 1.4432006106093374, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7276811599731445, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8709542751312256, "num_tokens": 430823817.0, "step": 11345 }, { "epoch": 1.4433278208879279, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5443332195281982, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8658077716827393, "num_tokens": 430863416.0, "step": 11346 }, { "epoch": 1.4434550311665182, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.743168592453003, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.862921416759491, "num_tokens": 430894017.0, "step": 11347 }, { "epoch": 1.4435822414451087, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002536773681640625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5337560176849365, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8666502237319946, "num_tokens": 430935023.0, "step": 11348 }, { "epoch": 1.4437094517236992, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7497408390045166, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8714035153388977, "num_tokens": 430971663.0, "step": 11349 }, { "epoch": 1.4438366620022898, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5269854068756104, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8652937412261963, "num_tokens": 431018919.0, "step": 11350 }, { "epoch": 1.4439638722808803, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.766225576400757, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8802459239959717, "num_tokens": 431050076.0, "step": 11351 }, { "epoch": 1.4440910825594708, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.542041063308716, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8578606843948364, "num_tokens": 431085535.0, "step": 11352 }, { "epoch": 1.4442182928380614, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.756139039993286, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8761245012283325, "num_tokens": 431123187.0, "step": 11353 }, { "epoch": 1.4443455031166519, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.555797576904297, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8604930639266968, "num_tokens": 431157867.0, "step": 11354 }, { "epoch": 1.4444727133952424, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.772934675216675, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8657308220863342, "num_tokens": 431191344.0, "step": 11355 }, { "epoch": 1.4445999236738327, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.538161039352417, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8732609152793884, "num_tokens": 431231019.0, "step": 11356 }, { "epoch": 1.4447271339524232, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.764409065246582, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8701669573783875, "num_tokens": 431265660.0, "step": 11357 }, { "epoch": 1.4448543442310138, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.58640193939209, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8800516128540039, "num_tokens": 431298666.0, "step": 11358 }, { "epoch": 1.4449815545096043, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7506320476531982, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8846502304077148, "num_tokens": 431327516.0, "step": 11359 }, { "epoch": 1.4451087647881948, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5560379028320312, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8701882362365723, "num_tokens": 431366244.0, "step": 11360 }, { "epoch": 1.4452359750667854, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.727322578430176, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8633294701576233, "num_tokens": 431402440.0, "step": 11361 }, { "epoch": 1.445363185345376, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.55454683303833, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8868497014045715, "num_tokens": 431436628.0, "step": 11362 }, { "epoch": 1.4454903956239664, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7133097648620605, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8840845823287964, "num_tokens": 431476343.0, "step": 11363 }, { "epoch": 1.445617605902557, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.548811435699463, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8777493238449097, "num_tokens": 431513741.0, "step": 11364 }, { "epoch": 1.4457448161811475, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7168049812316895, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8582021594047546, "num_tokens": 431556227.0, "step": 11365 }, { "epoch": 1.445872026459738, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.552572250366211, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.872920572757721, "num_tokens": 431592004.0, "step": 11366 }, { "epoch": 1.4459992367383285, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7113025188446045, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.866552472114563, "num_tokens": 431631286.0, "step": 11367 }, { "epoch": 1.446126447016919, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.547503709793091, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8686263561248779, "num_tokens": 431669269.0, "step": 11368 }, { "epoch": 1.4462536572955096, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7206506729125977, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8625934720039368, "num_tokens": 431708061.0, "step": 11369 }, { "epoch": 1.4463808675741001, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.546168804168701, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8750358819961548, "num_tokens": 431745855.0, "step": 11370 }, { "epoch": 1.4465080778526904, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7196125984191895, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8723539710044861, "num_tokens": 431784481.0, "step": 11371 }, { "epoch": 1.446635288131281, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5421929359436035, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8641238212585449, "num_tokens": 431822724.0, "step": 11372 }, { "epoch": 1.4467624984098715, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7157673835754395, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8763061165809631, "num_tokens": 431859415.0, "step": 11373 }, { "epoch": 1.446889708688462, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5295426845550537, "learning_rate": 1e-05, "loss": 0.3819, "mean_token_accuracy": 0.8911334872245789, "num_tokens": 431899432.0, "step": 11374 }, { "epoch": 1.4470169189670525, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7227120399475098, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8756463527679443, "num_tokens": 431935546.0, "step": 11375 }, { "epoch": 1.447144129245643, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5416100025177, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8728082180023193, "num_tokens": 431972366.0, "step": 11376 }, { "epoch": 1.4472713395242336, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.729661226272583, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.858648419380188, "num_tokens": 432007795.0, "step": 11377 }, { "epoch": 1.4473985498028241, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5518386363983154, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8503912687301636, "num_tokens": 432047589.0, "step": 11378 }, { "epoch": 1.4475257600814146, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7107150554656982, "learning_rate": 1e-05, "loss": 0.4044, "mean_token_accuracy": 0.8804225921630859, "num_tokens": 432085615.0, "step": 11379 }, { "epoch": 1.447652970360005, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5469717979431152, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8761018514633179, "num_tokens": 432121112.0, "step": 11380 }, { "epoch": 1.4477801806385955, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7123565673828125, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8706693053245544, "num_tokens": 432156429.0, "step": 11381 }, { "epoch": 1.447907390917186, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.555537462234497, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.8563412427902222, "num_tokens": 432195653.0, "step": 11382 }, { "epoch": 1.4480346011957765, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.708967447280884, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8717019557952881, "num_tokens": 432230997.0, "step": 11383 }, { "epoch": 1.448161811474367, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.544677257537842, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8633782267570496, "num_tokens": 432270868.0, "step": 11384 }, { "epoch": 1.4482890217529576, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.701446056365967, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8732905387878418, "num_tokens": 432309999.0, "step": 11385 }, { "epoch": 1.4484162320315481, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5416624546051025, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8822705745697021, "num_tokens": 432345240.0, "step": 11386 }, { "epoch": 1.4485434423101387, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6971242427825928, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8626265525817871, "num_tokens": 432391478.0, "step": 11387 }, { "epoch": 1.4486706525887292, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.532477378845215, "learning_rate": 1e-05, "loss": 0.3873, "mean_token_accuracy": 0.8859849572181702, "num_tokens": 432430601.0, "step": 11388 }, { "epoch": 1.4487978628673197, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.721942186355591, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8678812980651855, "num_tokens": 432467260.0, "step": 11389 }, { "epoch": 1.4489250731459102, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.542072296142578, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.855343759059906, "num_tokens": 432506270.0, "step": 11390 }, { "epoch": 1.4490522834245008, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.719588041305542, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8674076795578003, "num_tokens": 432543688.0, "step": 11391 }, { "epoch": 1.4491794937030913, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5303263664245605, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8797014951705933, "num_tokens": 432583522.0, "step": 11392 }, { "epoch": 1.4493067039816818, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7184195518493652, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.885322093963623, "num_tokens": 432622471.0, "step": 11393 }, { "epoch": 1.4494339142602723, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.527702808380127, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8827859163284302, "num_tokens": 432666689.0, "step": 11394 }, { "epoch": 1.4495611245388629, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7311220169067383, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8759440183639526, "num_tokens": 432707816.0, "step": 11395 }, { "epoch": 1.4496883348174532, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.526212453842163, "learning_rate": 1e-05, "loss": 0.3935, "mean_token_accuracy": 0.8854263424873352, "num_tokens": 432746873.0, "step": 11396 }, { "epoch": 1.4498155450960437, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.752720832824707, "learning_rate": 1e-05, "loss": 0.3961, "mean_token_accuracy": 0.8844617605209351, "num_tokens": 432777825.0, "step": 11397 }, { "epoch": 1.4499427553746342, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5183470249176025, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8694894909858704, "num_tokens": 432816806.0, "step": 11398 }, { "epoch": 1.4500699656532248, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.761070966720581, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8803690671920776, "num_tokens": 432855726.0, "step": 11399 }, { "epoch": 1.4501971759318153, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5411722660064697, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8624210357666016, "num_tokens": 432891526.0, "step": 11400 }, { "epoch": 1.4503243862104058, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7635536193847656, "learning_rate": 1e-05, "loss": 0.3661, "mean_token_accuracy": 0.8946563005447388, "num_tokens": 432921995.0, "step": 11401 }, { "epoch": 1.4504515964889964, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.536356210708618, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8667982816696167, "num_tokens": 432962120.0, "step": 11402 }, { "epoch": 1.4505788067675869, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.764045238494873, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.870538592338562, "num_tokens": 433001944.0, "step": 11403 }, { "epoch": 1.4507060170461774, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5501930713653564, "learning_rate": 1e-05, "loss": 0.4931, "mean_token_accuracy": 0.8505477905273438, "num_tokens": 433040984.0, "step": 11404 }, { "epoch": 1.4508332273247677, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7672653198242188, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8683239817619324, "num_tokens": 433077914.0, "step": 11405 }, { "epoch": 1.4509604376033582, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5530989170074463, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8803541660308838, "num_tokens": 433111219.0, "step": 11406 }, { "epoch": 1.4510876478819488, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.747657060623169, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8875413537025452, "num_tokens": 433143204.0, "step": 11407 }, { "epoch": 1.4512148581605393, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5582292079925537, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8870981931686401, "num_tokens": 433175361.0, "step": 11408 }, { "epoch": 1.4513420684391298, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.732534408569336, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8808297514915466, "num_tokens": 433215303.0, "step": 11409 }, { "epoch": 1.4514692787177204, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.542851209640503, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8771851062774658, "num_tokens": 433255301.0, "step": 11410 }, { "epoch": 1.4515964889963109, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.746790647506714, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8625594973564148, "num_tokens": 433291110.0, "step": 11411 }, { "epoch": 1.4517236992749014, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.546586036682129, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8709046840667725, "num_tokens": 433330910.0, "step": 11412 }, { "epoch": 1.451850909553492, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7507200241088867, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8738125562667847, "num_tokens": 433364297.0, "step": 11413 }, { "epoch": 1.4519781198320825, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5383143424987793, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8688082098960876, "num_tokens": 433405900.0, "step": 11414 }, { "epoch": 1.452105330110673, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.75846266746521, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.862679123878479, "num_tokens": 433442639.0, "step": 11415 }, { "epoch": 1.4522325403892635, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.571943998336792, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.866629958152771, "num_tokens": 433471657.0, "step": 11416 }, { "epoch": 1.452359750667854, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7365307807922363, "learning_rate": 1e-05, "loss": 0.4917, "mean_token_accuracy": 0.8526858687400818, "num_tokens": 433510936.0, "step": 11417 }, { "epoch": 1.4524869609464446, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.557786226272583, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8696198463439941, "num_tokens": 433546940.0, "step": 11418 }, { "epoch": 1.4526141712250351, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.72306752204895, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8734538555145264, "num_tokens": 433588701.0, "step": 11419 }, { "epoch": 1.4527413815036254, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5503766536712646, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8708144426345825, "num_tokens": 433625678.0, "step": 11420 }, { "epoch": 1.452868591782216, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.737196207046509, "learning_rate": 1e-05, "loss": 0.5065, "mean_token_accuracy": 0.8548517823219299, "num_tokens": 433660624.0, "step": 11421 }, { "epoch": 1.4529958020608065, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5377612113952637, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8795162439346313, "num_tokens": 433702737.0, "step": 11422 }, { "epoch": 1.453123012339397, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.729200839996338, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8703141212463379, "num_tokens": 433741866.0, "step": 11423 }, { "epoch": 1.4532502226179875, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5459446907043457, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8627485036849976, "num_tokens": 433779869.0, "step": 11424 }, { "epoch": 1.453377432896578, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.726576805114746, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8825870156288147, "num_tokens": 433819629.0, "step": 11425 }, { "epoch": 1.4535046431751686, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.544750452041626, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.877717137336731, "num_tokens": 433855330.0, "step": 11426 }, { "epoch": 1.4536318534537591, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.736530303955078, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8646150827407837, "num_tokens": 433896130.0, "step": 11427 }, { "epoch": 1.4537590637323496, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5474674701690674, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8810831308364868, "num_tokens": 433928514.0, "step": 11428 }, { "epoch": 1.45388627401094, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.725902795791626, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8674423694610596, "num_tokens": 433969213.0, "step": 11429 }, { "epoch": 1.4540134842895305, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.552875280380249, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8573201894760132, "num_tokens": 434007902.0, "step": 11430 }, { "epoch": 1.454140694568121, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.735640287399292, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8589792847633362, "num_tokens": 434043243.0, "step": 11431 }, { "epoch": 1.4542679048467115, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5470187664031982, "learning_rate": 1e-05, "loss": 0.3911, "mean_token_accuracy": 0.8842864036560059, "num_tokens": 434081075.0, "step": 11432 }, { "epoch": 1.454395115125302, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.725313425064087, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8701505661010742, "num_tokens": 434121230.0, "step": 11433 }, { "epoch": 1.4545223254038926, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5444014072418213, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8669651746749878, "num_tokens": 434158324.0, "step": 11434 }, { "epoch": 1.4546495356824831, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7171549797058105, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8731993436813354, "num_tokens": 434203910.0, "step": 11435 }, { "epoch": 1.4547767459610736, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5467607975006104, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8823714256286621, "num_tokens": 434238915.0, "step": 11436 }, { "epoch": 1.4549039562396642, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7140424251556396, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8744750022888184, "num_tokens": 434280049.0, "step": 11437 }, { "epoch": 1.4550311665182547, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5364491939544678, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8817137479782104, "num_tokens": 434323980.0, "step": 11438 }, { "epoch": 1.4551583767968452, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7253544330596924, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8663275241851807, "num_tokens": 434363386.0, "step": 11439 }, { "epoch": 1.4552855870754358, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5349507331848145, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8825117349624634, "num_tokens": 434399480.0, "step": 11440 }, { "epoch": 1.4554127973540263, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7237844467163086, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8690937757492065, "num_tokens": 434445662.0, "step": 11441 }, { "epoch": 1.4555400076326168, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.544846296310425, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8740566968917847, "num_tokens": 434480926.0, "step": 11442 }, { "epoch": 1.4556672179112073, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7343199253082275, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8612889051437378, "num_tokens": 434520018.0, "step": 11443 }, { "epoch": 1.4557944281897979, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5534422397613525, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8594213128089905, "num_tokens": 434555163.0, "step": 11444 }, { "epoch": 1.4559216384683882, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.745511531829834, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8829922676086426, "num_tokens": 434586353.0, "step": 11445 }, { "epoch": 1.4560488487469787, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.534656524658203, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8774324655532837, "num_tokens": 434630558.0, "step": 11446 }, { "epoch": 1.4561760590255692, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7415668964385986, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8709465265274048, "num_tokens": 434672949.0, "step": 11447 }, { "epoch": 1.4563032693041598, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5412330627441406, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8671597242355347, "num_tokens": 434709207.0, "step": 11448 }, { "epoch": 1.4564304795827503, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.735757350921631, "learning_rate": 1e-05, "loss": 0.4025, "mean_token_accuracy": 0.8827073574066162, "num_tokens": 434745787.0, "step": 11449 }, { "epoch": 1.4565576898613408, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.542295455932617, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8725860118865967, "num_tokens": 434785419.0, "step": 11450 }, { "epoch": 1.4566849001399313, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.747875213623047, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8639854192733765, "num_tokens": 434818217.0, "step": 11451 }, { "epoch": 1.4568121104185219, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.542711019515991, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8622389435768127, "num_tokens": 434855654.0, "step": 11452 }, { "epoch": 1.4569393206971124, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.760606527328491, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8581412434577942, "num_tokens": 434890657.0, "step": 11453 }, { "epoch": 1.4570665309757027, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5322723388671875, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8819897770881653, "num_tokens": 434929164.0, "step": 11454 }, { "epoch": 1.4571937412542932, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000255584716796875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7608754634857178, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8647069334983826, "num_tokens": 434966017.0, "step": 11455 }, { "epoch": 1.4573209515328838, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.535125970840454, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8808540105819702, "num_tokens": 435002076.0, "step": 11456 }, { "epoch": 1.4574481618114743, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7702157497406006, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.879159152507782, "num_tokens": 435030058.0, "step": 11457 }, { "epoch": 1.4575753720900648, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.541015386581421, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8651142716407776, "num_tokens": 435067816.0, "step": 11458 }, { "epoch": 1.4577025823686554, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.760711431503296, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8755639791488647, "num_tokens": 435106078.0, "step": 11459 }, { "epoch": 1.4578297926472459, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5308001041412354, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8826045393943787, "num_tokens": 435144558.0, "step": 11460 }, { "epoch": 1.4579570029258364, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7603137493133545, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.867820143699646, "num_tokens": 435183083.0, "step": 11461 }, { "epoch": 1.458084213204427, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5253992080688477, "learning_rate": 1e-05, "loss": 0.3788, "mean_token_accuracy": 0.8887200951576233, "num_tokens": 435223459.0, "step": 11462 }, { "epoch": 1.4582114234830175, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.765437126159668, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8779218792915344, "num_tokens": 435261491.0, "step": 11463 }, { "epoch": 1.458338633761608, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5516281127929688, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.860859751701355, "num_tokens": 435299634.0, "step": 11464 }, { "epoch": 1.4584658440401985, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.764430284500122, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8558553457260132, "num_tokens": 435339645.0, "step": 11465 }, { "epoch": 1.458593054318789, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.53528094291687, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8701870441436768, "num_tokens": 435382635.0, "step": 11466 }, { "epoch": 1.4587202645973796, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7653565406799316, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8626998662948608, "num_tokens": 435421653.0, "step": 11467 }, { "epoch": 1.45884747487597, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.550197124481201, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8617777824401855, "num_tokens": 435460135.0, "step": 11468 }, { "epoch": 1.4589746851545604, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7528765201568604, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8868273496627808, "num_tokens": 435497717.0, "step": 11469 }, { "epoch": 1.459101895433151, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.557218313217163, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8596899509429932, "num_tokens": 435535247.0, "step": 11470 }, { "epoch": 1.4592291057117415, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7450482845306396, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8789166212081909, "num_tokens": 435576841.0, "step": 11471 }, { "epoch": 1.459356315990332, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5424094200134277, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.875020444393158, "num_tokens": 435618197.0, "step": 11472 }, { "epoch": 1.4594835262689225, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.75518536567688, "learning_rate": 1e-05, "loss": 0.5239, "mean_token_accuracy": 0.848737359046936, "num_tokens": 435663327.0, "step": 11473 }, { "epoch": 1.459610736547513, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.560671806335449, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8701558113098145, "num_tokens": 435697795.0, "step": 11474 }, { "epoch": 1.4597379468261036, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.742469072341919, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8620046973228455, "num_tokens": 435734308.0, "step": 11475 }, { "epoch": 1.4598651571046941, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.546175479888916, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8645801544189453, "num_tokens": 435772575.0, "step": 11476 }, { "epoch": 1.4599923673832846, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.739074945449829, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8833428025245667, "num_tokens": 435807578.0, "step": 11477 }, { "epoch": 1.460119577661875, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.544936418533325, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8706253170967102, "num_tokens": 435849035.0, "step": 11478 }, { "epoch": 1.4602467879404655, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7499916553497314, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8737819194793701, "num_tokens": 435884912.0, "step": 11479 }, { "epoch": 1.460373998219056, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.537449359893799, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8798101544380188, "num_tokens": 435928822.0, "step": 11480 }, { "epoch": 1.4605012084976465, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7483770847320557, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.879965603351593, "num_tokens": 435963906.0, "step": 11481 }, { "epoch": 1.460628418776237, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5488269329071045, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8552501201629639, "num_tokens": 436002826.0, "step": 11482 }, { "epoch": 1.4607556290548276, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.735090494155884, "learning_rate": 1e-05, "loss": 0.3917, "mean_token_accuracy": 0.8867736458778381, "num_tokens": 436039043.0, "step": 11483 }, { "epoch": 1.4608828393334181, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.543584108352661, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8807843923568726, "num_tokens": 436078081.0, "step": 11484 }, { "epoch": 1.4610100496120086, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7638466358184814, "learning_rate": 1e-05, "loss": 0.4888, "mean_token_accuracy": 0.8563989400863647, "num_tokens": 436110415.0, "step": 11485 }, { "epoch": 1.4611372598905992, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5287435054779053, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8801018595695496, "num_tokens": 436152473.0, "step": 11486 }, { "epoch": 1.4612644701691897, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7615742683410645, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8683137893676758, "num_tokens": 436192410.0, "step": 11487 }, { "epoch": 1.4613916804477802, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5367608070373535, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8618635535240173, "num_tokens": 436230605.0, "step": 11488 }, { "epoch": 1.4615188907263708, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7627112865448, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8776010274887085, "num_tokens": 436267951.0, "step": 11489 }, { "epoch": 1.4616461010049613, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5416810512542725, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8547223806381226, "num_tokens": 436306685.0, "step": 11490 }, { "epoch": 1.4617733112835518, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7592859268188477, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.874826192855835, "num_tokens": 436345568.0, "step": 11491 }, { "epoch": 1.4619005215621423, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5264716148376465, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8723922967910767, "num_tokens": 436387580.0, "step": 11492 }, { "epoch": 1.4620277318407329, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7762904167175293, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.868362545967102, "num_tokens": 436423000.0, "step": 11493 }, { "epoch": 1.4621549421193232, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.534842014312744, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8735653162002563, "num_tokens": 436464366.0, "step": 11494 }, { "epoch": 1.4622821523979137, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.771721839904785, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8600136637687683, "num_tokens": 436505385.0, "step": 11495 }, { "epoch": 1.4624093626765042, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.547495126724243, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8607372641563416, "num_tokens": 436546251.0, "step": 11496 }, { "epoch": 1.4625365729550948, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7733240127563477, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8622553944587708, "num_tokens": 436582136.0, "step": 11497 }, { "epoch": 1.4626637832336853, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.542844295501709, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8722084164619446, "num_tokens": 436620027.0, "step": 11498 }, { "epoch": 1.4627909935122758, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.779141426086426, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8677208423614502, "num_tokens": 436656720.0, "step": 11499 }, { "epoch": 1.4629182037908663, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.548271894454956, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.879845917224884, "num_tokens": 436693813.0, "step": 11500 }, { "epoch": 1.4630454140694569, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7608654499053955, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8752667903900146, "num_tokens": 436732023.0, "step": 11501 }, { "epoch": 1.4631726243480474, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.548705577850342, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8669052124023438, "num_tokens": 436772330.0, "step": 11502 }, { "epoch": 1.4632998346266377, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7511749267578125, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8667697906494141, "num_tokens": 436814593.0, "step": 11503 }, { "epoch": 1.4634270449052282, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5608069896698, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8717552423477173, "num_tokens": 436849951.0, "step": 11504 }, { "epoch": 1.4635542551838188, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7532501220703125, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8565518856048584, "num_tokens": 436889113.0, "step": 11505 }, { "epoch": 1.4636814654624093, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5640013217926025, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.880767822265625, "num_tokens": 436921009.0, "step": 11506 }, { "epoch": 1.4638086757409998, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.741274118423462, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8743130564689636, "num_tokens": 436958425.0, "step": 11507 }, { "epoch": 1.4639358860195903, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.564596652984619, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8740828037261963, "num_tokens": 436992995.0, "step": 11508 }, { "epoch": 1.4640630962981809, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7263755798339844, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8775848150253296, "num_tokens": 437028884.0, "step": 11509 }, { "epoch": 1.4641903065767714, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5542328357696533, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8749554753303528, "num_tokens": 437071753.0, "step": 11510 }, { "epoch": 1.464317516855362, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.723604202270508, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8801939487457275, "num_tokens": 437108783.0, "step": 11511 }, { "epoch": 1.4644447271339525, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.586411714553833, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.875745952129364, "num_tokens": 437143234.0, "step": 11512 }, { "epoch": 1.464571937412543, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6994822025299072, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8668851852416992, "num_tokens": 437182363.0, "step": 11513 }, { "epoch": 1.4646991476911335, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5510106086730957, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.875994861125946, "num_tokens": 437224680.0, "step": 11514 }, { "epoch": 1.464826357969724, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7122182846069336, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8735074400901794, "num_tokens": 437261154.0, "step": 11515 }, { "epoch": 1.4649535682483146, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5483486652374268, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8715195655822754, "num_tokens": 437301654.0, "step": 11516 }, { "epoch": 1.465080778526905, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7120859622955322, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8704888820648193, "num_tokens": 437338437.0, "step": 11517 }, { "epoch": 1.4652079888054954, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.543767213821411, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8660805225372314, "num_tokens": 437384258.0, "step": 11518 }, { "epoch": 1.465335199084086, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7289888858795166, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8583897352218628, "num_tokens": 437421762.0, "step": 11519 }, { "epoch": 1.4654624093626765, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5442793369293213, "learning_rate": 1e-05, "loss": 0.3998, "mean_token_accuracy": 0.8800942897796631, "num_tokens": 437456797.0, "step": 11520 }, { "epoch": 1.465589619641267, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7488484382629395, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8823436498641968, "num_tokens": 437484435.0, "step": 11521 }, { "epoch": 1.4657168299198575, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.534855842590332, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8710557222366333, "num_tokens": 437522573.0, "step": 11522 }, { "epoch": 1.465844040198448, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.750091791152954, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8600084781646729, "num_tokens": 437560204.0, "step": 11523 }, { "epoch": 1.4659712504770386, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5394415855407715, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.867538571357727, "num_tokens": 437595845.0, "step": 11524 }, { "epoch": 1.466098460755629, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7419981956481934, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.875751256942749, "num_tokens": 437631985.0, "step": 11525 }, { "epoch": 1.4662256710342196, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5364959239959717, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8620114326477051, "num_tokens": 437677024.0, "step": 11526 }, { "epoch": 1.46635288131281, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.741941213607788, "learning_rate": 1e-05, "loss": 0.391, "mean_token_accuracy": 0.8874301314353943, "num_tokens": 437714980.0, "step": 11527 }, { "epoch": 1.4664800915914005, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5412561893463135, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8691859245300293, "num_tokens": 437752444.0, "step": 11528 }, { "epoch": 1.466607301869991, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7506003379821777, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8661986589431763, "num_tokens": 437787032.0, "step": 11529 }, { "epoch": 1.4667345121485815, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5325534343719482, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8697521686553955, "num_tokens": 437827560.0, "step": 11530 }, { "epoch": 1.466861722427172, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7558560371398926, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8671108484268188, "num_tokens": 437865339.0, "step": 11531 }, { "epoch": 1.4669889327057626, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5451393127441406, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8712923526763916, "num_tokens": 437899769.0, "step": 11532 }, { "epoch": 1.467116142984353, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.758474349975586, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.874833345413208, "num_tokens": 437934738.0, "step": 11533 }, { "epoch": 1.4672433532629436, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5365853309631348, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8649384379386902, "num_tokens": 437975948.0, "step": 11534 }, { "epoch": 1.4673705635415342, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.764023542404175, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8797103762626648, "num_tokens": 438010081.0, "step": 11535 }, { "epoch": 1.4674977738201247, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5505964756011963, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8748528957366943, "num_tokens": 438044069.0, "step": 11536 }, { "epoch": 1.4676249840987152, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.758845329284668, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8741234540939331, "num_tokens": 438080866.0, "step": 11537 }, { "epoch": 1.4677521943773058, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5431876182556152, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8717184066772461, "num_tokens": 438120550.0, "step": 11538 }, { "epoch": 1.4678794046558963, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7552828788757324, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8669039011001587, "num_tokens": 438156968.0, "step": 11539 }, { "epoch": 1.4680066149344868, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5581157207489014, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.873106062412262, "num_tokens": 438191470.0, "step": 11540 }, { "epoch": 1.4681338252130773, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.768979072570801, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8585177063941956, "num_tokens": 438225944.0, "step": 11541 }, { "epoch": 1.4682610354916679, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.544734239578247, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8717857003211975, "num_tokens": 438263759.0, "step": 11542 }, { "epoch": 1.4683882457702582, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.74601149559021, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.873691976070404, "num_tokens": 438302371.0, "step": 11543 }, { "epoch": 1.4685154560488487, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5520572662353516, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8534340262413025, "num_tokens": 438342006.0, "step": 11544 }, { "epoch": 1.4686426663274392, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7662129402160645, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8635405898094177, "num_tokens": 438375649.0, "step": 11545 }, { "epoch": 1.4687698766060298, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.539924383163452, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.877762496471405, "num_tokens": 438417566.0, "step": 11546 }, { "epoch": 1.4688970868846203, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7643182277679443, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8782469034194946, "num_tokens": 438452219.0, "step": 11547 }, { "epoch": 1.4690242971632108, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.52881121635437, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8833526372909546, "num_tokens": 438490503.0, "step": 11548 }, { "epoch": 1.4691515074418013, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.786018133163452, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8846445679664612, "num_tokens": 438521792.0, "step": 11549 }, { "epoch": 1.4692787177203919, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.601693868637085, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8800821304321289, "num_tokens": 438557408.0, "step": 11550 }, { "epoch": 1.4694059279989824, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7245051860809326, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.8841829299926758, "num_tokens": 438601424.0, "step": 11551 }, { "epoch": 1.4695331382775727, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5653581619262695, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8764904737472534, "num_tokens": 438638772.0, "step": 11552 }, { "epoch": 1.4696603485561632, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7279486656188965, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.86761075258255, "num_tokens": 438676906.0, "step": 11553 }, { "epoch": 1.4697875588347538, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.57218074798584, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8562458157539368, "num_tokens": 438715790.0, "step": 11554 }, { "epoch": 1.4699147691133443, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.711890935897827, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8764872550964355, "num_tokens": 438757739.0, "step": 11555 }, { "epoch": 1.4700419793919348, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5645365715026855, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.87260502576828, "num_tokens": 438797336.0, "step": 11556 }, { "epoch": 1.4701691896705253, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.707261085510254, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.868577241897583, "num_tokens": 438833452.0, "step": 11557 }, { "epoch": 1.4702963999491159, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.551051378250122, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.873897135257721, "num_tokens": 438874346.0, "step": 11558 }, { "epoch": 1.4704236102277064, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.723299980163574, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8592929840087891, "num_tokens": 438908888.0, "step": 11559 }, { "epoch": 1.470550820506297, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.542198896408081, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8756811618804932, "num_tokens": 438953127.0, "step": 11560 }, { "epoch": 1.4706780307848875, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7390613555908203, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8687726259231567, "num_tokens": 438986485.0, "step": 11561 }, { "epoch": 1.470805241063478, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.525541067123413, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8713697195053101, "num_tokens": 439030958.0, "step": 11562 }, { "epoch": 1.4709324513420685, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.746680498123169, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8635067939758301, "num_tokens": 439069139.0, "step": 11563 }, { "epoch": 1.471059661620659, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5302977561950684, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8782932758331299, "num_tokens": 439106526.0, "step": 11564 }, { "epoch": 1.4711868718992496, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.752495050430298, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8734467029571533, "num_tokens": 439146037.0, "step": 11565 }, { "epoch": 1.47131408217784, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.544142961502075, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8825588226318359, "num_tokens": 439177239.0, "step": 11566 }, { "epoch": 1.4714412924564304, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.758336305618286, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8733664155006409, "num_tokens": 439211009.0, "step": 11567 }, { "epoch": 1.471568502735021, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.534470558166504, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.87370765209198, "num_tokens": 439250656.0, "step": 11568 }, { "epoch": 1.4716957130136115, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.769798994064331, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8544389605522156, "num_tokens": 439285103.0, "step": 11569 }, { "epoch": 1.471822923292202, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.540832996368408, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8624013066291809, "num_tokens": 439325104.0, "step": 11570 }, { "epoch": 1.4719501335707925, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7783353328704834, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.858651876449585, "num_tokens": 439357760.0, "step": 11571 }, { "epoch": 1.472077343849383, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5351758003234863, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8679363131523132, "num_tokens": 439397823.0, "step": 11572 }, { "epoch": 1.4722045541279736, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7834043502807617, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8632553219795227, "num_tokens": 439435989.0, "step": 11573 }, { "epoch": 1.472331764406564, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.532336711883545, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8806661367416382, "num_tokens": 439474272.0, "step": 11574 }, { "epoch": 1.4724589746851546, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.780517578125, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8810567855834961, "num_tokens": 439512729.0, "step": 11575 }, { "epoch": 1.472586184963745, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.553293466567993, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8604142665863037, "num_tokens": 439550368.0, "step": 11576 }, { "epoch": 1.4727133952423355, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7697513103485107, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8713027238845825, "num_tokens": 439591244.0, "step": 11577 }, { "epoch": 1.472840605520926, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.545742988586426, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8680351972579956, "num_tokens": 439632389.0, "step": 11578 }, { "epoch": 1.4729678157995165, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7616918087005615, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8589142560958862, "num_tokens": 439674543.0, "step": 11579 }, { "epoch": 1.473095026078107, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5670995712280273, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8703336715698242, "num_tokens": 439709009.0, "step": 11580 }, { "epoch": 1.4732222363566976, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7353153228759766, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8748327493667603, "num_tokens": 439754154.0, "step": 11581 }, { "epoch": 1.473349446635288, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5497238636016846, "learning_rate": 1e-05, "loss": 0.3825, "mean_token_accuracy": 0.8904054164886475, "num_tokens": 439794986.0, "step": 11582 }, { "epoch": 1.4734766569138786, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7411341667175293, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8663915991783142, "num_tokens": 439833402.0, "step": 11583 }, { "epoch": 1.4736038671924692, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.541966676712036, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.884116530418396, "num_tokens": 439878670.0, "step": 11584 }, { "epoch": 1.4737310774710597, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7353336811065674, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8794741034507751, "num_tokens": 439916505.0, "step": 11585 }, { "epoch": 1.4738582877496502, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.549110174179077, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8736938834190369, "num_tokens": 439955833.0, "step": 11586 }, { "epoch": 1.4739854980282407, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.736968517303467, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8835530281066895, "num_tokens": 439994105.0, "step": 11587 }, { "epoch": 1.4741127083068313, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5555055141448975, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.877924382686615, "num_tokens": 440030298.0, "step": 11588 }, { "epoch": 1.4742399185854218, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7391438484191895, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8776841163635254, "num_tokens": 440064792.0, "step": 11589 }, { "epoch": 1.4743671288640123, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5715551376342773, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.871487557888031, "num_tokens": 440097886.0, "step": 11590 }, { "epoch": 1.4744943391426026, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.720141887664795, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8762401342391968, "num_tokens": 440133488.0, "step": 11591 }, { "epoch": 1.4746215494211932, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5526652336120605, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8783900141716003, "num_tokens": 440169448.0, "step": 11592 }, { "epoch": 1.4747487596997837, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7182235717773438, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.8795188665390015, "num_tokens": 440204755.0, "step": 11593 }, { "epoch": 1.4748759699783742, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5734238624572754, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8683732748031616, "num_tokens": 440239129.0, "step": 11594 }, { "epoch": 1.4750031802569648, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6991448402404785, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8726054430007935, "num_tokens": 440280962.0, "step": 11595 }, { "epoch": 1.4751303905355553, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5622034072875977, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.877602219581604, "num_tokens": 440315900.0, "step": 11596 }, { "epoch": 1.4752576008141458, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.677687406539917, "learning_rate": 1e-05, "loss": 0.4098, "mean_token_accuracy": 0.8792823553085327, "num_tokens": 440356699.0, "step": 11597 }, { "epoch": 1.4753848110927363, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5532405376434326, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8783519268035889, "num_tokens": 440398655.0, "step": 11598 }, { "epoch": 1.4755120213713269, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6948962211608887, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8639346361160278, "num_tokens": 440436204.0, "step": 11599 }, { "epoch": 1.4756392316499174, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5428993701934814, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8740353584289551, "num_tokens": 440478363.0, "step": 11600 }, { "epoch": 1.4757664419285077, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.690290927886963, "learning_rate": 1e-05, "loss": 0.3968, "mean_token_accuracy": 0.8833622932434082, "num_tokens": 440515454.0, "step": 11601 }, { "epoch": 1.4758936522070982, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5516014099121094, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8722819089889526, "num_tokens": 440553526.0, "step": 11602 }, { "epoch": 1.4760208624856888, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6966686248779297, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8610355854034424, "num_tokens": 440591446.0, "step": 11603 }, { "epoch": 1.4761480727642793, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.547780752182007, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8806769847869873, "num_tokens": 440630505.0, "step": 11604 }, { "epoch": 1.4762752830428698, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.693070888519287, "learning_rate": 1e-05, "loss": 0.3857, "mean_token_accuracy": 0.8856747150421143, "num_tokens": 440669318.0, "step": 11605 }, { "epoch": 1.4764024933214603, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5310447216033936, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.8804750442504883, "num_tokens": 440711218.0, "step": 11606 }, { "epoch": 1.4765297036000509, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7052667140960693, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8715143203735352, "num_tokens": 440754581.0, "step": 11607 }, { "epoch": 1.4766569138786414, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.535256862640381, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8716915845870972, "num_tokens": 440795714.0, "step": 11608 }, { "epoch": 1.476784124157232, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7281441688537598, "learning_rate": 1e-05, "loss": 0.3811, "mean_token_accuracy": 0.8871679306030273, "num_tokens": 440827423.0, "step": 11609 }, { "epoch": 1.4769113344358225, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.532534122467041, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8626974821090698, "num_tokens": 440868857.0, "step": 11610 }, { "epoch": 1.477038544714413, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7595291137695312, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8578834533691406, "num_tokens": 440904788.0, "step": 11611 }, { "epoch": 1.4771657549930035, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.516399383544922, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8710474967956543, "num_tokens": 440949013.0, "step": 11612 }, { "epoch": 1.477292965271594, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.771209478378296, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8605778217315674, "num_tokens": 440987392.0, "step": 11613 }, { "epoch": 1.4774201755501846, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5124781131744385, "learning_rate": 1e-05, "loss": 0.3453, "mean_token_accuracy": 0.8985799551010132, "num_tokens": 441024917.0, "step": 11614 }, { "epoch": 1.477547385828775, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.77807354927063, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8792669773101807, "num_tokens": 441066355.0, "step": 11615 }, { "epoch": 1.4776745961073654, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5341644287109375, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8649424314498901, "num_tokens": 441100192.0, "step": 11616 }, { "epoch": 1.477801806385956, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.787997007369995, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8629074096679688, "num_tokens": 441144627.0, "step": 11617 }, { "epoch": 1.4779290166645465, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.533872127532959, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8757469654083252, "num_tokens": 441181656.0, "step": 11618 }, { "epoch": 1.478056226943137, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.800656318664551, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8677750825881958, "num_tokens": 441215476.0, "step": 11619 }, { "epoch": 1.4781834372217275, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.530880928039551, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.866576611995697, "num_tokens": 441262919.0, "step": 11620 }, { "epoch": 1.478310647500318, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8027031421661377, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8693830966949463, "num_tokens": 441302602.0, "step": 11621 }, { "epoch": 1.4784378577789086, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5508596897125244, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8741685152053833, "num_tokens": 441335432.0, "step": 11622 }, { "epoch": 1.478565068057499, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7949068546295166, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8633430600166321, "num_tokens": 441377849.0, "step": 11623 }, { "epoch": 1.4786922783360896, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.568527936935425, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8726032376289368, "num_tokens": 441413225.0, "step": 11624 }, { "epoch": 1.47881948861468, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7815849781036377, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8784692883491516, "num_tokens": 441448041.0, "step": 11625 }, { "epoch": 1.4789466988932705, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.563727378845215, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8839004039764404, "num_tokens": 441483919.0, "step": 11626 }, { "epoch": 1.479073909171861, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7692060470581055, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8680130839347839, "num_tokens": 441518313.0, "step": 11627 }, { "epoch": 1.4792011194504515, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5624029636383057, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8683040142059326, "num_tokens": 441557248.0, "step": 11628 }, { "epoch": 1.479328329729042, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.763305902481079, "learning_rate": 1e-05, "loss": 0.5373, "mean_token_accuracy": 0.8435341715812683, "num_tokens": 441599759.0, "step": 11629 }, { "epoch": 1.4794555400076326, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5794451236724854, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8677845597267151, "num_tokens": 441631643.0, "step": 11630 }, { "epoch": 1.479582750286223, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7448666095733643, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8609656095504761, "num_tokens": 441669804.0, "step": 11631 }, { "epoch": 1.4797099605648136, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.572964668273926, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8610933423042297, "num_tokens": 441708661.0, "step": 11632 }, { "epoch": 1.4798371708434042, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7252371311187744, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8722698092460632, "num_tokens": 441745510.0, "step": 11633 }, { "epoch": 1.4799643811219947, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.590318202972412, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.87119460105896, "num_tokens": 441781482.0, "step": 11634 }, { "epoch": 1.4800915914005852, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7066216468811035, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8817380666732788, "num_tokens": 441814299.0, "step": 11635 }, { "epoch": 1.4802188016791757, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.566244125366211, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8564493656158447, "num_tokens": 441859742.0, "step": 11636 }, { "epoch": 1.4803460119577663, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.698578357696533, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8831433057785034, "num_tokens": 441896563.0, "step": 11637 }, { "epoch": 1.4804732222363568, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.559866428375244, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8802512884140015, "num_tokens": 441935380.0, "step": 11638 }, { "epoch": 1.4806004325149473, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7034590244293213, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.862531304359436, "num_tokens": 441977077.0, "step": 11639 }, { "epoch": 1.4807276427935376, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.555185079574585, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8738076090812683, "num_tokens": 442020538.0, "step": 11640 }, { "epoch": 1.4808548530721282, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6989448070526123, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8737973570823669, "num_tokens": 442057093.0, "step": 11641 }, { "epoch": 1.4809820633507187, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5688116550445557, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8672237396240234, "num_tokens": 442092364.0, "step": 11642 }, { "epoch": 1.4811092736293092, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.693042039871216, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8673602342605591, "num_tokens": 442134181.0, "step": 11643 }, { "epoch": 1.4812364839078997, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.553668737411499, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8565310835838318, "num_tokens": 442174987.0, "step": 11644 }, { "epoch": 1.4813636941864903, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.7079074382781982, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8770349621772766, "num_tokens": 442204214.0, "step": 11645 }, { "epoch": 1.4814909044650808, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5445234775543213, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8814235925674438, "num_tokens": 442243227.0, "step": 11646 }, { "epoch": 1.4816181147436713, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.7069692611694336, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8726009130477905, "num_tokens": 442282107.0, "step": 11647 }, { "epoch": 1.4817453250222619, "ewc_loss": 0.0634765625, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5474300384521484, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8796427249908447, "num_tokens": 442320445.0, "step": 11648 }, { "epoch": 1.4818725353008524, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.718867063522339, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8660466074943542, "num_tokens": 442358626.0, "step": 11649 }, { "epoch": 1.4819997455794427, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5334889888763428, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8708341121673584, "num_tokens": 442405095.0, "step": 11650 }, { "epoch": 1.4821269558580332, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7241909503936768, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.870393693447113, "num_tokens": 442442131.0, "step": 11651 }, { "epoch": 1.4822541661366238, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.54856538772583, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8569999933242798, "num_tokens": 442479660.0, "step": 11652 }, { "epoch": 1.4823813764152143, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7210328578948975, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8666137456893921, "num_tokens": 442523659.0, "step": 11653 }, { "epoch": 1.4825085866938048, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.536740303039551, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8664441704750061, "num_tokens": 442564445.0, "step": 11654 }, { "epoch": 1.4826357969723953, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7451555728912354, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.872585117816925, "num_tokens": 442597662.0, "step": 11655 }, { "epoch": 1.4827630072509859, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5414912700653076, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.865705132484436, "num_tokens": 442633365.0, "step": 11656 }, { "epoch": 1.4828902175295764, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7493033409118652, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8701719641685486, "num_tokens": 442668389.0, "step": 11657 }, { "epoch": 1.483017427808167, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5720064640045166, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.87397301197052, "num_tokens": 442695089.0, "step": 11658 }, { "epoch": 1.4831446380867574, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7206404209136963, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8731735944747925, "num_tokens": 442738602.0, "step": 11659 }, { "epoch": 1.483271848365348, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.546734571456909, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8693864345550537, "num_tokens": 442777400.0, "step": 11660 }, { "epoch": 1.4833990586439385, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.721998691558838, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8739539980888367, "num_tokens": 442815502.0, "step": 11661 }, { "epoch": 1.483526268922529, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.545067548751831, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8615909218788147, "num_tokens": 442858418.0, "step": 11662 }, { "epoch": 1.4836534792011196, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002574920654296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7127013206481934, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8875830173492432, "num_tokens": 442896073.0, "step": 11663 }, { "epoch": 1.48378068947971, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.54472017288208, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8703361749649048, "num_tokens": 442940719.0, "step": 11664 }, { "epoch": 1.4839078997583004, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7235958576202393, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8813272714614868, "num_tokens": 442977724.0, "step": 11665 }, { "epoch": 1.484035110036891, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5380327701568604, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.881443977355957, "num_tokens": 443017754.0, "step": 11666 }, { "epoch": 1.4841623203154815, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7519266605377197, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8736038208007812, "num_tokens": 443048661.0, "step": 11667 }, { "epoch": 1.484289530594072, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5223515033721924, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8704460859298706, "num_tokens": 443094445.0, "step": 11668 }, { "epoch": 1.4844167408726625, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.749239444732666, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8833080530166626, "num_tokens": 443134468.0, "step": 11669 }, { "epoch": 1.484543951151253, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5347208976745605, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8755887150764465, "num_tokens": 443174458.0, "step": 11670 }, { "epoch": 1.4846711614298436, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7585391998291016, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8798734545707703, "num_tokens": 443212421.0, "step": 11671 }, { "epoch": 1.484798371708434, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.545132637023926, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.847489595413208, "num_tokens": 443252083.0, "step": 11672 }, { "epoch": 1.4849255819870246, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.767575740814209, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8626759648323059, "num_tokens": 443286776.0, "step": 11673 }, { "epoch": 1.485052792265615, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5281639099121094, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8783799409866333, "num_tokens": 443325734.0, "step": 11674 }, { "epoch": 1.4851800025442055, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.771599531173706, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8702509999275208, "num_tokens": 443362637.0, "step": 11675 }, { "epoch": 1.485307212822796, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5271167755126953, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8816567659378052, "num_tokens": 443401629.0, "step": 11676 }, { "epoch": 1.4854344231013865, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.778275728225708, "learning_rate": 1e-05, "loss": 0.3895, "mean_token_accuracy": 0.886918306350708, "num_tokens": 443437619.0, "step": 11677 }, { "epoch": 1.485561633379977, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5539162158966064, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8581138253211975, "num_tokens": 443472175.0, "step": 11678 }, { "epoch": 1.4856888436585676, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.778596878051758, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8665390610694885, "num_tokens": 443506904.0, "step": 11679 }, { "epoch": 1.485816053937158, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5469141006469727, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8597978353500366, "num_tokens": 443548141.0, "step": 11680 }, { "epoch": 1.4859432642157486, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7647294998168945, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8712445497512817, "num_tokens": 443586609.0, "step": 11681 }, { "epoch": 1.4860704744943392, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.547356367111206, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8746740818023682, "num_tokens": 443628792.0, "step": 11682 }, { "epoch": 1.4861976847729297, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7630455493927, "learning_rate": 1e-05, "loss": 0.5023, "mean_token_accuracy": 0.8537263870239258, "num_tokens": 443671336.0, "step": 11683 }, { "epoch": 1.4863248950515202, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5488052368164062, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.870747447013855, "num_tokens": 443708987.0, "step": 11684 }, { "epoch": 1.4864521053301107, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7487258911132812, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8733003735542297, "num_tokens": 443749968.0, "step": 11685 }, { "epoch": 1.4865793156087013, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5690464973449707, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8748542070388794, "num_tokens": 443786287.0, "step": 11686 }, { "epoch": 1.4867065258872918, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.742142915725708, "learning_rate": 1e-05, "loss": 0.3987, "mean_token_accuracy": 0.8848741054534912, "num_tokens": 443820729.0, "step": 11687 }, { "epoch": 1.4868337361658823, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.549720525741577, "learning_rate": 1e-05, "loss": 0.3881, "mean_token_accuracy": 0.8847107887268066, "num_tokens": 443859401.0, "step": 11688 }, { "epoch": 1.4869609464444726, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7355611324310303, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8882114291191101, "num_tokens": 443895697.0, "step": 11689 }, { "epoch": 1.4870881567230632, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.550647497177124, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8774161338806152, "num_tokens": 443935229.0, "step": 11690 }, { "epoch": 1.4872153670016537, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7494475841522217, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8703535795211792, "num_tokens": 443967544.0, "step": 11691 }, { "epoch": 1.4873425772802442, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.545393943786621, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8684598207473755, "num_tokens": 444006973.0, "step": 11692 }, { "epoch": 1.4874697875588347, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7439537048339844, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.8796154856681824, "num_tokens": 444041143.0, "step": 11693 }, { "epoch": 1.4875969978374253, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5390217304229736, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8632707595825195, "num_tokens": 444086279.0, "step": 11694 }, { "epoch": 1.4877242081160158, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7452549934387207, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8726806640625, "num_tokens": 444123559.0, "step": 11695 }, { "epoch": 1.4878514183946063, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.555025339126587, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8612952828407288, "num_tokens": 444159197.0, "step": 11696 }, { "epoch": 1.4879786286731969, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7607061862945557, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8740326166152954, "num_tokens": 444190217.0, "step": 11697 }, { "epoch": 1.4881058389517874, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.549886465072632, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.886999249458313, "num_tokens": 444225241.0, "step": 11698 }, { "epoch": 1.4882330492303777, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002593994140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7509572505950928, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8574104309082031, "num_tokens": 444263702.0, "step": 11699 }, { "epoch": 1.4883602595089682, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.539762496948242, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8727084398269653, "num_tokens": 444303911.0, "step": 11700 }, { "epoch": 1.4884874697875587, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.735748291015625, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8711965084075928, "num_tokens": 444344705.0, "step": 11701 }, { "epoch": 1.4886146800661493, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5487060546875, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8850754499435425, "num_tokens": 444381598.0, "step": 11702 }, { "epoch": 1.4887418903447398, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7523930072784424, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8770273327827454, "num_tokens": 444413077.0, "step": 11703 }, { "epoch": 1.4888691006233303, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5419976711273193, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8746067881584167, "num_tokens": 444453028.0, "step": 11704 }, { "epoch": 1.4889963109019209, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7575085163116455, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8692845106124878, "num_tokens": 444486382.0, "step": 11705 }, { "epoch": 1.4891235211805114, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.545379638671875, "learning_rate": 1e-05, "loss": 0.4013, "mean_token_accuracy": 0.8802323937416077, "num_tokens": 444523299.0, "step": 11706 }, { "epoch": 1.489250731459102, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.757795572280884, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8734431266784668, "num_tokens": 444561772.0, "step": 11707 }, { "epoch": 1.4893779417376924, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.541503667831421, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.859387993812561, "num_tokens": 444603056.0, "step": 11708 }, { "epoch": 1.489505152016283, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7565603256225586, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8843028545379639, "num_tokens": 444640469.0, "step": 11709 }, { "epoch": 1.4896323622948735, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.557435989379883, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8719663619995117, "num_tokens": 444680737.0, "step": 11710 }, { "epoch": 1.489759572573464, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.754025459289551, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.8849848508834839, "num_tokens": 444713828.0, "step": 11711 }, { "epoch": 1.4898867828520546, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.553511381149292, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8645628690719604, "num_tokens": 444755823.0, "step": 11712 }, { "epoch": 1.490013993130645, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.747450828552246, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8756805658340454, "num_tokens": 444791558.0, "step": 11713 }, { "epoch": 1.4901412034092354, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5391464233398438, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8771011233329773, "num_tokens": 444832037.0, "step": 11714 }, { "epoch": 1.490268413687826, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7589452266693115, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.860186755657196, "num_tokens": 444870092.0, "step": 11715 }, { "epoch": 1.4903956239664164, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5404186248779297, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8722158074378967, "num_tokens": 444909267.0, "step": 11716 }, { "epoch": 1.490522834245007, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7531545162200928, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8694373369216919, "num_tokens": 444946851.0, "step": 11717 }, { "epoch": 1.4906500445235975, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5408191680908203, "learning_rate": 1e-05, "loss": 0.3858, "mean_token_accuracy": 0.889204204082489, "num_tokens": 444986324.0, "step": 11718 }, { "epoch": 1.490777254802188, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.754722833633423, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8811485767364502, "num_tokens": 445020443.0, "step": 11719 }, { "epoch": 1.4909044650807786, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.558830499649048, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8550721406936646, "num_tokens": 445057592.0, "step": 11720 }, { "epoch": 1.491031675359369, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.74430513381958, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.8867039680480957, "num_tokens": 445091513.0, "step": 11721 }, { "epoch": 1.4911588856379596, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.548672914505005, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8579269051551819, "num_tokens": 445132579.0, "step": 11722 }, { "epoch": 1.49128609591655, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.760194778442383, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8514407873153687, "num_tokens": 445167336.0, "step": 11723 }, { "epoch": 1.4914133061951405, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.545823097229004, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8614031076431274, "num_tokens": 445208873.0, "step": 11724 }, { "epoch": 1.491540516473731, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.755359649658203, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8709872961044312, "num_tokens": 445245649.0, "step": 11725 }, { "epoch": 1.4916677267523215, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5387332439422607, "learning_rate": 1e-05, "loss": 0.3904, "mean_token_accuracy": 0.886140763759613, "num_tokens": 445285890.0, "step": 11726 }, { "epoch": 1.491794937030912, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.757441997528076, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8636857867240906, "num_tokens": 445323408.0, "step": 11727 }, { "epoch": 1.4919221473095026, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5438032150268555, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.881722092628479, "num_tokens": 445361086.0, "step": 11728 }, { "epoch": 1.492049357588093, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.744570016860962, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.875072717666626, "num_tokens": 445404726.0, "step": 11729 }, { "epoch": 1.4921765678666836, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5462334156036377, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8739262819290161, "num_tokens": 445444190.0, "step": 11730 }, { "epoch": 1.4923037781452742, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7569122314453125, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8641043901443481, "num_tokens": 445484395.0, "step": 11731 }, { "epoch": 1.4924309884238647, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.567803144454956, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8694945573806763, "num_tokens": 445518104.0, "step": 11732 }, { "epoch": 1.4925581987024552, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7479360103607178, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8633226752281189, "num_tokens": 445554761.0, "step": 11733 }, { "epoch": 1.4926854089810457, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.552689552307129, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8677175641059875, "num_tokens": 445595331.0, "step": 11734 }, { "epoch": 1.4928126192596363, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.721430778503418, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8781028389930725, "num_tokens": 445639280.0, "step": 11735 }, { "epoch": 1.4929398295382268, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.558342218399048, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8715940117835999, "num_tokens": 445675437.0, "step": 11736 }, { "epoch": 1.4930670398168173, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.729318380355835, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8764336109161377, "num_tokens": 445712419.0, "step": 11737 }, { "epoch": 1.4931942500954076, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.570054531097412, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8540216684341431, "num_tokens": 445751465.0, "step": 11738 }, { "epoch": 1.4933214603739982, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7244207859039307, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.8842122554779053, "num_tokens": 445788145.0, "step": 11739 }, { "epoch": 1.4934486706525887, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5613927841186523, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8773494958877563, "num_tokens": 445825954.0, "step": 11740 }, { "epoch": 1.4935758809311792, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.716212749481201, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.8585689067840576, "num_tokens": 445866207.0, "step": 11741 }, { "epoch": 1.4937030912097697, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.553168535232544, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8611212968826294, "num_tokens": 445904679.0, "step": 11742 }, { "epoch": 1.4938303014883603, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.735402822494507, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8589782118797302, "num_tokens": 445936001.0, "step": 11743 }, { "epoch": 1.4939575117669508, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.563887119293213, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.865734338760376, "num_tokens": 445969212.0, "step": 11744 }, { "epoch": 1.4940847220455413, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7012646198272705, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8697909712791443, "num_tokens": 446013455.0, "step": 11745 }, { "epoch": 1.4942119323241319, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5799145698547363, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.867424726486206, "num_tokens": 446044152.0, "step": 11746 }, { "epoch": 1.4943391426027224, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6881840229034424, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8575662970542908, "num_tokens": 446087047.0, "step": 11747 }, { "epoch": 1.4944663528813127, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5647776126861572, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8790569305419922, "num_tokens": 446120095.0, "step": 11748 }, { "epoch": 1.4945935631599032, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.689938545227051, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8770753741264343, "num_tokens": 446156827.0, "step": 11749 }, { "epoch": 1.4947207734384937, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5568487644195557, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8705406785011292, "num_tokens": 446197101.0, "step": 11750 }, { "epoch": 1.4948479837170843, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.695633888244629, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8692501187324524, "num_tokens": 446230442.0, "step": 11751 }, { "epoch": 1.4949751939956748, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5495314598083496, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.8812406063079834, "num_tokens": 446268395.0, "step": 11752 }, { "epoch": 1.4951024042742653, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.691761016845703, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8758276104927063, "num_tokens": 446307855.0, "step": 11753 }, { "epoch": 1.4952296145528559, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.544065237045288, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8712916374206543, "num_tokens": 446348808.0, "step": 11754 }, { "epoch": 1.4953568248314464, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7253572940826416, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8772704601287842, "num_tokens": 446380580.0, "step": 11755 }, { "epoch": 1.495484035110037, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.543057680130005, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8778281211853027, "num_tokens": 446414420.0, "step": 11756 }, { "epoch": 1.4956112453886274, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7080881595611572, "learning_rate": 1e-05, "loss": 0.3832, "mean_token_accuracy": 0.8889670372009277, "num_tokens": 446450220.0, "step": 11757 }, { "epoch": 1.495738455667218, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5512564182281494, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8862311840057373, "num_tokens": 446483823.0, "step": 11758 }, { "epoch": 1.4958656659458085, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.717085599899292, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8742258548736572, "num_tokens": 446522317.0, "step": 11759 }, { "epoch": 1.495992876224399, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.542419672012329, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8652288317680359, "num_tokens": 446564491.0, "step": 11760 }, { "epoch": 1.4961200865029896, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7165567874908447, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.874275803565979, "num_tokens": 446605490.0, "step": 11761 }, { "epoch": 1.49624729678158, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.539909601211548, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8611717224121094, "num_tokens": 446647385.0, "step": 11762 }, { "epoch": 1.4963745070601704, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.729902505874634, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8670766949653625, "num_tokens": 446682895.0, "step": 11763 }, { "epoch": 1.496501717338761, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5342085361480713, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8614529967308044, "num_tokens": 446724724.0, "step": 11764 }, { "epoch": 1.4966289276173514, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7389461994171143, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8727118372917175, "num_tokens": 446763433.0, "step": 11765 }, { "epoch": 1.496756137895942, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.529832363128662, "learning_rate": 1e-05, "loss": 0.5051, "mean_token_accuracy": 0.8470785617828369, "num_tokens": 446809275.0, "step": 11766 }, { "epoch": 1.4968833481745325, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.748053550720215, "learning_rate": 1e-05, "loss": 0.3925, "mean_token_accuracy": 0.8851237893104553, "num_tokens": 446840619.0, "step": 11767 }, { "epoch": 1.497010558453123, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.532153844833374, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8581110239028931, "num_tokens": 446879979.0, "step": 11768 }, { "epoch": 1.4971377687317136, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.754703998565674, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8820643424987793, "num_tokens": 446913729.0, "step": 11769 }, { "epoch": 1.497264979010304, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5360071659088135, "learning_rate": 1e-05, "loss": 0.4967, "mean_token_accuracy": 0.8586839437484741, "num_tokens": 446951843.0, "step": 11770 }, { "epoch": 1.4973921892888946, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7691891193389893, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.868503212928772, "num_tokens": 446988091.0, "step": 11771 }, { "epoch": 1.497519399567485, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.538555145263672, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8634568452835083, "num_tokens": 447023833.0, "step": 11772 }, { "epoch": 1.4976466098460754, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.773379325866699, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8777673244476318, "num_tokens": 447056232.0, "step": 11773 }, { "epoch": 1.497773820124666, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5356385707855225, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.881977915763855, "num_tokens": 447094524.0, "step": 11774 }, { "epoch": 1.4979010304032565, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7810862064361572, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8667452931404114, "num_tokens": 447129076.0, "step": 11775 }, { "epoch": 1.498028240681847, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.536710262298584, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8710076808929443, "num_tokens": 447168651.0, "step": 11776 }, { "epoch": 1.4981554509604376, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7786452770233154, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8712019324302673, "num_tokens": 447207491.0, "step": 11777 }, { "epoch": 1.498282661239028, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.548050880432129, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8788717985153198, "num_tokens": 447242955.0, "step": 11778 }, { "epoch": 1.4984098715176186, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.8350701332092285, "learning_rate": 1e-05, "loss": 0.3852, "mean_token_accuracy": 0.890083372592926, "num_tokens": 447277634.0, "step": 11779 }, { "epoch": 1.4985370817962091, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.540517568588257, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8656402230262756, "num_tokens": 447315255.0, "step": 11780 }, { "epoch": 1.4986642920747997, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7914047241210938, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.883216142654419, "num_tokens": 447353801.0, "step": 11781 }, { "epoch": 1.4987915023533902, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5520291328430176, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.869213879108429, "num_tokens": 447388098.0, "step": 11782 }, { "epoch": 1.4989187126319807, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.794593334197998, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.870450496673584, "num_tokens": 447424860.0, "step": 11783 }, { "epoch": 1.4990459229105713, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5577633380889893, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.852806031703949, "num_tokens": 447465179.0, "step": 11784 }, { "epoch": 1.4991731331891618, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7833032608032227, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8580994606018066, "num_tokens": 447505960.0, "step": 11785 }, { "epoch": 1.4993003434677523, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.560424327850342, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8620628118515015, "num_tokens": 447545410.0, "step": 11786 }, { "epoch": 1.4994275537463426, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7784974575042725, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8668451309204102, "num_tokens": 447586294.0, "step": 11787 }, { "epoch": 1.4995547640249332, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.553679943084717, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8730279803276062, "num_tokens": 447628841.0, "step": 11788 }, { "epoch": 1.4996819743035237, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.769054889678955, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8708760738372803, "num_tokens": 447669456.0, "step": 11789 }, { "epoch": 1.4998091845821142, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5708537101745605, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8779659271240234, "num_tokens": 447703297.0, "step": 11790 }, { "epoch": 1.4999363948607047, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.781376838684082, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8569382429122925, "num_tokens": 447742497.0, "step": 11791 }, { "epoch": 1.5000636051392953, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5543038845062256, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.871755063533783, "num_tokens": 447783956.0, "step": 11792 }, { "epoch": 1.5001908154178858, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7686080932617188, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8704659342765808, "num_tokens": 447824527.0, "step": 11793 }, { "epoch": 1.5003180256964763, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5603013038635254, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8682847023010254, "num_tokens": 447863676.0, "step": 11794 }, { "epoch": 1.5004452359750666, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7491984367370605, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8574672937393188, "num_tokens": 447906157.0, "step": 11795 }, { "epoch": 1.5005724462536572, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.558137893676758, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8852900862693787, "num_tokens": 447944016.0, "step": 11796 }, { "epoch": 1.5006996565322477, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.747096538543701, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8818303346633911, "num_tokens": 447979500.0, "step": 11797 }, { "epoch": 1.5008268668108382, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.569323778152466, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8570089340209961, "num_tokens": 448012770.0, "step": 11798 }, { "epoch": 1.5009540770894287, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7460973262786865, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8718369007110596, "num_tokens": 448048227.0, "step": 11799 }, { "epoch": 1.5010812873680193, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5791566371917725, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8615872859954834, "num_tokens": 448081672.0, "step": 11800 }, { "epoch": 1.5012084976466098, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.726895570755005, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8695693016052246, "num_tokens": 448122693.0, "step": 11801 }, { "epoch": 1.5013357079252003, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5748178958892822, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8612221479415894, "num_tokens": 448161246.0, "step": 11802 }, { "epoch": 1.5014629182037909, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.716371536254883, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.868114709854126, "num_tokens": 448205233.0, "step": 11803 }, { "epoch": 1.5015901284823814, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5657567977905273, "learning_rate": 1e-05, "loss": 0.5074, "mean_token_accuracy": 0.8481409549713135, "num_tokens": 448247404.0, "step": 11804 }, { "epoch": 1.501717338760972, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7151012420654297, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8744866847991943, "num_tokens": 448286874.0, "step": 11805 }, { "epoch": 1.5018445490395624, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.574899196624756, "learning_rate": 1e-05, "loss": 0.504, "mean_token_accuracy": 0.8518117666244507, "num_tokens": 448325461.0, "step": 11806 }, { "epoch": 1.501971759318153, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.68670392036438, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8781082034111023, "num_tokens": 448366871.0, "step": 11807 }, { "epoch": 1.5020989695967435, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.586515426635742, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8786841034889221, "num_tokens": 448396798.0, "step": 11808 }, { "epoch": 1.502226179875334, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.67887806892395, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.872109055519104, "num_tokens": 448437782.0, "step": 11809 }, { "epoch": 1.5023533901539246, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5829498767852783, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8745012283325195, "num_tokens": 448469892.0, "step": 11810 }, { "epoch": 1.502480600432515, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6684823036193848, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.866503119468689, "num_tokens": 448508233.0, "step": 11811 }, { "epoch": 1.5026078107111056, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5880215167999268, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.858132541179657, "num_tokens": 448541071.0, "step": 11812 }, { "epoch": 1.502735020989696, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6796133518218994, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8637591600418091, "num_tokens": 448571348.0, "step": 11813 }, { "epoch": 1.5028622312682864, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5726075172424316, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8719199895858765, "num_tokens": 448611889.0, "step": 11814 }, { "epoch": 1.502989441546877, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6697609424591064, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8765803575515747, "num_tokens": 448646683.0, "step": 11815 }, { "epoch": 1.5031166518254675, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.557340383529663, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8728088736534119, "num_tokens": 448688154.0, "step": 11816 }, { "epoch": 1.503243862104058, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6861155033111572, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8674482107162476, "num_tokens": 448726343.0, "step": 11817 }, { "epoch": 1.5033710723826486, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5365753173828125, "learning_rate": 1e-05, "loss": 0.3815, "mean_token_accuracy": 0.8877864480018616, "num_tokens": 448770376.0, "step": 11818 }, { "epoch": 1.503498282661239, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.718454360961914, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8808923959732056, "num_tokens": 448799247.0, "step": 11819 }, { "epoch": 1.5036254929398294, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.535820245742798, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.861213207244873, "num_tokens": 448839379.0, "step": 11820 }, { "epoch": 1.50375270321842, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7479000091552734, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8526537418365479, "num_tokens": 448873924.0, "step": 11821 }, { "epoch": 1.5038799134970104, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5164713859558105, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8673447966575623, "num_tokens": 448915765.0, "step": 11822 }, { "epoch": 1.504007123775601, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.757185697555542, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8780641555786133, "num_tokens": 448953116.0, "step": 11823 }, { "epoch": 1.5041343340541915, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5307419300079346, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8649140000343323, "num_tokens": 448989065.0, "step": 11824 }, { "epoch": 1.504261544332782, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7602343559265137, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8649075627326965, "num_tokens": 449029608.0, "step": 11825 }, { "epoch": 1.5043887546113726, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5336287021636963, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8605935573577881, "num_tokens": 449068288.0, "step": 11826 }, { "epoch": 1.504515964889963, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7709105014801025, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8692283630371094, "num_tokens": 449107754.0, "step": 11827 }, { "epoch": 1.5046431751685536, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5078964233398438, "learning_rate": 1e-05, "loss": 0.3643, "mean_token_accuracy": 0.8934850096702576, "num_tokens": 449148680.0, "step": 11828 }, { "epoch": 1.5047703854471441, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.794567346572876, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8691301345825195, "num_tokens": 449191616.0, "step": 11829 }, { "epoch": 1.5048975957257347, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5357470512390137, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.8792536854743958, "num_tokens": 449227149.0, "step": 11830 }, { "epoch": 1.5050248060043252, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.842017412185669, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8676050901412964, "num_tokens": 449259186.0, "step": 11831 }, { "epoch": 1.5051520162829157, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.544783592224121, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8807253837585449, "num_tokens": 449294403.0, "step": 11832 }, { "epoch": 1.5052792265615063, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.808469295501709, "learning_rate": 1e-05, "loss": 0.3904, "mean_token_accuracy": 0.8892930746078491, "num_tokens": 449334205.0, "step": 11833 }, { "epoch": 1.5054064368400968, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5665640830993652, "learning_rate": 1e-05, "loss": 0.5125, "mean_token_accuracy": 0.853268027305603, "num_tokens": 449372501.0, "step": 11834 }, { "epoch": 1.5055336471186873, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.806913137435913, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8596265912055969, "num_tokens": 449410394.0, "step": 11835 }, { "epoch": 1.5056608573972778, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5567729473114014, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8735195398330688, "num_tokens": 449449494.0, "step": 11836 }, { "epoch": 1.5057880676758684, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8096916675567627, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8663185834884644, "num_tokens": 449485086.0, "step": 11837 }, { "epoch": 1.5059152779544587, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.571838855743408, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8703439831733704, "num_tokens": 449518681.0, "step": 11838 }, { "epoch": 1.5060424882330492, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.773834705352783, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8761752247810364, "num_tokens": 449557439.0, "step": 11839 }, { "epoch": 1.5061696985116397, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5643961429595947, "learning_rate": 1e-05, "loss": 0.4864, "mean_token_accuracy": 0.8555211424827576, "num_tokens": 449600571.0, "step": 11840 }, { "epoch": 1.5062969087902303, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.770176887512207, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8603187799453735, "num_tokens": 449639323.0, "step": 11841 }, { "epoch": 1.5064241190688208, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5587046146392822, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8670032024383545, "num_tokens": 449680082.0, "step": 11842 }, { "epoch": 1.5065513293474113, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.767735004425049, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.856806218624115, "num_tokens": 449717655.0, "step": 11843 }, { "epoch": 1.5066785396260016, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.556767225265503, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.881300687789917, "num_tokens": 449756833.0, "step": 11844 }, { "epoch": 1.5068057499045922, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7610843181610107, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8713165521621704, "num_tokens": 449796525.0, "step": 11845 }, { "epoch": 1.5069329601831827, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5517570972442627, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8691843748092651, "num_tokens": 449837731.0, "step": 11846 }, { "epoch": 1.5070601704617732, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.81581711769104, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8775954842567444, "num_tokens": 449876235.0, "step": 11847 }, { "epoch": 1.5071873807403637, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5492897033691406, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8708818554878235, "num_tokens": 449916153.0, "step": 11848 }, { "epoch": 1.5073145910189543, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.784085512161255, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8563010692596436, "num_tokens": 449957969.0, "step": 11849 }, { "epoch": 1.5074418012975448, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5429515838623047, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8746646046638489, "num_tokens": 449997464.0, "step": 11850 }, { "epoch": 1.5075690115761353, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7821667194366455, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8610193729400635, "num_tokens": 450038102.0, "step": 11851 }, { "epoch": 1.5076962218547258, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.552734851837158, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8781780004501343, "num_tokens": 450073832.0, "step": 11852 }, { "epoch": 1.5078234321333164, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7765328884124756, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8791251182556152, "num_tokens": 450108662.0, "step": 11853 }, { "epoch": 1.507950642411907, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5548019409179688, "learning_rate": 1e-05, "loss": 0.3899, "mean_token_accuracy": 0.885957658290863, "num_tokens": 450144601.0, "step": 11854 }, { "epoch": 1.5080778526904974, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7597029209136963, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8658859729766846, "num_tokens": 450190112.0, "step": 11855 }, { "epoch": 1.508205062969088, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.571272134780884, "learning_rate": 1e-05, "loss": 0.3998, "mean_token_accuracy": 0.8818392157554626, "num_tokens": 450227660.0, "step": 11856 }, { "epoch": 1.5083322732476785, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.742125988006592, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8780894875526428, "num_tokens": 450265243.0, "step": 11857 }, { "epoch": 1.508459483526269, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5817432403564453, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8636724948883057, "num_tokens": 450303382.0, "step": 11858 }, { "epoch": 1.5085866938048595, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7490484714508057, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8735065460205078, "num_tokens": 450335458.0, "step": 11859 }, { "epoch": 1.50871390408345, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5760703086853027, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.867743968963623, "num_tokens": 450373375.0, "step": 11860 }, { "epoch": 1.5088411143620406, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.737051010131836, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8802943229675293, "num_tokens": 450406258.0, "step": 11861 }, { "epoch": 1.508968324640631, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.571094036102295, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8779594898223877, "num_tokens": 450441558.0, "step": 11862 }, { "epoch": 1.5090955349192214, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7192187309265137, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8770557641983032, "num_tokens": 450477685.0, "step": 11863 }, { "epoch": 1.509222745197812, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.564230442047119, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8562582731246948, "num_tokens": 450516661.0, "step": 11864 }, { "epoch": 1.5093499554764025, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.708627223968506, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8725776672363281, "num_tokens": 450555091.0, "step": 11865 }, { "epoch": 1.509477165754993, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5580995082855225, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8679898977279663, "num_tokens": 450596172.0, "step": 11866 }, { "epoch": 1.5096043760335836, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.706394910812378, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8659579157829285, "num_tokens": 450636902.0, "step": 11867 }, { "epoch": 1.509731586312174, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.559439182281494, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.856863796710968, "num_tokens": 450682826.0, "step": 11868 }, { "epoch": 1.5098587965907644, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7080137729644775, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8608023524284363, "num_tokens": 450722995.0, "step": 11869 }, { "epoch": 1.509986006869355, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5474884510040283, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8775414228439331, "num_tokens": 450762109.0, "step": 11870 }, { "epoch": 1.5101132171479454, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.704702377319336, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8732290267944336, "num_tokens": 450799543.0, "step": 11871 }, { "epoch": 1.510240427426536, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5486884117126465, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8673581480979919, "num_tokens": 450840424.0, "step": 11872 }, { "epoch": 1.5103676377051265, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.713644504547119, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.876217782497406, "num_tokens": 450877652.0, "step": 11873 }, { "epoch": 1.510494847983717, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5569145679473877, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8507314324378967, "num_tokens": 450916171.0, "step": 11874 }, { "epoch": 1.5106220582623076, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7232418060302734, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8751485347747803, "num_tokens": 450950455.0, "step": 11875 }, { "epoch": 1.510749268540898, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.540135622024536, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8696803450584412, "num_tokens": 450990747.0, "step": 11876 }, { "epoch": 1.5108764788194886, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.725105047225952, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8764591813087463, "num_tokens": 451029396.0, "step": 11877 }, { "epoch": 1.5110036890980791, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5452096462249756, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8691351413726807, "num_tokens": 451065071.0, "step": 11878 }, { "epoch": 1.5111308993766697, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7323522567749023, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8746909499168396, "num_tokens": 451105487.0, "step": 11879 }, { "epoch": 1.5112581096552602, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5583512783050537, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8610265254974365, "num_tokens": 451141538.0, "step": 11880 }, { "epoch": 1.5113853199338507, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.729485034942627, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8843305110931396, "num_tokens": 451175080.0, "step": 11881 }, { "epoch": 1.5115125302124413, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.542217969894409, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8703084588050842, "num_tokens": 451216117.0, "step": 11882 }, { "epoch": 1.5116397404910318, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7406251430511475, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8739302158355713, "num_tokens": 451251585.0, "step": 11883 }, { "epoch": 1.5117669507696223, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5354464054107666, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8584034442901611, "num_tokens": 451290401.0, "step": 11884 }, { "epoch": 1.5118941610482128, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.746394395828247, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8627921938896179, "num_tokens": 451331896.0, "step": 11885 }, { "epoch": 1.5120213713268034, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.538367509841919, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.874618411064148, "num_tokens": 451368802.0, "step": 11886 }, { "epoch": 1.5121485816053937, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.757249116897583, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8830152750015259, "num_tokens": 451400101.0, "step": 11887 }, { "epoch": 1.5122757918839842, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5324723720550537, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8699824213981628, "num_tokens": 451444732.0, "step": 11888 }, { "epoch": 1.5124030021625747, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.753081798553467, "learning_rate": 1e-05, "loss": 0.3786, "mean_token_accuracy": 0.8924933671951294, "num_tokens": 451481467.0, "step": 11889 }, { "epoch": 1.5125302124411653, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5436253547668457, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8703356981277466, "num_tokens": 451520048.0, "step": 11890 }, { "epoch": 1.5126574227197558, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7596263885498047, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8726012706756592, "num_tokens": 451555121.0, "step": 11891 }, { "epoch": 1.5127846329983463, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.542234420776367, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8696742653846741, "num_tokens": 451594894.0, "step": 11892 }, { "epoch": 1.5129118432769366, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7745563983917236, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8765353560447693, "num_tokens": 451627175.0, "step": 11893 }, { "epoch": 1.5130390535555271, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.545478343963623, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.87811678647995, "num_tokens": 451664177.0, "step": 11894 }, { "epoch": 1.5131662638341177, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7698512077331543, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8781305551528931, "num_tokens": 451696964.0, "step": 11895 }, { "epoch": 1.5132934741127082, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5419397354125977, "learning_rate": 1e-05, "loss": 0.3887, "mean_token_accuracy": 0.8861571550369263, "num_tokens": 451735349.0, "step": 11896 }, { "epoch": 1.5134206843912987, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.776529550552368, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.881474494934082, "num_tokens": 451770410.0, "step": 11897 }, { "epoch": 1.5135478946698893, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.538325071334839, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8767879009246826, "num_tokens": 451812170.0, "step": 11898 }, { "epoch": 1.5136751049484798, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.764613628387451, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8724551200866699, "num_tokens": 451858079.0, "step": 11899 }, { "epoch": 1.5138023152270703, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5539722442626953, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.863366961479187, "num_tokens": 451901120.0, "step": 11900 }, { "epoch": 1.5139295255056608, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7704036235809326, "learning_rate": 1e-05, "loss": 0.5007, "mean_token_accuracy": 0.8580426573753357, "num_tokens": 451944990.0, "step": 11901 }, { "epoch": 1.5140567357842514, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.553403377532959, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8788507580757141, "num_tokens": 451983261.0, "step": 11902 }, { "epoch": 1.514183946062842, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7695226669311523, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8668636083602905, "num_tokens": 452019438.0, "step": 11903 }, { "epoch": 1.5143111563414324, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.571622610092163, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8558863401412964, "num_tokens": 452055941.0, "step": 11904 }, { "epoch": 1.514438366620023, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7468183040618896, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8812729120254517, "num_tokens": 452092625.0, "step": 11905 }, { "epoch": 1.5145655768986135, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.567046880722046, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8667635917663574, "num_tokens": 452126279.0, "step": 11906 }, { "epoch": 1.514692787177204, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7360522747039795, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8762871026992798, "num_tokens": 452163506.0, "step": 11907 }, { "epoch": 1.5148199974557945, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5736124515533447, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8612600564956665, "num_tokens": 452200898.0, "step": 11908 }, { "epoch": 1.514947207734385, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.727213144302368, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8603340983390808, "num_tokens": 452237772.0, "step": 11909 }, { "epoch": 1.5150744180129756, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5744998455047607, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8692006468772888, "num_tokens": 452272741.0, "step": 11910 }, { "epoch": 1.515201628291566, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7176127433776855, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8682417869567871, "num_tokens": 452307674.0, "step": 11911 }, { "epoch": 1.5153288385701564, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.707204818725586, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8611578941345215, "num_tokens": 452352186.0, "step": 11912 }, { "epoch": 1.515456048848747, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.631049156188965, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.878615140914917, "num_tokens": 452390371.0, "step": 11913 }, { "epoch": 1.5155832591273375, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.628450870513916, "learning_rate": 1e-05, "loss": 0.3942, "mean_token_accuracy": 0.8848568797111511, "num_tokens": 452424545.0, "step": 11914 }, { "epoch": 1.515710469405928, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6204400062561035, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.867460310459137, "num_tokens": 452456543.0, "step": 11915 }, { "epoch": 1.5158376796845185, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.619436740875244, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8604781627655029, "num_tokens": 452497884.0, "step": 11916 }, { "epoch": 1.515964889963109, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5975453853607178, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8651295900344849, "num_tokens": 452542596.0, "step": 11917 }, { "epoch": 1.5160921002416994, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6199233531951904, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.878990888595581, "num_tokens": 452581385.0, "step": 11918 }, { "epoch": 1.51621931052029, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.8866937160491943, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8684543371200562, "num_tokens": 452621399.0, "step": 11919 }, { "epoch": 1.5163465207988804, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5399088859558105, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8657943606376648, "num_tokens": 452657782.0, "step": 11920 }, { "epoch": 1.516473731077471, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.73359751701355, "learning_rate": 1e-05, "loss": 0.4017, "mean_token_accuracy": 0.8828632831573486, "num_tokens": 452696239.0, "step": 11921 }, { "epoch": 1.5166009413560615, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.51534104347229, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.863609790802002, "num_tokens": 452734235.0, "step": 11922 }, { "epoch": 1.516728151634652, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.768845558166504, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8768444061279297, "num_tokens": 452769151.0, "step": 11923 }, { "epoch": 1.5168553619132426, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5126724243164062, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8765920400619507, "num_tokens": 452808944.0, "step": 11924 }, { "epoch": 1.516982572191833, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.795771598815918, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8682942986488342, "num_tokens": 452845815.0, "step": 11925 }, { "epoch": 1.5171097824704236, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.517378091812134, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8720382452011108, "num_tokens": 452888010.0, "step": 11926 }, { "epoch": 1.5172369927490141, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8202881813049316, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8811355233192444, "num_tokens": 452924149.0, "step": 11927 }, { "epoch": 1.5173642030276047, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.526994228363037, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8760346174240112, "num_tokens": 452964436.0, "step": 11928 }, { "epoch": 1.5174914133061952, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.83906888961792, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8650239706039429, "num_tokens": 453000455.0, "step": 11929 }, { "epoch": 1.5176186235847857, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.527519464492798, "learning_rate": 1e-05, "loss": 0.3935, "mean_token_accuracy": 0.8854599595069885, "num_tokens": 453040419.0, "step": 11930 }, { "epoch": 1.5177458338633762, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8471908569335938, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8688204288482666, "num_tokens": 453077668.0, "step": 11931 }, { "epoch": 1.5178730441419668, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5489025115966797, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8766630291938782, "num_tokens": 453115423.0, "step": 11932 }, { "epoch": 1.5180002544205573, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8510055541992188, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8689034581184387, "num_tokens": 453150529.0, "step": 11933 }, { "epoch": 1.5181274646991478, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5683603286743164, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8715139031410217, "num_tokens": 453189123.0, "step": 11934 }, { "epoch": 1.5182546749777384, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.821286201477051, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.88475102186203, "num_tokens": 453231381.0, "step": 11935 }, { "epoch": 1.5183818852563287, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5796027183532715, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8594003319740295, "num_tokens": 453272134.0, "step": 11936 }, { "epoch": 1.5185090955349192, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.809800148010254, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8655908107757568, "num_tokens": 453306950.0, "step": 11937 }, { "epoch": 1.5186363058135097, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5724117755889893, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8594468235969543, "num_tokens": 453354206.0, "step": 11938 }, { "epoch": 1.5187635160921003, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8072893619537354, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8671850562095642, "num_tokens": 453388033.0, "step": 11939 }, { "epoch": 1.5188907263706908, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5939061641693115, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8660451769828796, "num_tokens": 453421636.0, "step": 11940 }, { "epoch": 1.5190179366492813, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7611873149871826, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8736941814422607, "num_tokens": 453467994.0, "step": 11941 }, { "epoch": 1.5191451469278716, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.579202175140381, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8568911552429199, "num_tokens": 453508627.0, "step": 11942 }, { "epoch": 1.5192723572064621, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.750180959701538, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.87153160572052, "num_tokens": 453547286.0, "step": 11943 }, { "epoch": 1.5193995674850527, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.582099199295044, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8482060432434082, "num_tokens": 453588485.0, "step": 11944 }, { "epoch": 1.5195267777636432, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.719991683959961, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8668215274810791, "num_tokens": 453633103.0, "step": 11945 }, { "epoch": 1.5196539880422337, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.581631660461426, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8756580948829651, "num_tokens": 453671817.0, "step": 11946 }, { "epoch": 1.5197811983208243, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7226691246032715, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8707864880561829, "num_tokens": 453704704.0, "step": 11947 }, { "epoch": 1.5199084085994148, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5836215019226074, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8644448518753052, "num_tokens": 453743899.0, "step": 11948 }, { "epoch": 1.5200356188780053, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.700976848602295, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8767423033714294, "num_tokens": 453783052.0, "step": 11949 }, { "epoch": 1.5201628291565958, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.568881034851074, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8640548586845398, "num_tokens": 453825522.0, "step": 11950 }, { "epoch": 1.5202900394351864, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6973698139190674, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.876274824142456, "num_tokens": 453859353.0, "step": 11951 }, { "epoch": 1.520417249713777, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.56154203414917, "learning_rate": 1e-05, "loss": 0.3751, "mean_token_accuracy": 0.8894516229629517, "num_tokens": 453896792.0, "step": 11952 }, { "epoch": 1.5205444599923674, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6975347995758057, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8735897541046143, "num_tokens": 453939679.0, "step": 11953 }, { "epoch": 1.520671670270958, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5855183601379395, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.855731725692749, "num_tokens": 453970171.0, "step": 11954 }, { "epoch": 1.5207988805495485, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6918623447418213, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8609975576400757, "num_tokens": 454008470.0, "step": 11955 }, { "epoch": 1.520926090828139, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5731897354125977, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8621623516082764, "num_tokens": 454047711.0, "step": 11956 }, { "epoch": 1.5210533011067295, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6913576126098633, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.868482768535614, "num_tokens": 454083494.0, "step": 11957 }, { "epoch": 1.52118051138532, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5751655101776123, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.874620795249939, "num_tokens": 454115220.0, "step": 11958 }, { "epoch": 1.5213077216639106, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6812009811401367, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.875775933265686, "num_tokens": 454152754.0, "step": 11959 }, { "epoch": 1.521434931942501, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.571078300476074, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8730127811431885, "num_tokens": 454185357.0, "step": 11960 }, { "epoch": 1.5215621422210914, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.688703775405884, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8797968029975891, "num_tokens": 454217277.0, "step": 11961 }, { "epoch": 1.521689352499682, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5641214847564697, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8540461659431458, "num_tokens": 454253252.0, "step": 11962 }, { "epoch": 1.5218165627782725, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6814846992492676, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8807869553565979, "num_tokens": 454292779.0, "step": 11963 }, { "epoch": 1.521943773056863, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.550114870071411, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8706448078155518, "num_tokens": 454336077.0, "step": 11964 }, { "epoch": 1.5220709833354535, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.693554639816284, "learning_rate": 1e-05, "loss": 0.4033, "mean_token_accuracy": 0.882282018661499, "num_tokens": 454369088.0, "step": 11965 }, { "epoch": 1.5221981936140438, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5436973571777344, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.862119197845459, "num_tokens": 454413057.0, "step": 11966 }, { "epoch": 1.5223254038926344, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6921653747558594, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.8844724893569946, "num_tokens": 454447458.0, "step": 11967 }, { "epoch": 1.522452614171225, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5686135292053223, "learning_rate": 1e-05, "loss": 0.4015, "mean_token_accuracy": 0.8837765455245972, "num_tokens": 454479057.0, "step": 11968 }, { "epoch": 1.5225798244498154, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6921963691711426, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8601173162460327, "num_tokens": 454520676.0, "step": 11969 }, { "epoch": 1.522707034728406, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.559069871902466, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.863950252532959, "num_tokens": 454558021.0, "step": 11970 }, { "epoch": 1.5228342450069965, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.691305160522461, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8760018348693848, "num_tokens": 454594701.0, "step": 11971 }, { "epoch": 1.522961455285587, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5542352199554443, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8471345901489258, "num_tokens": 454638746.0, "step": 11972 }, { "epoch": 1.5230886655641775, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6897695064544678, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8755877017974854, "num_tokens": 454676040.0, "step": 11973 }, { "epoch": 1.523215875842768, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5836219787597656, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8695534467697144, "num_tokens": 454713262.0, "step": 11974 }, { "epoch": 1.5233430861213586, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6775500774383545, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8673684597015381, "num_tokens": 454748613.0, "step": 11975 }, { "epoch": 1.5234702963999491, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5631532669067383, "learning_rate": 1e-05, "loss": 0.3817, "mean_token_accuracy": 0.889142632484436, "num_tokens": 454782040.0, "step": 11976 }, { "epoch": 1.5235975066785397, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6689515113830566, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8611689805984497, "num_tokens": 454828394.0, "step": 11977 }, { "epoch": 1.5237247169571302, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.567328929901123, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8618331551551819, "num_tokens": 454866324.0, "step": 11978 }, { "epoch": 1.5238519272357207, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6673316955566406, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.875686764717102, "num_tokens": 454907458.0, "step": 11979 }, { "epoch": 1.5239791375143112, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.55259370803833, "learning_rate": 1e-05, "loss": 0.3906, "mean_token_accuracy": 0.8853553533554077, "num_tokens": 454948159.0, "step": 11980 }, { "epoch": 1.5241063477929018, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.68658185005188, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8777016401290894, "num_tokens": 454984760.0, "step": 11981 }, { "epoch": 1.5242335580714923, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5468671321868896, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8701507449150085, "num_tokens": 455026670.0, "step": 11982 }, { "epoch": 1.5243607683500828, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7304847240448, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8692495822906494, "num_tokens": 455063470.0, "step": 11983 }, { "epoch": 1.5244879786286734, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.53161883354187, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8844287395477295, "num_tokens": 455097787.0, "step": 11984 }, { "epoch": 1.5246151889072637, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7220113277435303, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8861165642738342, "num_tokens": 455137987.0, "step": 11985 }, { "epoch": 1.5247423991858542, "ewc_loss": 0.06396484375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5162193775177, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8552206754684448, "num_tokens": 455184048.0, "step": 11986 }, { "epoch": 1.5248696094644447, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7484211921691895, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8749457597732544, "num_tokens": 455219876.0, "step": 11987 }, { "epoch": 1.5249968197430352, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5318384170532227, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8723835945129395, "num_tokens": 455253491.0, "step": 11988 }, { "epoch": 1.5251240300216258, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.759986162185669, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8668922185897827, "num_tokens": 455293183.0, "step": 11989 }, { "epoch": 1.5252512403002163, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5195682048797607, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8861463069915771, "num_tokens": 455333970.0, "step": 11990 }, { "epoch": 1.5253784505788066, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7794666290283203, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.869320273399353, "num_tokens": 455371547.0, "step": 11991 }, { "epoch": 1.5255056608573971, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5449039936065674, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8707407712936401, "num_tokens": 455403427.0, "step": 11992 }, { "epoch": 1.5256328711359877, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7889978885650635, "learning_rate": 1e-05, "loss": 0.3886, "mean_token_accuracy": 0.8943331837654114, "num_tokens": 455434707.0, "step": 11993 }, { "epoch": 1.5257600814145782, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.525507688522339, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8770480751991272, "num_tokens": 455473255.0, "step": 11994 }, { "epoch": 1.5258872916931687, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.80285906791687, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8724030256271362, "num_tokens": 455504631.0, "step": 11995 }, { "epoch": 1.5260145019717593, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.545875310897827, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8821485042572021, "num_tokens": 455542652.0, "step": 11996 }, { "epoch": 1.5261417122503498, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7949092388153076, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8758595585823059, "num_tokens": 455577912.0, "step": 11997 }, { "epoch": 1.5262689225289403, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5380425453186035, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.876328706741333, "num_tokens": 455617900.0, "step": 11998 }, { "epoch": 1.5263961328075308, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8002495765686035, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8786969184875488, "num_tokens": 455656779.0, "step": 11999 }, { "epoch": 1.5265233430861214, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5474660396575928, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8759583830833435, "num_tokens": 455692620.0, "step": 12000 }, { "epoch": 1.526650553364712, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8010995388031006, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8761966824531555, "num_tokens": 455729805.0, "step": 12001 }, { "epoch": 1.5267777636433024, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.557846784591675, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8770455121994019, "num_tokens": 455767529.0, "step": 12002 }, { "epoch": 1.526904973921893, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7790884971618652, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8710809946060181, "num_tokens": 455807898.0, "step": 12003 }, { "epoch": 1.5270321842004835, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.563472032546997, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8706090450286865, "num_tokens": 455848726.0, "step": 12004 }, { "epoch": 1.527159394479074, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.767805814743042, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.868984580039978, "num_tokens": 455892363.0, "step": 12005 }, { "epoch": 1.5272866047576645, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5698516368865967, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8605412244796753, "num_tokens": 455930104.0, "step": 12006 }, { "epoch": 1.527413815036255, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7652387619018555, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8661358952522278, "num_tokens": 455964272.0, "step": 12007 }, { "epoch": 1.5275410253148456, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5794126987457275, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8682851791381836, "num_tokens": 456003020.0, "step": 12008 }, { "epoch": 1.527668235593436, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7411744594573975, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8716383576393127, "num_tokens": 456040305.0, "step": 12009 }, { "epoch": 1.5277954458720264, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5709292888641357, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8648952841758728, "num_tokens": 456080477.0, "step": 12010 }, { "epoch": 1.527922656150617, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.731132984161377, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.88016676902771, "num_tokens": 456116941.0, "step": 12011 }, { "epoch": 1.5280498664292075, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.566934585571289, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8657519221305847, "num_tokens": 456156184.0, "step": 12012 }, { "epoch": 1.528177076707798, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002613067626953125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7195589542388916, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8578702211380005, "num_tokens": 456200920.0, "step": 12013 }, { "epoch": 1.5283042869863885, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.570117712020874, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8805358409881592, "num_tokens": 456235632.0, "step": 12014 }, { "epoch": 1.5284314972649788, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.730360984802246, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8728148937225342, "num_tokens": 456264898.0, "step": 12015 }, { "epoch": 1.5285587075435694, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5616652965545654, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8681317567825317, "num_tokens": 456304972.0, "step": 12016 }, { "epoch": 1.52868591782216, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7220025062561035, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8695594072341919, "num_tokens": 456344496.0, "step": 12017 }, { "epoch": 1.5288131281007504, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5615651607513428, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8784180879592896, "num_tokens": 456382408.0, "step": 12018 }, { "epoch": 1.528940338379341, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.724010705947876, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8653894662857056, "num_tokens": 456424946.0, "step": 12019 }, { "epoch": 1.5290675486579315, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5424463748931885, "learning_rate": 1e-05, "loss": 0.4121, "mean_token_accuracy": 0.8805122375488281, "num_tokens": 456467432.0, "step": 12020 }, { "epoch": 1.529194758936522, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7355082035064697, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8772203326225281, "num_tokens": 456501557.0, "step": 12021 }, { "epoch": 1.5293219692151125, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5641164779663086, "learning_rate": 1e-05, "loss": 0.5249, "mean_token_accuracy": 0.8466417789459229, "num_tokens": 456540500.0, "step": 12022 }, { "epoch": 1.529449179493703, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.723339557647705, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8777838945388794, "num_tokens": 456574403.0, "step": 12023 }, { "epoch": 1.5295763897722936, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5591042041778564, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8763930201530457, "num_tokens": 456610044.0, "step": 12024 }, { "epoch": 1.5297036000508841, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7092459201812744, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8737161159515381, "num_tokens": 456650289.0, "step": 12025 }, { "epoch": 1.5298308103294747, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.557420492172241, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8677120208740234, "num_tokens": 456687193.0, "step": 12026 }, { "epoch": 1.5299580206080652, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.727867603302002, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8717852234840393, "num_tokens": 456720553.0, "step": 12027 }, { "epoch": 1.5300852308866557, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5679094791412354, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8663060665130615, "num_tokens": 456754736.0, "step": 12028 }, { "epoch": 1.5302124411652462, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7052395343780518, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8857115507125854, "num_tokens": 456791268.0, "step": 12029 }, { "epoch": 1.5303396514438368, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5416080951690674, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8810185790061951, "num_tokens": 456831728.0, "step": 12030 }, { "epoch": 1.5304668617224273, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7142560482025146, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8786801695823669, "num_tokens": 456869375.0, "step": 12031 }, { "epoch": 1.5305940720010178, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.549942970275879, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8633904457092285, "num_tokens": 456909376.0, "step": 12032 }, { "epoch": 1.5307212822796084, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.718139886856079, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8829885721206665, "num_tokens": 456945851.0, "step": 12033 }, { "epoch": 1.5308484925581987, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.573975086212158, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8699150681495667, "num_tokens": 456979833.0, "step": 12034 }, { "epoch": 1.5309757028367892, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.694617748260498, "learning_rate": 1e-05, "loss": 0.3846, "mean_token_accuracy": 0.888870358467102, "num_tokens": 457016032.0, "step": 12035 }, { "epoch": 1.5311029131153797, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.565120220184326, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8869873285293579, "num_tokens": 457050015.0, "step": 12036 }, { "epoch": 1.5312301233939702, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.69374418258667, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.874286413192749, "num_tokens": 457087287.0, "step": 12037 }, { "epoch": 1.5313573336725608, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5629208087921143, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8696164488792419, "num_tokens": 457123237.0, "step": 12038 }, { "epoch": 1.5314845439511513, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6913256645202637, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8691853284835815, "num_tokens": 457161834.0, "step": 12039 }, { "epoch": 1.5316117542297416, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5562827587127686, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.862488329410553, "num_tokens": 457203672.0, "step": 12040 }, { "epoch": 1.5317389645083321, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.693368911743164, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8828855752944946, "num_tokens": 457238486.0, "step": 12041 }, { "epoch": 1.5318661747869227, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.572233200073242, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8578060865402222, "num_tokens": 457270817.0, "step": 12042 }, { "epoch": 1.5319933850655132, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.692521810531616, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8724926114082336, "num_tokens": 457306812.0, "step": 12043 }, { "epoch": 1.5321205953441037, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.557476758956909, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8804599046707153, "num_tokens": 457340893.0, "step": 12044 }, { "epoch": 1.5322478056226942, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.690573215484619, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8746235370635986, "num_tokens": 457381668.0, "step": 12045 }, { "epoch": 1.5323750159012848, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5614118576049805, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.860412061214447, "num_tokens": 457417498.0, "step": 12046 }, { "epoch": 1.5325022261798753, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.683389902114868, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8724165558815002, "num_tokens": 457460300.0, "step": 12047 }, { "epoch": 1.5326294364584658, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.556086540222168, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8626589775085449, "num_tokens": 457502030.0, "step": 12048 }, { "epoch": 1.5327566467370564, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.685253620147705, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8805567026138306, "num_tokens": 457540170.0, "step": 12049 }, { "epoch": 1.532883857015647, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.542316198348999, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.878241777420044, "num_tokens": 457585834.0, "step": 12050 }, { "epoch": 1.5330110672942374, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6966419219970703, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8824693560600281, "num_tokens": 457622027.0, "step": 12051 }, { "epoch": 1.533138277572828, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.549694538116455, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8647259473800659, "num_tokens": 457659129.0, "step": 12052 }, { "epoch": 1.5332654878514185, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6977336406707764, "learning_rate": 1e-05, "loss": 0.3869, "mean_token_accuracy": 0.8878681063652039, "num_tokens": 457695046.0, "step": 12053 }, { "epoch": 1.533392698130009, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.542109727859497, "learning_rate": 1e-05, "loss": 0.4121, "mean_token_accuracy": 0.8779915571212769, "num_tokens": 457735650.0, "step": 12054 }, { "epoch": 1.5335199084085995, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7127959728240967, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8803646564483643, "num_tokens": 457767395.0, "step": 12055 }, { "epoch": 1.53364711868719, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5437841415405273, "learning_rate": 1e-05, "loss": 0.3799, "mean_token_accuracy": 0.885454535484314, "num_tokens": 457800100.0, "step": 12056 }, { "epoch": 1.5337743289657806, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7247090339660645, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8759530782699585, "num_tokens": 457837497.0, "step": 12057 }, { "epoch": 1.533901539244371, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5325303077697754, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8749935030937195, "num_tokens": 457875711.0, "step": 12058 }, { "epoch": 1.5340287495229614, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.732707977294922, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8596059083938599, "num_tokens": 457914888.0, "step": 12059 }, { "epoch": 1.534155959801552, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5367233753204346, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8568947315216064, "num_tokens": 457953315.0, "step": 12060 }, { "epoch": 1.5342831700801425, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7463290691375732, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8591295480728149, "num_tokens": 457990857.0, "step": 12061 }, { "epoch": 1.534410380358733, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5244171619415283, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8657985925674438, "num_tokens": 458034060.0, "step": 12062 }, { "epoch": 1.5345375906373235, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.76776123046875, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8654683828353882, "num_tokens": 458074093.0, "step": 12063 }, { "epoch": 1.5346648009159138, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.535923719406128, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.879345178604126, "num_tokens": 458107374.0, "step": 12064 }, { "epoch": 1.5347920111945044, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7673563957214355, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8747437596321106, "num_tokens": 458141873.0, "step": 12065 }, { "epoch": 1.534919221473095, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5334208011627197, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8728182315826416, "num_tokens": 458181483.0, "step": 12066 }, { "epoch": 1.5350464317516854, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.778914451599121, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8645548820495605, "num_tokens": 458223644.0, "step": 12067 }, { "epoch": 1.535173642030276, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.537489175796509, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8734191656112671, "num_tokens": 458263518.0, "step": 12068 }, { "epoch": 1.5353008523088665, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7924411296844482, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8786468505859375, "num_tokens": 458296621.0, "step": 12069 }, { "epoch": 1.535428062587457, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.542551040649414, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.866089940071106, "num_tokens": 458335161.0, "step": 12070 }, { "epoch": 1.5355552728660475, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.794647693634033, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8616011142730713, "num_tokens": 458373758.0, "step": 12071 }, { "epoch": 1.535682483144638, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5452141761779785, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8762203454971313, "num_tokens": 458412662.0, "step": 12072 }, { "epoch": 1.5358096934232286, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.791438341140747, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8633811473846436, "num_tokens": 458455314.0, "step": 12073 }, { "epoch": 1.5359369037018191, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.552443504333496, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8522101640701294, "num_tokens": 458498633.0, "step": 12074 }, { "epoch": 1.5360641139804097, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7922682762145996, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8690609931945801, "num_tokens": 458538920.0, "step": 12075 }, { "epoch": 1.5361913242590002, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5571486949920654, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8704646825790405, "num_tokens": 458574333.0, "step": 12076 }, { "epoch": 1.5363185345375907, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7876250743865967, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8640313148498535, "num_tokens": 458612131.0, "step": 12077 }, { "epoch": 1.5364457448161812, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5644781589508057, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8689707517623901, "num_tokens": 458652169.0, "step": 12078 }, { "epoch": 1.5365729550947718, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.79826021194458, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8649803400039673, "num_tokens": 458687816.0, "step": 12079 }, { "epoch": 1.5367001653733623, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5505473613739014, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.8778456449508667, "num_tokens": 458723600.0, "step": 12080 }, { "epoch": 1.5368273756519528, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7933576107025146, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8584572076797485, "num_tokens": 458761863.0, "step": 12081 }, { "epoch": 1.5369545859305433, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5798850059509277, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.85926353931427, "num_tokens": 458794986.0, "step": 12082 }, { "epoch": 1.5370817962091337, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.773118495941162, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8623937964439392, "num_tokens": 458833939.0, "step": 12083 }, { "epoch": 1.5372090064877242, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5716516971588135, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8608156442642212, "num_tokens": 458874522.0, "step": 12084 }, { "epoch": 1.5373362167663147, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.764143466949463, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8500031232833862, "num_tokens": 458911794.0, "step": 12085 }, { "epoch": 1.5374634270449052, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5645453929901123, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8698238730430603, "num_tokens": 458951979.0, "step": 12086 }, { "epoch": 1.5375906373234958, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.750476837158203, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8620734810829163, "num_tokens": 458993657.0, "step": 12087 }, { "epoch": 1.5377178476020863, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.558311939239502, "learning_rate": 1e-05, "loss": 0.3923, "mean_token_accuracy": 0.8835665583610535, "num_tokens": 459030216.0, "step": 12088 }, { "epoch": 1.5378450578806766, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.764925003051758, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8604456782341003, "num_tokens": 459066451.0, "step": 12089 }, { "epoch": 1.5379722681592671, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5559446811676025, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8803222179412842, "num_tokens": 459106928.0, "step": 12090 }, { "epoch": 1.5380994784378577, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7685766220092773, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8600897192955017, "num_tokens": 459142769.0, "step": 12091 }, { "epoch": 1.5382266887164482, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5715479850769043, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8752154111862183, "num_tokens": 459178895.0, "step": 12092 }, { "epoch": 1.5383538989950387, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000263214111328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7436459064483643, "learning_rate": 1e-05, "loss": 0.3924, "mean_token_accuracy": 0.884995698928833, "num_tokens": 459214973.0, "step": 12093 }, { "epoch": 1.5384811092736292, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.56559157371521, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8693886995315552, "num_tokens": 459251279.0, "step": 12094 }, { "epoch": 1.5386083195522198, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7506959438323975, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8520176410675049, "num_tokens": 459285587.0, "step": 12095 }, { "epoch": 1.5387355298308103, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5653445720672607, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8590346574783325, "num_tokens": 459326103.0, "step": 12096 }, { "epoch": 1.5388627401094008, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7387216091156006, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8698304891586304, "num_tokens": 459367075.0, "step": 12097 }, { "epoch": 1.5389899503879914, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.564655303955078, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8627068400382996, "num_tokens": 459408164.0, "step": 12098 }, { "epoch": 1.5391171606665819, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.738226890563965, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8616648316383362, "num_tokens": 459443081.0, "step": 12099 }, { "epoch": 1.5392443709451724, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5615687370300293, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8659564256668091, "num_tokens": 459481469.0, "step": 12100 }, { "epoch": 1.539371581223763, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7241218090057373, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8695421814918518, "num_tokens": 459519319.0, "step": 12101 }, { "epoch": 1.5394987915023535, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5606043338775635, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8708154559135437, "num_tokens": 459559998.0, "step": 12102 }, { "epoch": 1.539626001780944, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7163126468658447, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8846363425254822, "num_tokens": 459596765.0, "step": 12103 }, { "epoch": 1.5397532120595345, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5690155029296875, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.8526641130447388, "num_tokens": 459634023.0, "step": 12104 }, { "epoch": 1.539880422338125, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7037620544433594, "learning_rate": 1e-05, "loss": 0.3964, "mean_token_accuracy": 0.8850480318069458, "num_tokens": 459675355.0, "step": 12105 }, { "epoch": 1.5400076326167156, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5706889629364014, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.859618067741394, "num_tokens": 459716788.0, "step": 12106 }, { "epoch": 1.5401348428953059, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7042319774627686, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8760980367660522, "num_tokens": 459752580.0, "step": 12107 }, { "epoch": 1.5402620531738964, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5698230266571045, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8759656548500061, "num_tokens": 459788594.0, "step": 12108 }, { "epoch": 1.540389263452487, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.701686382293701, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.883031964302063, "num_tokens": 459826806.0, "step": 12109 }, { "epoch": 1.5405164737310775, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.548222780227661, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8770092725753784, "num_tokens": 459870157.0, "step": 12110 }, { "epoch": 1.540643684009668, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7104218006134033, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8763854503631592, "num_tokens": 459909081.0, "step": 12111 }, { "epoch": 1.5407708942882585, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5638747215270996, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8781706094741821, "num_tokens": 459942886.0, "step": 12112 }, { "epoch": 1.5408981045668488, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7146694660186768, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8710833787918091, "num_tokens": 459979351.0, "step": 12113 }, { "epoch": 1.5410253148454394, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5436129570007324, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8677962422370911, "num_tokens": 460026058.0, "step": 12114 }, { "epoch": 1.54115252512403, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7168400287628174, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8674193620681763, "num_tokens": 460066414.0, "step": 12115 }, { "epoch": 1.5412797354026204, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.54402232170105, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8661956191062927, "num_tokens": 460107747.0, "step": 12116 }, { "epoch": 1.541406945681211, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.731342077255249, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8748786449432373, "num_tokens": 460144418.0, "step": 12117 }, { "epoch": 1.5415341559598015, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.540452241897583, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8615090847015381, "num_tokens": 460186061.0, "step": 12118 }, { "epoch": 1.541661366238392, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.739438533782959, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8654861450195312, "num_tokens": 460222672.0, "step": 12119 }, { "epoch": 1.5417885765169825, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5285301208496094, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8793339729309082, "num_tokens": 460265992.0, "step": 12120 }, { "epoch": 1.541915786795573, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.751333713531494, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8727951049804688, "num_tokens": 460305498.0, "step": 12121 }, { "epoch": 1.5420429970741636, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.556405544281006, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8658878207206726, "num_tokens": 460336013.0, "step": 12122 }, { "epoch": 1.5421702073527541, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7425663471221924, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8720870018005371, "num_tokens": 460373452.0, "step": 12123 }, { "epoch": 1.5422974176313446, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5425446033477783, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8722941875457764, "num_tokens": 460414059.0, "step": 12124 }, { "epoch": 1.5424246279099352, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.751037836074829, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.8846454620361328, "num_tokens": 460446098.0, "step": 12125 }, { "epoch": 1.5425518381885257, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.537278413772583, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8718301653862, "num_tokens": 460485384.0, "step": 12126 }, { "epoch": 1.5426790484671162, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.781031370162964, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8645723462104797, "num_tokens": 460517200.0, "step": 12127 }, { "epoch": 1.5428062587457068, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5515975952148438, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8794506192207336, "num_tokens": 460547557.0, "step": 12128 }, { "epoch": 1.5429334690242973, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.760946750640869, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8791882991790771, "num_tokens": 460583189.0, "step": 12129 }, { "epoch": 1.5430606793028878, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.551042079925537, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8750276565551758, "num_tokens": 460623619.0, "step": 12130 }, { "epoch": 1.5431878895814783, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7670576572418213, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8669947385787964, "num_tokens": 460666667.0, "step": 12131 }, { "epoch": 1.5433150998600687, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.557096242904663, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8709831237792969, "num_tokens": 460704292.0, "step": 12132 }, { "epoch": 1.5434423101386592, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7678818702697754, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8745219111442566, "num_tokens": 460748153.0, "step": 12133 }, { "epoch": 1.5435695204172497, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.542998790740967, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8763470649719238, "num_tokens": 460787663.0, "step": 12134 }, { "epoch": 1.5436967306958402, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7729053497314453, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8790173530578613, "num_tokens": 460821643.0, "step": 12135 }, { "epoch": 1.5438239409744308, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.569563627243042, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8628396391868591, "num_tokens": 460857180.0, "step": 12136 }, { "epoch": 1.5439511512530213, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.775529384613037, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8682631254196167, "num_tokens": 460891573.0, "step": 12137 }, { "epoch": 1.5440783615316116, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5504348278045654, "learning_rate": 1e-05, "loss": 0.3918, "mean_token_accuracy": 0.8840137720108032, "num_tokens": 460927727.0, "step": 12138 }, { "epoch": 1.5442055718102021, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7746522426605225, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8619693517684937, "num_tokens": 460965531.0, "step": 12139 }, { "epoch": 1.5443327820887927, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5546443462371826, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8710126876831055, "num_tokens": 461002900.0, "step": 12140 }, { "epoch": 1.5444599923673832, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7631921768188477, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8723609447479248, "num_tokens": 461041569.0, "step": 12141 }, { "epoch": 1.5445872026459737, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.561873435974121, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.861729621887207, "num_tokens": 461080453.0, "step": 12142 }, { "epoch": 1.5447144129245642, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7551991939544678, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8675995469093323, "num_tokens": 461120963.0, "step": 12143 }, { "epoch": 1.5448416232031548, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5661256313323975, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8751192092895508, "num_tokens": 461156441.0, "step": 12144 }, { "epoch": 1.5449688334817453, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7511253356933594, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8815171718597412, "num_tokens": 461189058.0, "step": 12145 }, { "epoch": 1.5450960437603358, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5632948875427246, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8788917660713196, "num_tokens": 461223962.0, "step": 12146 }, { "epoch": 1.5452232540389264, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.73270845413208, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8798038959503174, "num_tokens": 461262114.0, "step": 12147 }, { "epoch": 1.5453504643175169, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5723373889923096, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8731027841567993, "num_tokens": 461294633.0, "step": 12148 }, { "epoch": 1.5454776745961074, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7212917804718018, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8669044375419617, "num_tokens": 461338549.0, "step": 12149 }, { "epoch": 1.545604884874698, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5833497047424316, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8678171634674072, "num_tokens": 461372612.0, "step": 12150 }, { "epoch": 1.5457320951532885, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7301502227783203, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8609458804130554, "num_tokens": 461404060.0, "step": 12151 }, { "epoch": 1.545859305431879, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.57887601852417, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8692139983177185, "num_tokens": 461436728.0, "step": 12152 }, { "epoch": 1.5459865157104695, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7023539543151855, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8817529678344727, "num_tokens": 461475418.0, "step": 12153 }, { "epoch": 1.54611372598906, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.57313871383667, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8621024489402771, "num_tokens": 461512094.0, "step": 12154 }, { "epoch": 1.5462409362676506, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.707036256790161, "learning_rate": 1e-05, "loss": 0.4982, "mean_token_accuracy": 0.8528767824172974, "num_tokens": 461553454.0, "step": 12155 }, { "epoch": 1.5463681465462409, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.589890956878662, "learning_rate": 1e-05, "loss": 0.5229, "mean_token_accuracy": 0.8501943349838257, "num_tokens": 461590818.0, "step": 12156 }, { "epoch": 1.5464953568248314, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.662780284881592, "learning_rate": 1e-05, "loss": 0.3972, "mean_token_accuracy": 0.8840241432189941, "num_tokens": 461638692.0, "step": 12157 }, { "epoch": 1.546622567103422, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.576322317123413, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8726595640182495, "num_tokens": 461681040.0, "step": 12158 }, { "epoch": 1.5467497773820125, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6784839630126953, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8709042072296143, "num_tokens": 461716999.0, "step": 12159 }, { "epoch": 1.546876987660603, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.575984477996826, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8636659979820251, "num_tokens": 461755053.0, "step": 12160 }, { "epoch": 1.5470041979391935, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6647701263427734, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8761172890663147, "num_tokens": 461795384.0, "step": 12161 }, { "epoch": 1.5471314082177838, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5576059818267822, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8736851811408997, "num_tokens": 461843951.0, "step": 12162 }, { "epoch": 1.5472586184963744, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.685789108276367, "learning_rate": 1e-05, "loss": 0.4911, "mean_token_accuracy": 0.8551943302154541, "num_tokens": 461884172.0, "step": 12163 }, { "epoch": 1.5473858287749649, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.553623676300049, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8688420057296753, "num_tokens": 461928493.0, "step": 12164 }, { "epoch": 1.5475130390535554, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6989755630493164, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.875095009803772, "num_tokens": 461964462.0, "step": 12165 }, { "epoch": 1.547640249332146, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.548367738723755, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8772690892219543, "num_tokens": 462004123.0, "step": 12166 }, { "epoch": 1.5477674596107365, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7128889560699463, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.852202296257019, "num_tokens": 462042906.0, "step": 12167 }, { "epoch": 1.547894669889327, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5393457412719727, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8818149566650391, "num_tokens": 462083802.0, "step": 12168 }, { "epoch": 1.5480218801679175, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.724839448928833, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8661870360374451, "num_tokens": 462119861.0, "step": 12169 }, { "epoch": 1.548149090446508, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5347228050231934, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8644814491271973, "num_tokens": 462162445.0, "step": 12170 }, { "epoch": 1.5482763007250986, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7509663105010986, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8696302771568298, "num_tokens": 462195176.0, "step": 12171 }, { "epoch": 1.5484035110036891, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.528219699859619, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8642467260360718, "num_tokens": 462236155.0, "step": 12172 }, { "epoch": 1.5485307212822796, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7841622829437256, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8716446161270142, "num_tokens": 462267099.0, "step": 12173 }, { "epoch": 1.5486579315608702, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.519716262817383, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8582512140274048, "num_tokens": 462311485.0, "step": 12174 }, { "epoch": 1.5487851418394607, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7897026538848877, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8713060617446899, "num_tokens": 462349078.0, "step": 12175 }, { "epoch": 1.5489123521180512, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5420665740966797, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8693081140518188, "num_tokens": 462381477.0, "step": 12176 }, { "epoch": 1.5490395623966418, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8017518520355225, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8707304000854492, "num_tokens": 462415996.0, "step": 12177 }, { "epoch": 1.5491667726752323, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5506677627563477, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8751646280288696, "num_tokens": 462446737.0, "step": 12178 }, { "epoch": 1.5492939829538228, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.786541700363159, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8652622699737549, "num_tokens": 462487742.0, "step": 12179 }, { "epoch": 1.5494211932324133, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.537271499633789, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8745870590209961, "num_tokens": 462529231.0, "step": 12180 }, { "epoch": 1.5495484035110036, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.79992938041687, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8735735416412354, "num_tokens": 462565608.0, "step": 12181 }, { "epoch": 1.5496756137895942, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5526490211486816, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.873400092124939, "num_tokens": 462600919.0, "step": 12182 }, { "epoch": 1.5498028240681847, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.798184633255005, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.867667555809021, "num_tokens": 462635594.0, "step": 12183 }, { "epoch": 1.5499300343467752, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5623538494110107, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.857970654964447, "num_tokens": 462672349.0, "step": 12184 }, { "epoch": 1.5500572446253658, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.801217555999756, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8754647970199585, "num_tokens": 462710286.0, "step": 12185 }, { "epoch": 1.5501844549039563, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.573307991027832, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8693377375602722, "num_tokens": 462744721.0, "step": 12186 }, { "epoch": 1.5503116651825466, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7693676948547363, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8843710422515869, "num_tokens": 462782812.0, "step": 12187 }, { "epoch": 1.5504388754611371, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.549414873123169, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8792588114738464, "num_tokens": 462827919.0, "step": 12188 }, { "epoch": 1.5505660857397277, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.784555673599243, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8662233352661133, "num_tokens": 462864878.0, "step": 12189 }, { "epoch": 1.5506932960183182, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5520365238189697, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8788257837295532, "num_tokens": 462906213.0, "step": 12190 }, { "epoch": 1.5508205062969087, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.802067518234253, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.8506863117218018, "num_tokens": 462943078.0, "step": 12191 }, { "epoch": 1.5509477165754992, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5511529445648193, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.880394458770752, "num_tokens": 462982444.0, "step": 12192 }, { "epoch": 1.5510749268540898, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7907490730285645, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8801919221878052, "num_tokens": 463024335.0, "step": 12193 }, { "epoch": 1.5512021371326803, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5603270530700684, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8726508617401123, "num_tokens": 463062278.0, "step": 12194 }, { "epoch": 1.5513293474112708, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7909321784973145, "learning_rate": 1e-05, "loss": 0.5075, "mean_token_accuracy": 0.8499815464019775, "num_tokens": 463100257.0, "step": 12195 }, { "epoch": 1.5514565576898613, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5693771839141846, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.874754786491394, "num_tokens": 463139468.0, "step": 12196 }, { "epoch": 1.5515837679684519, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7766213417053223, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8662889003753662, "num_tokens": 463176145.0, "step": 12197 }, { "epoch": 1.5517109782470424, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.576807975769043, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8793066740036011, "num_tokens": 463208899.0, "step": 12198 }, { "epoch": 1.551838188525633, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.755255937576294, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8743121027946472, "num_tokens": 463250163.0, "step": 12199 }, { "epoch": 1.5519653988042235, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5769758224487305, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8674409985542297, "num_tokens": 463285447.0, "step": 12200 }, { "epoch": 1.552092609082814, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7383933067321777, "learning_rate": 1e-05, "loss": 0.4004, "mean_token_accuracy": 0.8867911100387573, "num_tokens": 463322601.0, "step": 12201 }, { "epoch": 1.5522198193614045, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5647459030151367, "learning_rate": 1e-05, "loss": 0.4121, "mean_token_accuracy": 0.8811764121055603, "num_tokens": 463360624.0, "step": 12202 }, { "epoch": 1.552347029639995, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7254836559295654, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8731468915939331, "num_tokens": 463401662.0, "step": 12203 }, { "epoch": 1.5524742399185856, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5803403854370117, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8686491250991821, "num_tokens": 463433999.0, "step": 12204 }, { "epoch": 1.5526014501971759, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7256736755371094, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8533992767333984, "num_tokens": 463473673.0, "step": 12205 }, { "epoch": 1.5527286604757664, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.588649272918701, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.858901858329773, "num_tokens": 463507197.0, "step": 12206 }, { "epoch": 1.552855870754357, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.716176986694336, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8714710474014282, "num_tokens": 463540777.0, "step": 12207 }, { "epoch": 1.5529830810329475, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.570448398590088, "learning_rate": 1e-05, "loss": 0.3736, "mean_token_accuracy": 0.8891686201095581, "num_tokens": 463573335.0, "step": 12208 }, { "epoch": 1.553110291311538, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6994893550872803, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8749294281005859, "num_tokens": 463613898.0, "step": 12209 }, { "epoch": 1.5532375015901285, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.562751293182373, "learning_rate": 1e-05, "loss": 0.3862, "mean_token_accuracy": 0.8846303224563599, "num_tokens": 463648948.0, "step": 12210 }, { "epoch": 1.5533647118687188, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7074472904205322, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8794200420379639, "num_tokens": 463683417.0, "step": 12211 }, { "epoch": 1.5534919221473094, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.57881236076355, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8633312582969666, "num_tokens": 463719533.0, "step": 12212 }, { "epoch": 1.5536191324258999, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.690047264099121, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.8818023204803467, "num_tokens": 463757740.0, "step": 12213 }, { "epoch": 1.5537463427044904, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5686111450195312, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.871148943901062, "num_tokens": 463795619.0, "step": 12214 }, { "epoch": 1.553873552983081, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.700315237045288, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8734878301620483, "num_tokens": 463829934.0, "step": 12215 }, { "epoch": 1.5540007632616715, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.57332706451416, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8493179082870483, "num_tokens": 463869584.0, "step": 12216 }, { "epoch": 1.554127973540262, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6884074211120605, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8568059802055359, "num_tokens": 463909220.0, "step": 12217 }, { "epoch": 1.5542551838188525, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.551953077316284, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.878185510635376, "num_tokens": 463949367.0, "step": 12218 }, { "epoch": 1.554382394097443, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.710908889770508, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.862589955329895, "num_tokens": 463986037.0, "step": 12219 }, { "epoch": 1.5545096043760336, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.553081512451172, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8804113864898682, "num_tokens": 464021903.0, "step": 12220 }, { "epoch": 1.554636814654624, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.709538698196411, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8772749304771423, "num_tokens": 464058738.0, "step": 12221 }, { "epoch": 1.5547640249332146, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.564846992492676, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8527113795280457, "num_tokens": 464090791.0, "step": 12222 }, { "epoch": 1.5548912352118052, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7069451808929443, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8702933192253113, "num_tokens": 464133897.0, "step": 12223 }, { "epoch": 1.5550184454903957, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5384302139282227, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8786662817001343, "num_tokens": 464173808.0, "step": 12224 }, { "epoch": 1.5551456557689862, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.722033977508545, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.8744127750396729, "num_tokens": 464209456.0, "step": 12225 }, { "epoch": 1.5552728660475768, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.543015480041504, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8625049591064453, "num_tokens": 464250192.0, "step": 12226 }, { "epoch": 1.5554000763261673, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.734811544418335, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8531591892242432, "num_tokens": 464291485.0, "step": 12227 }, { "epoch": 1.5555272866047578, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5448150634765625, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8633072972297668, "num_tokens": 464329963.0, "step": 12228 }, { "epoch": 1.5556544968833483, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.741060972213745, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8733195066452026, "num_tokens": 464366944.0, "step": 12229 }, { "epoch": 1.5557817071619386, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.539323091506958, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8643159866333008, "num_tokens": 464409070.0, "step": 12230 }, { "epoch": 1.5559089174405292, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7526121139526367, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.877554178237915, "num_tokens": 464448305.0, "step": 12231 }, { "epoch": 1.5560361277191197, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.560056209564209, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8531493544578552, "num_tokens": 464482453.0, "step": 12232 }, { "epoch": 1.5561633379977102, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.754176616668701, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8600492477416992, "num_tokens": 464518692.0, "step": 12233 }, { "epoch": 1.5562905482763008, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.545551300048828, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8647934198379517, "num_tokens": 464554502.0, "step": 12234 }, { "epoch": 1.5564177585548913, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7517871856689453, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8785147666931152, "num_tokens": 464591414.0, "step": 12235 }, { "epoch": 1.5565449688334816, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.557840585708618, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8657610416412354, "num_tokens": 464627116.0, "step": 12236 }, { "epoch": 1.5566721791120721, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7601373195648193, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8710733652114868, "num_tokens": 464661783.0, "step": 12237 }, { "epoch": 1.5567993893906626, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.567018747329712, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8604862093925476, "num_tokens": 464704851.0, "step": 12238 }, { "epoch": 1.5569265996692532, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7478537559509277, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.861437976360321, "num_tokens": 464746583.0, "step": 12239 }, { "epoch": 1.5570538099478437, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.554522752761841, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8656049966812134, "num_tokens": 464783275.0, "step": 12240 }, { "epoch": 1.5571810202264342, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7455599308013916, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8764458894729614, "num_tokens": 464818436.0, "step": 12241 }, { "epoch": 1.5573082305050248, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5465993881225586, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8690359592437744, "num_tokens": 464859708.0, "step": 12242 }, { "epoch": 1.5574354407836153, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7549221515655518, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8743678331375122, "num_tokens": 464894339.0, "step": 12243 }, { "epoch": 1.5575626510622058, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.549361228942871, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8780670166015625, "num_tokens": 464931517.0, "step": 12244 }, { "epoch": 1.5576898613407963, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7573611736297607, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8681926131248474, "num_tokens": 464971616.0, "step": 12245 }, { "epoch": 1.5578170716193869, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5448834896087646, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8789010643959045, "num_tokens": 465012551.0, "step": 12246 }, { "epoch": 1.5579442818979774, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7579803466796875, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8811277151107788, "num_tokens": 465054266.0, "step": 12247 }, { "epoch": 1.558071492176568, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.553389072418213, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8765302896499634, "num_tokens": 465090057.0, "step": 12248 }, { "epoch": 1.5581987024551585, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.746598720550537, "learning_rate": 1e-05, "loss": 0.3869, "mean_token_accuracy": 0.8877601027488708, "num_tokens": 465132169.0, "step": 12249 }, { "epoch": 1.558325912733749, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5636439323425293, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8691694736480713, "num_tokens": 465169922.0, "step": 12250 }, { "epoch": 1.5584531230123395, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7462613582611084, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8850756883621216, "num_tokens": 465207792.0, "step": 12251 }, { "epoch": 1.55858033329093, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5551605224609375, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8622301816940308, "num_tokens": 465250498.0, "step": 12252 }, { "epoch": 1.5587075435695206, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7634050846099854, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8654259443283081, "num_tokens": 465288321.0, "step": 12253 }, { "epoch": 1.5588347538481109, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5455322265625, "learning_rate": 1e-05, "loss": 0.3505, "mean_token_accuracy": 0.8965251445770264, "num_tokens": 465322241.0, "step": 12254 }, { "epoch": 1.5589619641267014, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7561371326446533, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8849199414253235, "num_tokens": 465359680.0, "step": 12255 }, { "epoch": 1.559089174405292, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5519909858703613, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8784184455871582, "num_tokens": 465396951.0, "step": 12256 }, { "epoch": 1.5592163846838825, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7688376903533936, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8686782717704773, "num_tokens": 465433087.0, "step": 12257 }, { "epoch": 1.559343594962473, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5530471801757812, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8768998980522156, "num_tokens": 465472810.0, "step": 12258 }, { "epoch": 1.5594708052410635, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.763385772705078, "learning_rate": 1e-05, "loss": 0.4767, "mean_token_accuracy": 0.8671477437019348, "num_tokens": 465512714.0, "step": 12259 }, { "epoch": 1.5595980155196538, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5632617473602295, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8566842079162598, "num_tokens": 465547629.0, "step": 12260 }, { "epoch": 1.5597252257982444, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.765573740005493, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8660728335380554, "num_tokens": 465584226.0, "step": 12261 }, { "epoch": 1.5598524360768349, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5469255447387695, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8585934638977051, "num_tokens": 465625854.0, "step": 12262 }, { "epoch": 1.5599796463554254, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7693636417388916, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8530156016349792, "num_tokens": 465660670.0, "step": 12263 }, { "epoch": 1.560106856634016, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.54904842376709, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8701786398887634, "num_tokens": 465700359.0, "step": 12264 }, { "epoch": 1.5602340669126065, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.8011372089385986, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8744187951087952, "num_tokens": 465733648.0, "step": 12265 }, { "epoch": 1.560361277191197, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.552741289138794, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8638693690299988, "num_tokens": 465769814.0, "step": 12266 }, { "epoch": 1.5604884874697875, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.784425973892212, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8777521252632141, "num_tokens": 465804637.0, "step": 12267 }, { "epoch": 1.560615697748378, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5566751956939697, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.884208083152771, "num_tokens": 465838649.0, "step": 12268 }, { "epoch": 1.5607429080269686, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.780137777328491, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8838828802108765, "num_tokens": 465870628.0, "step": 12269 }, { "epoch": 1.560870118305559, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.566466808319092, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8655670285224915, "num_tokens": 465905612.0, "step": 12270 }, { "epoch": 1.5609973285841496, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.760390043258667, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8693605661392212, "num_tokens": 465945964.0, "step": 12271 }, { "epoch": 1.5611245388627402, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5671205520629883, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8845196962356567, "num_tokens": 465979600.0, "step": 12272 }, { "epoch": 1.5612517491413307, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7523977756500244, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8654891848564148, "num_tokens": 466017406.0, "step": 12273 }, { "epoch": 1.5613789594199212, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.561833620071411, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8675883412361145, "num_tokens": 466059178.0, "step": 12274 }, { "epoch": 1.5615061696985117, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7513482570648193, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8675536513328552, "num_tokens": 466096385.0, "step": 12275 }, { "epoch": 1.5616333799771023, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5659029483795166, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8715454936027527, "num_tokens": 466131160.0, "step": 12276 }, { "epoch": 1.5617605902556928, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.736750364303589, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8730555772781372, "num_tokens": 466174842.0, "step": 12277 }, { "epoch": 1.5618878005342833, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.560049295425415, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8769692182540894, "num_tokens": 466212943.0, "step": 12278 }, { "epoch": 1.5620150108128736, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.727646827697754, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.885898768901825, "num_tokens": 466251560.0, "step": 12279 }, { "epoch": 1.5621422210914642, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.574876070022583, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8599924445152283, "num_tokens": 466287150.0, "step": 12280 }, { "epoch": 1.5622694313700547, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.719424247741699, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8699501156806946, "num_tokens": 466328508.0, "step": 12281 }, { "epoch": 1.5623966416486452, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.57719087600708, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8674260377883911, "num_tokens": 466362552.0, "step": 12282 }, { "epoch": 1.5625238519272358, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.780937433242798, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8667574524879456, "num_tokens": 466404253.0, "step": 12283 }, { "epoch": 1.5626510622058263, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.556673765182495, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8620485663414001, "num_tokens": 466443258.0, "step": 12284 }, { "epoch": 1.5627782724844166, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7496988773345947, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8692280650138855, "num_tokens": 466477575.0, "step": 12285 }, { "epoch": 1.5629054827630071, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5458054542541504, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8675776124000549, "num_tokens": 466516301.0, "step": 12286 }, { "epoch": 1.5630326930415976, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7599847316741943, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8607162237167358, "num_tokens": 466553904.0, "step": 12287 }, { "epoch": 1.5631599033201882, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.563646078109741, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8658628463745117, "num_tokens": 466591668.0, "step": 12288 }, { "epoch": 1.5632871135987787, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7490906715393066, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.870055615901947, "num_tokens": 466626484.0, "step": 12289 }, { "epoch": 1.5634143238773692, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5547523498535156, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8803982734680176, "num_tokens": 466663590.0, "step": 12290 }, { "epoch": 1.5635415341559598, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.764798879623413, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8840587139129639, "num_tokens": 466699856.0, "step": 12291 }, { "epoch": 1.5636687444345503, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.545480251312256, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.886102020740509, "num_tokens": 466738703.0, "step": 12292 }, { "epoch": 1.5637959547131408, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.770650625228882, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8766816854476929, "num_tokens": 466770948.0, "step": 12293 }, { "epoch": 1.5639231649917313, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5639421939849854, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8730225563049316, "num_tokens": 466804655.0, "step": 12294 }, { "epoch": 1.5640503752703219, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7495691776275635, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8771696090698242, "num_tokens": 466847592.0, "step": 12295 }, { "epoch": 1.5641775855489124, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5454351902008057, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8789753317832947, "num_tokens": 466887855.0, "step": 12296 }, { "epoch": 1.564304795827503, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7630252838134766, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8669981956481934, "num_tokens": 466925749.0, "step": 12297 }, { "epoch": 1.5644320061060935, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5398130416870117, "learning_rate": 1e-05, "loss": 0.3686, "mean_token_accuracy": 0.8935810923576355, "num_tokens": 466966487.0, "step": 12298 }, { "epoch": 1.564559216384684, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7748947143554688, "learning_rate": 1e-05, "loss": 0.5228, "mean_token_accuracy": 0.8480528593063354, "num_tokens": 467008673.0, "step": 12299 }, { "epoch": 1.5646864266632745, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.544184684753418, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8777728080749512, "num_tokens": 467048389.0, "step": 12300 }, { "epoch": 1.564813636941865, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.768972873687744, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.876750111579895, "num_tokens": 467087059.0, "step": 12301 }, { "epoch": 1.5649408472204556, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.554136276245117, "learning_rate": 1e-05, "loss": 0.3853, "mean_token_accuracy": 0.8888329267501831, "num_tokens": 467123003.0, "step": 12302 }, { "epoch": 1.5650680574990459, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.774928331375122, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8601037859916687, "num_tokens": 467160526.0, "step": 12303 }, { "epoch": 1.5651952677776364, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.567749261856079, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8691445589065552, "num_tokens": 467195846.0, "step": 12304 }, { "epoch": 1.565322478056227, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7728898525238037, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8604789972305298, "num_tokens": 467232412.0, "step": 12305 }, { "epoch": 1.5654496883348175, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5599000453948975, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8653271794319153, "num_tokens": 467269695.0, "step": 12306 }, { "epoch": 1.565576898613408, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7770049571990967, "learning_rate": 1e-05, "loss": 0.4871, "mean_token_accuracy": 0.8573246598243713, "num_tokens": 467304395.0, "step": 12307 }, { "epoch": 1.5657041088919985, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.545121908187866, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8704367280006409, "num_tokens": 467348606.0, "step": 12308 }, { "epoch": 1.5658313191705888, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7820863723754883, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8704788684844971, "num_tokens": 467382673.0, "step": 12309 }, { "epoch": 1.5659585294491793, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5413379669189453, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8785773515701294, "num_tokens": 467426291.0, "step": 12310 }, { "epoch": 1.5660857397277699, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7827258110046387, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8688346147537231, "num_tokens": 467469299.0, "step": 12311 }, { "epoch": 1.5662129500063604, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.54917311668396, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8651377558708191, "num_tokens": 467510876.0, "step": 12312 }, { "epoch": 1.566340160284951, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.8294930458068848, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.870475172996521, "num_tokens": 467546112.0, "step": 12313 }, { "epoch": 1.5664673705635415, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.55560040473938, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8734773397445679, "num_tokens": 467580009.0, "step": 12314 }, { "epoch": 1.566594580842132, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.797132968902588, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.86387038230896, "num_tokens": 467621009.0, "step": 12315 }, { "epoch": 1.5667217911207225, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.547367572784424, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8821998834609985, "num_tokens": 467662304.0, "step": 12316 }, { "epoch": 1.566849001399313, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.809973955154419, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8557302355766296, "num_tokens": 467697608.0, "step": 12317 }, { "epoch": 1.5669762116779036, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5549800395965576, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8734839558601379, "num_tokens": 467739978.0, "step": 12318 }, { "epoch": 1.567103421956494, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8101351261138916, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8605116605758667, "num_tokens": 467775055.0, "step": 12319 }, { "epoch": 1.5672306322350846, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.56663179397583, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8753641843795776, "num_tokens": 467812385.0, "step": 12320 }, { "epoch": 1.5673578425136752, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8101003170013428, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8562741875648499, "num_tokens": 467846784.0, "step": 12321 }, { "epoch": 1.5674850527922657, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5782663822174072, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8673394918441772, "num_tokens": 467880259.0, "step": 12322 }, { "epoch": 1.5676122630708562, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8144941329956055, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.871541440486908, "num_tokens": 467914613.0, "step": 12323 }, { "epoch": 1.5677394733494467, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5886502265930176, "learning_rate": 1e-05, "loss": 0.5068, "mean_token_accuracy": 0.8489528894424438, "num_tokens": 467953070.0, "step": 12324 }, { "epoch": 1.5678666836280373, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.771160840988159, "learning_rate": 1e-05, "loss": 0.4843, "mean_token_accuracy": 0.8584071397781372, "num_tokens": 467989404.0, "step": 12325 }, { "epoch": 1.5679938939066278, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5710184574127197, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8686230182647705, "num_tokens": 468030670.0, "step": 12326 }, { "epoch": 1.5681211041852183, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.766476631164551, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8691607713699341, "num_tokens": 468064130.0, "step": 12327 }, { "epoch": 1.5682483144638086, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5807957649230957, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8706289529800415, "num_tokens": 468098393.0, "step": 12328 }, { "epoch": 1.5683755247423992, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7498841285705566, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8822921514511108, "num_tokens": 468134961.0, "step": 12329 }, { "epoch": 1.5685027350209897, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.573171377182007, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8822368383407593, "num_tokens": 468169389.0, "step": 12330 }, { "epoch": 1.5686299452995802, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7423057556152344, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8628415465354919, "num_tokens": 468211504.0, "step": 12331 }, { "epoch": 1.5687571555781707, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.589456796646118, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8599218726158142, "num_tokens": 468245204.0, "step": 12332 }, { "epoch": 1.5688843658567613, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.737936019897461, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8726305365562439, "num_tokens": 468277054.0, "step": 12333 }, { "epoch": 1.5690115761353516, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.563878297805786, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8854309320449829, "num_tokens": 468316399.0, "step": 12334 }, { "epoch": 1.569138786413942, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.722702741622925, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8775812983512878, "num_tokens": 468351943.0, "step": 12335 }, { "epoch": 1.5692659966925326, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5695548057556152, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8695662021636963, "num_tokens": 468387443.0, "step": 12336 }, { "epoch": 1.5693932069711232, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7312984466552734, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8730310201644897, "num_tokens": 468420503.0, "step": 12337 }, { "epoch": 1.5695204172497137, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5722155570983887, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8653303384780884, "num_tokens": 468456171.0, "step": 12338 }, { "epoch": 1.5696476275283042, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7064802646636963, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8826324939727783, "num_tokens": 468494121.0, "step": 12339 }, { "epoch": 1.5697748378068948, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5565271377563477, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8727319240570068, "num_tokens": 468540960.0, "step": 12340 }, { "epoch": 1.5699020480854853, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7221477031707764, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8690992593765259, "num_tokens": 468577851.0, "step": 12341 }, { "epoch": 1.5700292583640758, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5725960731506348, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8695786595344543, "num_tokens": 468613210.0, "step": 12342 }, { "epoch": 1.5701564686426663, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7021703720092773, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8678088188171387, "num_tokens": 468655871.0, "step": 12343 }, { "epoch": 1.5702836789212569, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.565885305404663, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8797368407249451, "num_tokens": 468692757.0, "step": 12344 }, { "epoch": 1.5704108891998474, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7021546363830566, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8786617517471313, "num_tokens": 468730163.0, "step": 12345 }, { "epoch": 1.570538099478438, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.561668872833252, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8728163242340088, "num_tokens": 468768578.0, "step": 12346 }, { "epoch": 1.5706653097570284, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7037746906280518, "learning_rate": 1e-05, "loss": 0.3928, "mean_token_accuracy": 0.8834302425384521, "num_tokens": 468802027.0, "step": 12347 }, { "epoch": 1.570792520035619, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.562164783477783, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.861219048500061, "num_tokens": 468845266.0, "step": 12348 }, { "epoch": 1.5709197303142095, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7192792892456055, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8747021555900574, "num_tokens": 468879780.0, "step": 12349 }, { "epoch": 1.5710469405928, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5554275512695312, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8748750686645508, "num_tokens": 468916846.0, "step": 12350 }, { "epoch": 1.5711741508713906, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.720407724380493, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8835194110870361, "num_tokens": 468947382.0, "step": 12351 }, { "epoch": 1.5713013611499809, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.539313316345215, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.8847376108169556, "num_tokens": 468987258.0, "step": 12352 }, { "epoch": 1.5714285714285714, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7379300594329834, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8768969774246216, "num_tokens": 469024878.0, "step": 12353 }, { "epoch": 1.571555781707162, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.533139228820801, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8737900257110596, "num_tokens": 469065636.0, "step": 12354 }, { "epoch": 1.5716829919857525, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.755338191986084, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8702775239944458, "num_tokens": 469099817.0, "step": 12355 }, { "epoch": 1.571810202264343, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.545445442199707, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8680346608161926, "num_tokens": 469135185.0, "step": 12356 }, { "epoch": 1.5719374125429335, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7469499111175537, "learning_rate": 1e-05, "loss": 0.3874, "mean_token_accuracy": 0.8876861333847046, "num_tokens": 469173980.0, "step": 12357 }, { "epoch": 1.5720646228215238, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5557713508605957, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8761050701141357, "num_tokens": 469206939.0, "step": 12358 }, { "epoch": 1.5721918331001143, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.744145154953003, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8781870603561401, "num_tokens": 469245261.0, "step": 12359 }, { "epoch": 1.5723190433787049, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5482289791107178, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8700310587882996, "num_tokens": 469285569.0, "step": 12360 }, { "epoch": 1.5724462536572954, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7510950565338135, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8681089878082275, "num_tokens": 469325249.0, "step": 12361 }, { "epoch": 1.572573463935886, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5597424507141113, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8625591993331909, "num_tokens": 469360187.0, "step": 12362 }, { "epoch": 1.5727006742144765, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.73928165435791, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8675256371498108, "num_tokens": 469399363.0, "step": 12363 }, { "epoch": 1.572827884493067, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5514678955078125, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8687810897827148, "num_tokens": 469441523.0, "step": 12364 }, { "epoch": 1.5729550947716575, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7408077716827393, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8800694942474365, "num_tokens": 469475169.0, "step": 12365 }, { "epoch": 1.573082305050248, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.570528984069824, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8667848110198975, "num_tokens": 469509794.0, "step": 12366 }, { "epoch": 1.5732095153288386, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7316088676452637, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8667824268341064, "num_tokens": 469549007.0, "step": 12367 }, { "epoch": 1.573336725607429, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5482540130615234, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8756131529808044, "num_tokens": 469586959.0, "step": 12368 }, { "epoch": 1.5734639358860196, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7491116523742676, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8776010870933533, "num_tokens": 469617767.0, "step": 12369 }, { "epoch": 1.5735911461646102, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5506539344787598, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8674763441085815, "num_tokens": 469655300.0, "step": 12370 }, { "epoch": 1.5737183564432007, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.747560977935791, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8740531206130981, "num_tokens": 469693666.0, "step": 12371 }, { "epoch": 1.5738455667217912, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.545130968093872, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8733538389205933, "num_tokens": 469731081.0, "step": 12372 }, { "epoch": 1.5739727770003817, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.749908208847046, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8766148090362549, "num_tokens": 469768081.0, "step": 12373 }, { "epoch": 1.5740999872789723, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5479848384857178, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8703292012214661, "num_tokens": 469806992.0, "step": 12374 }, { "epoch": 1.5742271975575628, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.764481544494629, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8670881986618042, "num_tokens": 469844969.0, "step": 12375 }, { "epoch": 1.5743544078361533, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5466864109039307, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8773371577262878, "num_tokens": 469881558.0, "step": 12376 }, { "epoch": 1.5744816181147436, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.780634880065918, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8713784217834473, "num_tokens": 469914659.0, "step": 12377 }, { "epoch": 1.5746088283933342, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5479233264923096, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8663256168365479, "num_tokens": 469956012.0, "step": 12378 }, { "epoch": 1.5747360386719247, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7778539657592773, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8671582937240601, "num_tokens": 469993840.0, "step": 12379 }, { "epoch": 1.5748632489505152, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5456504821777344, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8612774610519409, "num_tokens": 470032741.0, "step": 12380 }, { "epoch": 1.5749904592291057, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7755303382873535, "learning_rate": 1e-05, "loss": 0.4003, "mean_token_accuracy": 0.882412314414978, "num_tokens": 470068969.0, "step": 12381 }, { "epoch": 1.5751176695076963, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.554042339324951, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8770170211791992, "num_tokens": 470108972.0, "step": 12382 }, { "epoch": 1.5752448797862866, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7620532512664795, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.8874981999397278, "num_tokens": 470146655.0, "step": 12383 }, { "epoch": 1.575372090064877, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5603201389312744, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8700098991394043, "num_tokens": 470184068.0, "step": 12384 }, { "epoch": 1.5754993003434676, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7666327953338623, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8694919347763062, "num_tokens": 470221328.0, "step": 12385 }, { "epoch": 1.5756265106220582, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5559120178222656, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8697752356529236, "num_tokens": 470261393.0, "step": 12386 }, { "epoch": 1.5757537209006487, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.775085926055908, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8706389665603638, "num_tokens": 470293705.0, "step": 12387 }, { "epoch": 1.5758809311792392, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5698912143707275, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.869520366191864, "num_tokens": 470329891.0, "step": 12388 }, { "epoch": 1.5760081414578297, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.755709409713745, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8732888698577881, "num_tokens": 470368763.0, "step": 12389 }, { "epoch": 1.5761353517364203, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.555957317352295, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8715884685516357, "num_tokens": 470407267.0, "step": 12390 }, { "epoch": 1.5762625620150108, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.749974012374878, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8698490858078003, "num_tokens": 470448872.0, "step": 12391 }, { "epoch": 1.5763897722936013, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5649189949035645, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8783762454986572, "num_tokens": 470486460.0, "step": 12392 }, { "epoch": 1.5765169825721919, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.733494281768799, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.882588803768158, "num_tokens": 470528800.0, "step": 12393 }, { "epoch": 1.5766441928507824, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5665204524993896, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8741211891174316, "num_tokens": 470568185.0, "step": 12394 }, { "epoch": 1.576771403129373, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.742745876312256, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8709225058555603, "num_tokens": 470607785.0, "step": 12395 }, { "epoch": 1.5768986134079634, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.555523157119751, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.86656653881073, "num_tokens": 470648536.0, "step": 12396 }, { "epoch": 1.577025823686554, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7321412563323975, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8789619207382202, "num_tokens": 470686071.0, "step": 12397 }, { "epoch": 1.5771530339651445, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5782978534698486, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8806868195533752, "num_tokens": 470715165.0, "step": 12398 }, { "epoch": 1.577280244243735, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7278637886047363, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8779381513595581, "num_tokens": 470748950.0, "step": 12399 }, { "epoch": 1.5774074545223256, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5579030513763428, "learning_rate": 1e-05, "loss": 0.3714, "mean_token_accuracy": 0.8890204429626465, "num_tokens": 470783856.0, "step": 12400 }, { "epoch": 1.5775346648009159, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7262542247772217, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.860808253288269, "num_tokens": 470823103.0, "step": 12401 }, { "epoch": 1.5776618750795064, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5629305839538574, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.8543238043785095, "num_tokens": 470865582.0, "step": 12402 }, { "epoch": 1.577789085358097, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7275335788726807, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8627504110336304, "num_tokens": 470901710.0, "step": 12403 }, { "epoch": 1.5779162956366874, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5555579662323, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8530029058456421, "num_tokens": 470944850.0, "step": 12404 }, { "epoch": 1.578043505915278, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7426719665527344, "learning_rate": 1e-05, "loss": 0.4853, "mean_token_accuracy": 0.8639898300170898, "num_tokens": 470985092.0, "step": 12405 }, { "epoch": 1.5781707161938685, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.545354127883911, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8785154819488525, "num_tokens": 471026018.0, "step": 12406 }, { "epoch": 1.5782979264724588, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7485785484313965, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8679921627044678, "num_tokens": 471061365.0, "step": 12407 }, { "epoch": 1.5784251367510493, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5514559745788574, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8750951290130615, "num_tokens": 471098142.0, "step": 12408 }, { "epoch": 1.5785523470296399, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.752760648727417, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8559050559997559, "num_tokens": 471138001.0, "step": 12409 }, { "epoch": 1.5786795573082304, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5559706687927246, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8762459754943848, "num_tokens": 471170392.0, "step": 12410 }, { "epoch": 1.578806767586821, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.743955135345459, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8643485307693481, "num_tokens": 471215704.0, "step": 12411 }, { "epoch": 1.5789339778654115, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5588769912719727, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8857811689376831, "num_tokens": 471252084.0, "step": 12412 }, { "epoch": 1.579061188144002, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7527923583984375, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8644436597824097, "num_tokens": 471287096.0, "step": 12413 }, { "epoch": 1.5791883984225925, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.55505108833313, "learning_rate": 1e-05, "loss": 0.3849, "mean_token_accuracy": 0.8864870071411133, "num_tokens": 471321732.0, "step": 12414 }, { "epoch": 1.579315608701183, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.757323741912842, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8701543211936951, "num_tokens": 471353680.0, "step": 12415 }, { "epoch": 1.5794428189797736, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5465221405029297, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.857469916343689, "num_tokens": 471395286.0, "step": 12416 }, { "epoch": 1.579570029258364, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7539587020874023, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8809573650360107, "num_tokens": 471433708.0, "step": 12417 }, { "epoch": 1.5796972395369546, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5499534606933594, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8707553148269653, "num_tokens": 471472655.0, "step": 12418 }, { "epoch": 1.5798244498155452, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.756251096725464, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8703795671463013, "num_tokens": 471507013.0, "step": 12419 }, { "epoch": 1.5799516600941357, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5665762424468994, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8739502429962158, "num_tokens": 471539951.0, "step": 12420 }, { "epoch": 1.5800788703727262, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7401812076568604, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8675245046615601, "num_tokens": 471580505.0, "step": 12421 }, { "epoch": 1.5802060806513167, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5699708461761475, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8818879127502441, "num_tokens": 471610408.0, "step": 12422 }, { "epoch": 1.5803332909299073, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.742889642715454, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8683146238327026, "num_tokens": 471648529.0, "step": 12423 }, { "epoch": 1.5804605012084978, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.573582172393799, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8731147050857544, "num_tokens": 471682182.0, "step": 12424 }, { "epoch": 1.5805877114870883, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.718580484390259, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8743084669113159, "num_tokens": 471719180.0, "step": 12425 }, { "epoch": 1.5807149217656786, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5555419921875, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.867983341217041, "num_tokens": 471759965.0, "step": 12426 }, { "epoch": 1.5808421320442692, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7250895500183105, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8823244571685791, "num_tokens": 471795405.0, "step": 12427 }, { "epoch": 1.5809693423228597, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.552168607711792, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.877902626991272, "num_tokens": 471837957.0, "step": 12428 }, { "epoch": 1.5810965526014502, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7302589416503906, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8699958324432373, "num_tokens": 471875422.0, "step": 12429 }, { "epoch": 1.5812237628800407, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.551044464111328, "learning_rate": 1e-05, "loss": 0.4156, "mean_token_accuracy": 0.8773326873779297, "num_tokens": 471913752.0, "step": 12430 }, { "epoch": 1.5813509731586313, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7360761165618896, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8640697002410889, "num_tokens": 471952362.0, "step": 12431 }, { "epoch": 1.5814781834372216, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5486929416656494, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8794642686843872, "num_tokens": 471990798.0, "step": 12432 }, { "epoch": 1.581605393715812, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.741424560546875, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8706689476966858, "num_tokens": 472027398.0, "step": 12433 }, { "epoch": 1.5817326039944026, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.536905527114868, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8758625984191895, "num_tokens": 472071725.0, "step": 12434 }, { "epoch": 1.5818598142729932, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7551567554473877, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8839041590690613, "num_tokens": 472106913.0, "step": 12435 }, { "epoch": 1.5819870245515837, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5617032051086426, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8714045882225037, "num_tokens": 472138136.0, "step": 12436 }, { "epoch": 1.5821142348301742, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7573583126068115, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8676796555519104, "num_tokens": 472174884.0, "step": 12437 }, { "epoch": 1.5822414451087647, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5471155643463135, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8672574162483215, "num_tokens": 472215862.0, "step": 12438 }, { "epoch": 1.5823686553873553, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7464470863342285, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8781986236572266, "num_tokens": 472255770.0, "step": 12439 }, { "epoch": 1.5824958656659458, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5581417083740234, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8626608848571777, "num_tokens": 472290815.0, "step": 12440 }, { "epoch": 1.5826230759445363, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7520055770874023, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.861215353012085, "num_tokens": 472328180.0, "step": 12441 }, { "epoch": 1.5827502862231269, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5535390377044678, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8626048564910889, "num_tokens": 472366519.0, "step": 12442 }, { "epoch": 1.5828774965017174, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7468185424804688, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8559790849685669, "num_tokens": 472405270.0, "step": 12443 }, { "epoch": 1.583004706780308, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.538602113723755, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8771504163742065, "num_tokens": 472447572.0, "step": 12444 }, { "epoch": 1.5831319170588984, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7480998039245605, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8749671578407288, "num_tokens": 472487869.0, "step": 12445 }, { "epoch": 1.583259127337489, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5586347579956055, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8741324543952942, "num_tokens": 472521856.0, "step": 12446 }, { "epoch": 1.5833863376160795, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.741856813430786, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8739887475967407, "num_tokens": 472566550.0, "step": 12447 }, { "epoch": 1.58351354789467, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.553931951522827, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8684347867965698, "num_tokens": 472605054.0, "step": 12448 }, { "epoch": 1.5836407581732606, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7343344688415527, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8735606074333191, "num_tokens": 472649938.0, "step": 12449 }, { "epoch": 1.5837679684518509, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5501177310943604, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8752105236053467, "num_tokens": 472688399.0, "step": 12450 }, { "epoch": 1.5838951787304414, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7546002864837646, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8660784363746643, "num_tokens": 472726383.0, "step": 12451 }, { "epoch": 1.584022389009032, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.566826105117798, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8662340641021729, "num_tokens": 472759692.0, "step": 12452 }, { "epoch": 1.5841495992876224, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.738438129425049, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.878139317035675, "num_tokens": 472799679.0, "step": 12453 }, { "epoch": 1.584276809566213, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.555584192276001, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.858543872833252, "num_tokens": 472842043.0, "step": 12454 }, { "epoch": 1.5844040198448035, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.747220039367676, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8666254281997681, "num_tokens": 472877594.0, "step": 12455 }, { "epoch": 1.5845312301233938, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5483779907226562, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.87450110912323, "num_tokens": 472914052.0, "step": 12456 }, { "epoch": 1.5846584404019843, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.751025915145874, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8735730648040771, "num_tokens": 472953359.0, "step": 12457 }, { "epoch": 1.5847856506805749, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5434653759002686, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8662360906600952, "num_tokens": 472993898.0, "step": 12458 }, { "epoch": 1.5849128609591654, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.760885715484619, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8756428360939026, "num_tokens": 473032026.0, "step": 12459 }, { "epoch": 1.585040071237756, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5303688049316406, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8784116506576538, "num_tokens": 473075282.0, "step": 12460 }, { "epoch": 1.5851672815163464, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7694942951202393, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.876482367515564, "num_tokens": 473112834.0, "step": 12461 }, { "epoch": 1.585294491794937, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5486092567443848, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8693400025367737, "num_tokens": 473150331.0, "step": 12462 }, { "epoch": 1.5854217020735275, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.773155450820923, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8777635097503662, "num_tokens": 473185175.0, "step": 12463 }, { "epoch": 1.585548912352118, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5494558811187744, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8627527952194214, "num_tokens": 473223546.0, "step": 12464 }, { "epoch": 1.5856761226307086, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.783416748046875, "learning_rate": 1e-05, "loss": 0.5018, "mean_token_accuracy": 0.8579251766204834, "num_tokens": 473260098.0, "step": 12465 }, { "epoch": 1.585803332909299, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.552252769470215, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8738046884536743, "num_tokens": 473299283.0, "step": 12466 }, { "epoch": 1.5859305431878896, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.786489963531494, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8671998977661133, "num_tokens": 473332511.0, "step": 12467 }, { "epoch": 1.5860577534664801, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5605545043945312, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8758686780929565, "num_tokens": 473367259.0, "step": 12468 }, { "epoch": 1.5861849637450707, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7616937160491943, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8704640865325928, "num_tokens": 473412421.0, "step": 12469 }, { "epoch": 1.5863121740236612, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5718328952789307, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8505812883377075, "num_tokens": 473450444.0, "step": 12470 }, { "epoch": 1.5864393843022517, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.764596700668335, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8694084286689758, "num_tokens": 473485093.0, "step": 12471 }, { "epoch": 1.5865665945808423, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.55423903465271, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8831005096435547, "num_tokens": 473522461.0, "step": 12472 }, { "epoch": 1.5866938048594328, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7600760459899902, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8666438460350037, "num_tokens": 473559860.0, "step": 12473 }, { "epoch": 1.5868210151380233, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5717780590057373, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8811368942260742, "num_tokens": 473590144.0, "step": 12474 }, { "epoch": 1.5869482254166136, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.745156764984131, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8769280910491943, "num_tokens": 473631384.0, "step": 12475 }, { "epoch": 1.5870754356952042, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5798826217651367, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8706314563751221, "num_tokens": 473667396.0, "step": 12476 }, { "epoch": 1.5872026459737947, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7436423301696777, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.87471604347229, "num_tokens": 473699856.0, "step": 12477 }, { "epoch": 1.5873298562523852, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.593001365661621, "learning_rate": 1e-05, "loss": 0.4979, "mean_token_accuracy": 0.851376473903656, "num_tokens": 473731164.0, "step": 12478 }, { "epoch": 1.5874570665309757, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7227964401245117, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.86919105052948, "num_tokens": 473767005.0, "step": 12479 }, { "epoch": 1.5875842768095663, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.57755446434021, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8691141605377197, "num_tokens": 473805753.0, "step": 12480 }, { "epoch": 1.5877114870881566, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.700629949569702, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8712606430053711, "num_tokens": 473845375.0, "step": 12481 }, { "epoch": 1.587838697366747, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.558887243270874, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8619856834411621, "num_tokens": 473889184.0, "step": 12482 }, { "epoch": 1.5879659076453376, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7020537853240967, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8797663450241089, "num_tokens": 473927917.0, "step": 12483 }, { "epoch": 1.5880931179239282, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.565473794937134, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8830457925796509, "num_tokens": 473964670.0, "step": 12484 }, { "epoch": 1.5882203282025187, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7085306644439697, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8865184783935547, "num_tokens": 473998455.0, "step": 12485 }, { "epoch": 1.5883475384811092, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5571200847625732, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8700731992721558, "num_tokens": 474041623.0, "step": 12486 }, { "epoch": 1.5884747487596997, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.706040143966675, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8693761825561523, "num_tokens": 474083294.0, "step": 12487 }, { "epoch": 1.5886019590382903, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5658557415008545, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8712229132652283, "num_tokens": 474117737.0, "step": 12488 }, { "epoch": 1.5887291693168808, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.709095001220703, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8685367107391357, "num_tokens": 474154467.0, "step": 12489 }, { "epoch": 1.5888563795954713, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5612027645111084, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8684712052345276, "num_tokens": 474191845.0, "step": 12490 }, { "epoch": 1.5889835898740619, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7153165340423584, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8747181296348572, "num_tokens": 474227787.0, "step": 12491 }, { "epoch": 1.5891108001526524, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.541569948196411, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.868033766746521, "num_tokens": 474271890.0, "step": 12492 }, { "epoch": 1.589238010431243, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.722050905227661, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.857979416847229, "num_tokens": 474314493.0, "step": 12493 }, { "epoch": 1.5893652207098334, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5496723651885986, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8806093335151672, "num_tokens": 474349348.0, "step": 12494 }, { "epoch": 1.589492430988424, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.742969512939453, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8653475046157837, "num_tokens": 474384145.0, "step": 12495 }, { "epoch": 1.5896196412670145, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.545398473739624, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8815889358520508, "num_tokens": 474419064.0, "step": 12496 }, { "epoch": 1.589746851545605, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.730132818222046, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8688596487045288, "num_tokens": 474458716.0, "step": 12497 }, { "epoch": 1.5898740618241956, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.54557204246521, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8768643140792847, "num_tokens": 474497672.0, "step": 12498 }, { "epoch": 1.5900012721027859, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7517940998077393, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8759807348251343, "num_tokens": 474530442.0, "step": 12499 }, { "epoch": 1.5901284823813764, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.538719654083252, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.854711651802063, "num_tokens": 474573103.0, "step": 12500 }, { "epoch": 1.590255692659967, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.749849557876587, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8789174556732178, "num_tokens": 474610687.0, "step": 12501 }, { "epoch": 1.5903829029385574, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5431854724884033, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8646849989891052, "num_tokens": 474650893.0, "step": 12502 }, { "epoch": 1.590510113217148, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7533676624298096, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8718273639678955, "num_tokens": 474690808.0, "step": 12503 }, { "epoch": 1.5906373234957385, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5424797534942627, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8810839653015137, "num_tokens": 474726814.0, "step": 12504 }, { "epoch": 1.5907645337743288, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.768186330795288, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8678338527679443, "num_tokens": 474765170.0, "step": 12505 }, { "epoch": 1.5908917440529193, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.541658639907837, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8672815561294556, "num_tokens": 474808466.0, "step": 12506 }, { "epoch": 1.5910189543315099, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7686917781829834, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8776014447212219, "num_tokens": 474847334.0, "step": 12507 }, { "epoch": 1.5911461646101004, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.55107045173645, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8715271353721619, "num_tokens": 474885937.0, "step": 12508 }, { "epoch": 1.591273374888691, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7717716693878174, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.875471830368042, "num_tokens": 474921089.0, "step": 12509 }, { "epoch": 1.5914005851672814, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5465829372406006, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8770726919174194, "num_tokens": 474961660.0, "step": 12510 }, { "epoch": 1.591527795445872, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.780047655105591, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8740004301071167, "num_tokens": 474997815.0, "step": 12511 }, { "epoch": 1.5916550057244625, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.553234815597534, "learning_rate": 1e-05, "loss": 0.5084, "mean_token_accuracy": 0.8452795147895813, "num_tokens": 475038334.0, "step": 12512 }, { "epoch": 1.591782216003053, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7821221351623535, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8740890026092529, "num_tokens": 475073522.0, "step": 12513 }, { "epoch": 1.5919094262816436, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.547684907913208, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8779264092445374, "num_tokens": 475111811.0, "step": 12514 }, { "epoch": 1.592036636560234, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7800562381744385, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.8880454301834106, "num_tokens": 475146496.0, "step": 12515 }, { "epoch": 1.5921638468388246, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.555185556411743, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8708480596542358, "num_tokens": 475187947.0, "step": 12516 }, { "epoch": 1.5922910571174151, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.779343366622925, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8728405833244324, "num_tokens": 475225420.0, "step": 12517 }, { "epoch": 1.5924182673960057, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5625646114349365, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8635770082473755, "num_tokens": 475263913.0, "step": 12518 }, { "epoch": 1.5925454776745962, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.778001070022583, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8828263878822327, "num_tokens": 475300446.0, "step": 12519 }, { "epoch": 1.5926726879531867, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5780751705169678, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8601125478744507, "num_tokens": 475338728.0, "step": 12520 }, { "epoch": 1.5927998982317773, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.774510622024536, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8712582588195801, "num_tokens": 475369453.0, "step": 12521 }, { "epoch": 1.5929271085103678, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5598220825195312, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8848543167114258, "num_tokens": 475407994.0, "step": 12522 }, { "epoch": 1.5930543187889583, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.774050712585449, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8674141764640808, "num_tokens": 475445480.0, "step": 12523 }, { "epoch": 1.5931815290675486, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5707151889801025, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8595610857009888, "num_tokens": 475483971.0, "step": 12524 }, { "epoch": 1.5933087393461391, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7582249641418457, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8739607334136963, "num_tokens": 475521332.0, "step": 12525 }, { "epoch": 1.5934359496247297, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5758421421051025, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8659706115722656, "num_tokens": 475556085.0, "step": 12526 }, { "epoch": 1.5935631599033202, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7910635471343994, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8674558997154236, "num_tokens": 475590966.0, "step": 12527 }, { "epoch": 1.5936903701819107, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.570465087890625, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8640034794807434, "num_tokens": 475627430.0, "step": 12528 }, { "epoch": 1.5938175804605013, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7597496509552, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8801428079605103, "num_tokens": 475662346.0, "step": 12529 }, { "epoch": 1.5939447907390916, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5646820068359375, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8629992008209229, "num_tokens": 475702312.0, "step": 12530 }, { "epoch": 1.594072001017682, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.746150493621826, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8716799020767212, "num_tokens": 475743997.0, "step": 12531 }, { "epoch": 1.5941992112962726, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5680782794952393, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8609060645103455, "num_tokens": 475783369.0, "step": 12532 }, { "epoch": 1.5943264215748632, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.74878191947937, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8852555751800537, "num_tokens": 475815225.0, "step": 12533 }, { "epoch": 1.5944536318534537, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.577058792114258, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8665629625320435, "num_tokens": 475847917.0, "step": 12534 }, { "epoch": 1.5945808421320442, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.73447322845459, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.876289963722229, "num_tokens": 475889308.0, "step": 12535 }, { "epoch": 1.5947080524106347, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.568464756011963, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8803336024284363, "num_tokens": 475930603.0, "step": 12536 }, { "epoch": 1.5948352626892253, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.722874164581299, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.878951907157898, "num_tokens": 475968288.0, "step": 12537 }, { "epoch": 1.5949624729678158, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5765438079833984, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.876685619354248, "num_tokens": 476002967.0, "step": 12538 }, { "epoch": 1.5950896832464063, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7265896797180176, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8734406232833862, "num_tokens": 476037958.0, "step": 12539 }, { "epoch": 1.5952168935249968, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5884788036346436, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8513677716255188, "num_tokens": 476072036.0, "step": 12540 }, { "epoch": 1.5953441038035874, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.701174259185791, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.875009298324585, "num_tokens": 476115665.0, "step": 12541 }, { "epoch": 1.595471314082178, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.575260639190674, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8665950298309326, "num_tokens": 476155380.0, "step": 12542 }, { "epoch": 1.5955985243607684, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.69803786277771, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8659787178039551, "num_tokens": 476194833.0, "step": 12543 }, { "epoch": 1.595725734639359, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5763602256774902, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8743764758110046, "num_tokens": 476233617.0, "step": 12544 }, { "epoch": 1.5958529449179495, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.706904411315918, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8676126599311829, "num_tokens": 476270312.0, "step": 12545 }, { "epoch": 1.59598015519654, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.558917999267578, "learning_rate": 1e-05, "loss": 0.3554, "mean_token_accuracy": 0.8970924615859985, "num_tokens": 476304466.0, "step": 12546 }, { "epoch": 1.5961073654751305, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7078583240509033, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.87515789270401, "num_tokens": 476345136.0, "step": 12547 }, { "epoch": 1.5962345757537209, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.562828779220581, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8761447668075562, "num_tokens": 476381546.0, "step": 12548 }, { "epoch": 1.5963617860323114, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.731010675430298, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8719804883003235, "num_tokens": 476413990.0, "step": 12549 }, { "epoch": 1.596488996310902, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.549729347229004, "learning_rate": 1e-05, "loss": 0.4013, "mean_token_accuracy": 0.8841062784194946, "num_tokens": 476451487.0, "step": 12550 }, { "epoch": 1.5966162065894924, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.732386589050293, "learning_rate": 1e-05, "loss": 0.4854, "mean_token_accuracy": 0.8587257862091064, "num_tokens": 476493071.0, "step": 12551 }, { "epoch": 1.596743416868083, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.549502372741699, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8785516023635864, "num_tokens": 476531390.0, "step": 12552 }, { "epoch": 1.5968706271466735, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7348132133483887, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8683570623397827, "num_tokens": 476572491.0, "step": 12553 }, { "epoch": 1.5969978374252638, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5476555824279785, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8639505505561829, "num_tokens": 476613897.0, "step": 12554 }, { "epoch": 1.5971250477038543, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7480130195617676, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8648799657821655, "num_tokens": 476650773.0, "step": 12555 }, { "epoch": 1.5972522579824449, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.552852153778076, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8722578883171082, "num_tokens": 476685546.0, "step": 12556 }, { "epoch": 1.5973794682610354, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.764646291732788, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.8548679947853088, "num_tokens": 476720569.0, "step": 12557 }, { "epoch": 1.597506678539626, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5363900661468506, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8753530979156494, "num_tokens": 476760162.0, "step": 12558 }, { "epoch": 1.5976338888182164, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.779160261154175, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8702384233474731, "num_tokens": 476793575.0, "step": 12559 }, { "epoch": 1.597761099096807, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.555697202682495, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8747859001159668, "num_tokens": 476826714.0, "step": 12560 }, { "epoch": 1.5978883093753975, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7679049968719482, "learning_rate": 1e-05, "loss": 0.3958, "mean_token_accuracy": 0.8863003253936768, "num_tokens": 476866715.0, "step": 12561 }, { "epoch": 1.598015519653988, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5430800914764404, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8671712875366211, "num_tokens": 476906036.0, "step": 12562 }, { "epoch": 1.5981427299325786, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7664663791656494, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8710335493087769, "num_tokens": 476945005.0, "step": 12563 }, { "epoch": 1.598269940211169, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5508227348327637, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8776489496231079, "num_tokens": 476989491.0, "step": 12564 }, { "epoch": 1.5983971504897596, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7728874683380127, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8703017830848694, "num_tokens": 477026062.0, "step": 12565 }, { "epoch": 1.5985243607683501, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.560753345489502, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.869164228439331, "num_tokens": 477060045.0, "step": 12566 }, { "epoch": 1.5986515710469407, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.760399341583252, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8842145204544067, "num_tokens": 477102343.0, "step": 12567 }, { "epoch": 1.5987787813255312, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5583202838897705, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.879867672920227, "num_tokens": 477136239.0, "step": 12568 }, { "epoch": 1.5989059916041217, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.756518602371216, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8693544864654541, "num_tokens": 477178291.0, "step": 12569 }, { "epoch": 1.5990332018827123, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5818285942077637, "learning_rate": 1e-05, "loss": 0.4876, "mean_token_accuracy": 0.8537218570709229, "num_tokens": 477213993.0, "step": 12570 }, { "epoch": 1.5991604121613028, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7521722316741943, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8622121810913086, "num_tokens": 477258384.0, "step": 12571 }, { "epoch": 1.5992876224398933, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.561527729034424, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8722068071365356, "num_tokens": 477297386.0, "step": 12572 }, { "epoch": 1.5994148327184836, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7428598403930664, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8804084062576294, "num_tokens": 477334214.0, "step": 12573 }, { "epoch": 1.5995420429970741, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.564613103866577, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8599286079406738, "num_tokens": 477374777.0, "step": 12574 }, { "epoch": 1.5996692532756647, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.737220048904419, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8702125549316406, "num_tokens": 477415206.0, "step": 12575 }, { "epoch": 1.5997964635542552, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.559548854827881, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8686603307723999, "num_tokens": 477453864.0, "step": 12576 }, { "epoch": 1.5999236738328457, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.754023313522339, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8554003238677979, "num_tokens": 477489236.0, "step": 12577 }, { "epoch": 1.6000508841114363, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.567563772201538, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.883220911026001, "num_tokens": 477524459.0, "step": 12578 }, { "epoch": 1.6001780943900266, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7420010566711426, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8685626983642578, "num_tokens": 477560045.0, "step": 12579 }, { "epoch": 1.600305304668617, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5658321380615234, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8594858646392822, "num_tokens": 477596438.0, "step": 12580 }, { "epoch": 1.6004325149472076, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7369396686553955, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8715929985046387, "num_tokens": 477634370.0, "step": 12581 }, { "epoch": 1.6005597252257981, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5766265392303467, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.864109992980957, "num_tokens": 477669798.0, "step": 12582 }, { "epoch": 1.6006869355043887, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7085628509521484, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8702649474143982, "num_tokens": 477714127.0, "step": 12583 }, { "epoch": 1.6008141457829792, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5664618015289307, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8761987686157227, "num_tokens": 477753176.0, "step": 12584 }, { "epoch": 1.6009413560615697, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.712146282196045, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.873458981513977, "num_tokens": 477789372.0, "step": 12585 }, { "epoch": 1.6010685663401603, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.57932186126709, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8673635125160217, "num_tokens": 477820891.0, "step": 12586 }, { "epoch": 1.6011957766187508, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7276611328125, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8700853586196899, "num_tokens": 477853952.0, "step": 12587 }, { "epoch": 1.6013229868973413, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.553088665008545, "learning_rate": 1e-05, "loss": 0.3971, "mean_token_accuracy": 0.8859109878540039, "num_tokens": 477888692.0, "step": 12588 }, { "epoch": 1.6014501971759318, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7162795066833496, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8833473920822144, "num_tokens": 477923769.0, "step": 12589 }, { "epoch": 1.6015774074545224, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5492329597473145, "learning_rate": 1e-05, "loss": 0.3834, "mean_token_accuracy": 0.8877993822097778, "num_tokens": 477959997.0, "step": 12590 }, { "epoch": 1.601704617733113, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.743380308151245, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8556557893753052, "num_tokens": 477996211.0, "step": 12591 }, { "epoch": 1.6018318280117034, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.560527801513672, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8725961446762085, "num_tokens": 478029817.0, "step": 12592 }, { "epoch": 1.601959038290294, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.735116958618164, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8802433013916016, "num_tokens": 478071216.0, "step": 12593 }, { "epoch": 1.6020862485688845, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5427935123443604, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8777264952659607, "num_tokens": 478112581.0, "step": 12594 }, { "epoch": 1.602213458847475, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7548632621765137, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8594756126403809, "num_tokens": 478148887.0, "step": 12595 }, { "epoch": 1.6023406691260655, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.592641830444336, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8607866764068604, "num_tokens": 478187580.0, "step": 12596 }, { "epoch": 1.6024678794046558, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7247087955474854, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8747233152389526, "num_tokens": 478222679.0, "step": 12597 }, { "epoch": 1.6025950896832464, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5638880729675293, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.877959132194519, "num_tokens": 478259464.0, "step": 12598 }, { "epoch": 1.602722299961837, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.714285373687744, "learning_rate": 1e-05, "loss": 0.3909, "mean_token_accuracy": 0.8854010701179504, "num_tokens": 478293790.0, "step": 12599 }, { "epoch": 1.6028495102404274, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.568216323852539, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8748806715011597, "num_tokens": 478329311.0, "step": 12600 }, { "epoch": 1.602976720519018, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7272145748138428, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8732703328132629, "num_tokens": 478362818.0, "step": 12601 }, { "epoch": 1.6031039307976085, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5575530529022217, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8694345355033875, "num_tokens": 478402783.0, "step": 12602 }, { "epoch": 1.6032311410761988, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7280538082122803, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8594183921813965, "num_tokens": 478438141.0, "step": 12603 }, { "epoch": 1.6033583513547893, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5638835430145264, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8791926503181458, "num_tokens": 478473176.0, "step": 12604 }, { "epoch": 1.6034855616333799, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7264814376831055, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8767897486686707, "num_tokens": 478511899.0, "step": 12605 }, { "epoch": 1.6036127719119704, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5532872676849365, "learning_rate": 1e-05, "loss": 0.3896, "mean_token_accuracy": 0.8853805065155029, "num_tokens": 478545018.0, "step": 12606 }, { "epoch": 1.603739982190561, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7414464950561523, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8663106560707092, "num_tokens": 478582384.0, "step": 12607 }, { "epoch": 1.6038671924691514, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.547734498977661, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8705316781997681, "num_tokens": 478621491.0, "step": 12608 }, { "epoch": 1.603994402747742, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7276673316955566, "learning_rate": 1e-05, "loss": 0.3801, "mean_token_accuracy": 0.8930362462997437, "num_tokens": 478662018.0, "step": 12609 }, { "epoch": 1.6041216130263325, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.545903444290161, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8701788187026978, "num_tokens": 478702730.0, "step": 12610 }, { "epoch": 1.604248823304923, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7499823570251465, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8695515990257263, "num_tokens": 478736696.0, "step": 12611 }, { "epoch": 1.6043760335835135, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.549471378326416, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8521549701690674, "num_tokens": 478775873.0, "step": 12612 }, { "epoch": 1.604503243862104, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.753664970397949, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8789588809013367, "num_tokens": 478810571.0, "step": 12613 }, { "epoch": 1.6046304541406946, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.550117015838623, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8595528602600098, "num_tokens": 478845696.0, "step": 12614 }, { "epoch": 1.6047576644192851, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.750638484954834, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8665521740913391, "num_tokens": 478884561.0, "step": 12615 }, { "epoch": 1.6048848746978757, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5710020065307617, "learning_rate": 1e-05, "loss": 0.5197, "mean_token_accuracy": 0.849557638168335, "num_tokens": 478921989.0, "step": 12616 }, { "epoch": 1.6050120849764662, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7457072734832764, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8712550401687622, "num_tokens": 478958346.0, "step": 12617 }, { "epoch": 1.6051392952550567, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.551274061203003, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8731385469436646, "num_tokens": 479000141.0, "step": 12618 }, { "epoch": 1.6052665055336472, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7418558597564697, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.878398597240448, "num_tokens": 479038785.0, "step": 12619 }, { "epoch": 1.6053937158122378, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5678133964538574, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8632978796958923, "num_tokens": 479074679.0, "step": 12620 }, { "epoch": 1.6055209260908283, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7507519721984863, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8772640228271484, "num_tokens": 479106745.0, "step": 12621 }, { "epoch": 1.6056481363694186, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.557748794555664, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8739123344421387, "num_tokens": 479145159.0, "step": 12622 }, { "epoch": 1.6057753466480091, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.73770809173584, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8840979337692261, "num_tokens": 479181033.0, "step": 12623 }, { "epoch": 1.6059025569265997, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.559917449951172, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8671263456344604, "num_tokens": 479223954.0, "step": 12624 }, { "epoch": 1.6060297672051902, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.736013650894165, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8575677871704102, "num_tokens": 479268676.0, "step": 12625 }, { "epoch": 1.6061569774837807, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5806803703308105, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8821672201156616, "num_tokens": 479305203.0, "step": 12626 }, { "epoch": 1.6062841877623713, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7334020137786865, "learning_rate": 1e-05, "loss": 0.5216, "mean_token_accuracy": 0.8457457423210144, "num_tokens": 479347358.0, "step": 12627 }, { "epoch": 1.6064113980409616, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5584452152252197, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8630410432815552, "num_tokens": 479392973.0, "step": 12628 }, { "epoch": 1.606538608319552, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.746859073638916, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8594294786453247, "num_tokens": 479430832.0, "step": 12629 }, { "epoch": 1.6066658185981426, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5535950660705566, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8594092726707458, "num_tokens": 479471363.0, "step": 12630 }, { "epoch": 1.6067930288767331, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7428970336914062, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8763893842697144, "num_tokens": 479506041.0, "step": 12631 }, { "epoch": 1.6069202391553237, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.574082136154175, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8721548318862915, "num_tokens": 479537290.0, "step": 12632 }, { "epoch": 1.6070474494339142, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.751345157623291, "learning_rate": 1e-05, "loss": 0.4908, "mean_token_accuracy": 0.8559896945953369, "num_tokens": 479573191.0, "step": 12633 }, { "epoch": 1.6071746597125047, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5532853603363037, "learning_rate": 1e-05, "loss": 0.3874, "mean_token_accuracy": 0.8848196268081665, "num_tokens": 479606210.0, "step": 12634 }, { "epoch": 1.6073018699910953, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.739931583404541, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.8863165378570557, "num_tokens": 479645198.0, "step": 12635 }, { "epoch": 1.6074290802696858, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.546675443649292, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8675765991210938, "num_tokens": 479686319.0, "step": 12636 }, { "epoch": 1.6075562905482763, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7544920444488525, "learning_rate": 1e-05, "loss": 0.3838, "mean_token_accuracy": 0.8920429944992065, "num_tokens": 479720170.0, "step": 12637 }, { "epoch": 1.6076835008268668, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5450823307037354, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.876952588558197, "num_tokens": 479757263.0, "step": 12638 }, { "epoch": 1.6078107111054574, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.767561912536621, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8694018125534058, "num_tokens": 479793050.0, "step": 12639 }, { "epoch": 1.607937921384048, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5642590522766113, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8685252070426941, "num_tokens": 479831705.0, "step": 12640 }, { "epoch": 1.6080651316626384, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.752060651779175, "learning_rate": 1e-05, "loss": 0.3923, "mean_token_accuracy": 0.8841521739959717, "num_tokens": 479863888.0, "step": 12641 }, { "epoch": 1.608192341941229, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.549326181411743, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8804267644882202, "num_tokens": 479902280.0, "step": 12642 }, { "epoch": 1.6083195522198195, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.760826587677002, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8748090267181396, "num_tokens": 479935649.0, "step": 12643 }, { "epoch": 1.60844676249841, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.558981418609619, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8691497445106506, "num_tokens": 479971834.0, "step": 12644 }, { "epoch": 1.6085739727770005, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.764124870300293, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8579006195068359, "num_tokens": 480011985.0, "step": 12645 }, { "epoch": 1.6087011830555908, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5663199424743652, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8591925501823425, "num_tokens": 480047833.0, "step": 12646 }, { "epoch": 1.6088283933341814, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7691943645477295, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8648375272750854, "num_tokens": 480080774.0, "step": 12647 }, { "epoch": 1.608955603612772, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5511269569396973, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.87284916639328, "num_tokens": 480119114.0, "step": 12648 }, { "epoch": 1.6090828138913624, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7664365768432617, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8654735684394836, "num_tokens": 480156624.0, "step": 12649 }, { "epoch": 1.609210024169953, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.560760021209717, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8721286058425903, "num_tokens": 480192573.0, "step": 12650 }, { "epoch": 1.6093372344485435, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7705819606781006, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8566710948944092, "num_tokens": 480232139.0, "step": 12651 }, { "epoch": 1.6094644447271338, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.564115047454834, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8632395267486572, "num_tokens": 480269811.0, "step": 12652 }, { "epoch": 1.6095916550057243, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7556562423706055, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8652610778808594, "num_tokens": 480307785.0, "step": 12653 }, { "epoch": 1.6097188652843148, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5571038722991943, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8693708777427673, "num_tokens": 480347028.0, "step": 12654 }, { "epoch": 1.6098460755629054, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7582719326019287, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8780558109283447, "num_tokens": 480383915.0, "step": 12655 }, { "epoch": 1.609973285841496, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.553190231323242, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.872771680355072, "num_tokens": 480424226.0, "step": 12656 }, { "epoch": 1.6101004961200864, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7611660957336426, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.881596565246582, "num_tokens": 480461920.0, "step": 12657 }, { "epoch": 1.610227706398677, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.556678295135498, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8782130479812622, "num_tokens": 480498509.0, "step": 12658 }, { "epoch": 1.6103549166772675, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7605628967285156, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8622761368751526, "num_tokens": 480541912.0, "step": 12659 }, { "epoch": 1.610482126955858, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5668880939483643, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8750616312026978, "num_tokens": 480576034.0, "step": 12660 }, { "epoch": 1.6106093372344485, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.770838737487793, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8642040491104126, "num_tokens": 480612265.0, "step": 12661 }, { "epoch": 1.610736547513039, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6196603775024414, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8529143929481506, "num_tokens": 480650667.0, "step": 12662 }, { "epoch": 1.6108637577916296, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7271459102630615, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8688527345657349, "num_tokens": 480688931.0, "step": 12663 }, { "epoch": 1.6109909680702201, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.600456953048706, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8639407157897949, "num_tokens": 480719525.0, "step": 12664 }, { "epoch": 1.6111181783488107, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.692176580429077, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8815836906433105, "num_tokens": 480755872.0, "step": 12665 }, { "epoch": 1.6112453886274012, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.585301160812378, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8746484518051147, "num_tokens": 480794994.0, "step": 12666 }, { "epoch": 1.6113725989059917, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6875102519989014, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8779004812240601, "num_tokens": 480828421.0, "step": 12667 }, { "epoch": 1.6114998091845822, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.576399564743042, "learning_rate": 1e-05, "loss": 0.3936, "mean_token_accuracy": 0.8825317621231079, "num_tokens": 480868571.0, "step": 12668 }, { "epoch": 1.6116270194631728, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.682121753692627, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.8833253383636475, "num_tokens": 480910897.0, "step": 12669 }, { "epoch": 1.6117542297417633, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5971486568450928, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8715935945510864, "num_tokens": 480942243.0, "step": 12670 }, { "epoch": 1.6118814400203536, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.682171106338501, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8612709045410156, "num_tokens": 480980510.0, "step": 12671 }, { "epoch": 1.6120086502989441, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.584052324295044, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8773493766784668, "num_tokens": 481016837.0, "step": 12672 }, { "epoch": 1.6121358605775347, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.677659273147583, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8709867000579834, "num_tokens": 481055862.0, "step": 12673 }, { "epoch": 1.6122630708561252, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.583765983581543, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8634499311447144, "num_tokens": 481094632.0, "step": 12674 }, { "epoch": 1.6123902811347157, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6636743545532227, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8714055418968201, "num_tokens": 481138468.0, "step": 12675 }, { "epoch": 1.6125174914133062, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.568640947341919, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.867068886756897, "num_tokens": 481182983.0, "step": 12676 }, { "epoch": 1.6126447016918966, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6639513969421387, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8722177743911743, "num_tokens": 481225805.0, "step": 12677 }, { "epoch": 1.612771911970487, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5816144943237305, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8753217458724976, "num_tokens": 481260532.0, "step": 12678 }, { "epoch": 1.6128991222490776, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6731345653533936, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8670852184295654, "num_tokens": 481300494.0, "step": 12679 }, { "epoch": 1.6130263325276681, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.559345245361328, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8846527338027954, "num_tokens": 481335244.0, "step": 12680 }, { "epoch": 1.6131535428062587, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.681776523590088, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8767257332801819, "num_tokens": 481370894.0, "step": 12681 }, { "epoch": 1.6132807530848492, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.553987503051758, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8669091463088989, "num_tokens": 481413570.0, "step": 12682 }, { "epoch": 1.6134079633634397, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6804087162017822, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8742903470993042, "num_tokens": 481461380.0, "step": 12683 }, { "epoch": 1.6135351736420303, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5708463191986084, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8703216314315796, "num_tokens": 481497738.0, "step": 12684 }, { "epoch": 1.6136623839206208, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.671133041381836, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8826184868812561, "num_tokens": 481541000.0, "step": 12685 }, { "epoch": 1.6137895941992113, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5617055892944336, "learning_rate": 1e-05, "loss": 0.3799, "mean_token_accuracy": 0.8891654014587402, "num_tokens": 481578012.0, "step": 12686 }, { "epoch": 1.6139168044778018, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7005743980407715, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8644134998321533, "num_tokens": 481612996.0, "step": 12687 }, { "epoch": 1.6140440147563924, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5693163871765137, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8724559545516968, "num_tokens": 481645570.0, "step": 12688 }, { "epoch": 1.614171225034983, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6775944232940674, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8794030547142029, "num_tokens": 481687873.0, "step": 12689 }, { "epoch": 1.6142984353135734, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.567444324493408, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8627486228942871, "num_tokens": 481728105.0, "step": 12690 }, { "epoch": 1.614425645592164, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6880459785461426, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8805800080299377, "num_tokens": 481763558.0, "step": 12691 }, { "epoch": 1.6145528558707545, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5662384033203125, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8584936261177063, "num_tokens": 481801153.0, "step": 12692 }, { "epoch": 1.614680066149345, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.696962833404541, "learning_rate": 1e-05, "loss": 0.3794, "mean_token_accuracy": 0.8892625570297241, "num_tokens": 481830523.0, "step": 12693 }, { "epoch": 1.6148072764279355, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5439236164093018, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8693971633911133, "num_tokens": 481870814.0, "step": 12694 }, { "epoch": 1.6149344867065258, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.709322690963745, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8739674091339111, "num_tokens": 481912279.0, "step": 12695 }, { "epoch": 1.6150616969851164, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.536921739578247, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8692383766174316, "num_tokens": 481957506.0, "step": 12696 }, { "epoch": 1.615188907263707, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7147765159606934, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8799754977226257, "num_tokens": 481996718.0, "step": 12697 }, { "epoch": 1.6153161175422974, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.541741371154785, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.870915412902832, "num_tokens": 482035891.0, "step": 12698 }, { "epoch": 1.615443327820888, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7455854415893555, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8728008270263672, "num_tokens": 482070381.0, "step": 12699 }, { "epoch": 1.6155705380994785, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.52337646484375, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8755283355712891, "num_tokens": 482114484.0, "step": 12700 }, { "epoch": 1.6156977483780688, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.77677321434021, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8716698288917542, "num_tokens": 482147978.0, "step": 12701 }, { "epoch": 1.6158249586566593, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5224153995513916, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8776657581329346, "num_tokens": 482193017.0, "step": 12702 }, { "epoch": 1.6159521689352498, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.801448345184326, "learning_rate": 1e-05, "loss": 0.4877, "mean_token_accuracy": 0.8565760850906372, "num_tokens": 482230856.0, "step": 12703 }, { "epoch": 1.6160793792138404, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5271737575531006, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8826064467430115, "num_tokens": 482267540.0, "step": 12704 }, { "epoch": 1.616206589492431, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.8138744831085205, "learning_rate": 1e-05, "loss": 0.4014, "mean_token_accuracy": 0.8824668526649475, "num_tokens": 482302256.0, "step": 12705 }, { "epoch": 1.6163337997710214, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5209758281707764, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8782812356948853, "num_tokens": 482340706.0, "step": 12706 }, { "epoch": 1.616461010049612, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.836359739303589, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.880420446395874, "num_tokens": 482377313.0, "step": 12707 }, { "epoch": 1.6165882203282025, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.537565231323242, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8841433525085449, "num_tokens": 482415510.0, "step": 12708 }, { "epoch": 1.616715430606793, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.839531183242798, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8710169792175293, "num_tokens": 482453568.0, "step": 12709 }, { "epoch": 1.6168426408853835, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5670528411865234, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8703911304473877, "num_tokens": 482487024.0, "step": 12710 }, { "epoch": 1.616969851163974, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.831380844116211, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.880561113357544, "num_tokens": 482524328.0, "step": 12711 }, { "epoch": 1.6170970614425646, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5745761394500732, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.8814271092414856, "num_tokens": 482557042.0, "step": 12712 }, { "epoch": 1.6172242717211551, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8291940689086914, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.8506518602371216, "num_tokens": 482594486.0, "step": 12713 }, { "epoch": 1.6173514819997457, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.574856758117676, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8720115423202515, "num_tokens": 482635335.0, "step": 12714 }, { "epoch": 1.6174786922783362, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.793910026550293, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8827071189880371, "num_tokens": 482678945.0, "step": 12715 }, { "epoch": 1.6176059025569267, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.574026107788086, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.87333744764328, "num_tokens": 482717864.0, "step": 12716 }, { "epoch": 1.6177331128355172, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7818989753723145, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8688490986824036, "num_tokens": 482756361.0, "step": 12717 }, { "epoch": 1.6178603231141078, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.58311128616333, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8765095472335815, "num_tokens": 482794658.0, "step": 12718 }, { "epoch": 1.6179875333926983, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.787428617477417, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8687885403633118, "num_tokens": 482829426.0, "step": 12719 }, { "epoch": 1.6181147436712886, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.577150344848633, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8781259059906006, "num_tokens": 482867891.0, "step": 12720 }, { "epoch": 1.6182419539498791, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.769296407699585, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8774714469909668, "num_tokens": 482902307.0, "step": 12721 }, { "epoch": 1.6183691642284697, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.578399658203125, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8691449761390686, "num_tokens": 482939751.0, "step": 12722 }, { "epoch": 1.6184963745070602, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7588229179382324, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8659775257110596, "num_tokens": 482978899.0, "step": 12723 }, { "epoch": 1.6186235847856507, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5814273357391357, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8811614513397217, "num_tokens": 483011086.0, "step": 12724 }, { "epoch": 1.6187507950642412, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7243382930755615, "learning_rate": 1e-05, "loss": 0.3886, "mean_token_accuracy": 0.8867604732513428, "num_tokens": 483054235.0, "step": 12725 }, { "epoch": 1.6188780053428315, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5868139266967773, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8760793209075928, "num_tokens": 483086843.0, "step": 12726 }, { "epoch": 1.619005215621422, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7335052490234375, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8758747577667236, "num_tokens": 483119198.0, "step": 12727 }, { "epoch": 1.6191324259000126, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5660512447357178, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.882020890712738, "num_tokens": 483160835.0, "step": 12728 }, { "epoch": 1.6192596361786031, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7368130683898926, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8667188286781311, "num_tokens": 483194906.0, "step": 12729 }, { "epoch": 1.6193868464571937, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.566951036453247, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8665258884429932, "num_tokens": 483231731.0, "step": 12730 }, { "epoch": 1.6195140567357842, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.725536823272705, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8741767406463623, "num_tokens": 483270058.0, "step": 12731 }, { "epoch": 1.6196412670143747, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.565408229827881, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8854459524154663, "num_tokens": 483303600.0, "step": 12732 }, { "epoch": 1.6197684772929652, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7317252159118652, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8751193284988403, "num_tokens": 483339933.0, "step": 12733 }, { "epoch": 1.6198956875715558, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5595579147338867, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8791419267654419, "num_tokens": 483377153.0, "step": 12734 }, { "epoch": 1.6200228978501463, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7350707054138184, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8859498500823975, "num_tokens": 483415089.0, "step": 12735 }, { "epoch": 1.6201501081287368, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.555431604385376, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8709568977355957, "num_tokens": 483454023.0, "step": 12736 }, { "epoch": 1.6202773184073274, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7315051555633545, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8729504346847534, "num_tokens": 483493572.0, "step": 12737 }, { "epoch": 1.6204045286859179, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.564423084259033, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8682466149330139, "num_tokens": 483529817.0, "step": 12738 }, { "epoch": 1.6205317389645084, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7329280376434326, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8679203987121582, "num_tokens": 483570905.0, "step": 12739 }, { "epoch": 1.620658949243099, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.545318841934204, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.868901252746582, "num_tokens": 483618177.0, "step": 12740 }, { "epoch": 1.6207861595216895, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.751913070678711, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8748593926429749, "num_tokens": 483651227.0, "step": 12741 }, { "epoch": 1.62091336980028, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5554020404815674, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8777203559875488, "num_tokens": 483685577.0, "step": 12742 }, { "epoch": 1.6210405800788705, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7638955116271973, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8597459197044373, "num_tokens": 483721842.0, "step": 12743 }, { "epoch": 1.6211677903574608, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5596423149108887, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8669002652168274, "num_tokens": 483758646.0, "step": 12744 }, { "epoch": 1.6212950006360514, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7663967609405518, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.85956209897995, "num_tokens": 483789216.0, "step": 12745 }, { "epoch": 1.621422210914642, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5409250259399414, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8831084370613098, "num_tokens": 483832901.0, "step": 12746 }, { "epoch": 1.6215494211932324, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7680509090423584, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8670413494110107, "num_tokens": 483874331.0, "step": 12747 }, { "epoch": 1.621676631471823, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5520362854003906, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8763267993927002, "num_tokens": 483911608.0, "step": 12748 }, { "epoch": 1.6218038417504135, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7859599590301514, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8737631440162659, "num_tokens": 483946419.0, "step": 12749 }, { "epoch": 1.6219310520290038, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5457780361175537, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8611718416213989, "num_tokens": 483989740.0, "step": 12750 }, { "epoch": 1.6220582623075943, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7826287746429443, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8654460310935974, "num_tokens": 484025672.0, "step": 12751 }, { "epoch": 1.6221854725861848, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5594537258148193, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8749616742134094, "num_tokens": 484060421.0, "step": 12752 }, { "epoch": 1.6223126828647754, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7685844898223877, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8741835355758667, "num_tokens": 484102238.0, "step": 12753 }, { "epoch": 1.622439893143366, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5581552982330322, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8763689994812012, "num_tokens": 484138314.0, "step": 12754 }, { "epoch": 1.6225671034219564, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.771399736404419, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8532687425613403, "num_tokens": 484175541.0, "step": 12755 }, { "epoch": 1.622694313700547, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.570467472076416, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8851081132888794, "num_tokens": 484211276.0, "step": 12756 }, { "epoch": 1.6228215239791375, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.769796371459961, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8536454439163208, "num_tokens": 484250192.0, "step": 12757 }, { "epoch": 1.622948734257728, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5638427734375, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8804875612258911, "num_tokens": 484287330.0, "step": 12758 }, { "epoch": 1.6230759445363185, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7758355140686035, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8682735562324524, "num_tokens": 484319937.0, "step": 12759 }, { "epoch": 1.623203154814909, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.571946859359741, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8673089742660522, "num_tokens": 484353305.0, "step": 12760 }, { "epoch": 1.6233303650934996, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7629497051239014, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8591440916061401, "num_tokens": 484391970.0, "step": 12761 }, { "epoch": 1.6234575753720901, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.552206516265869, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8784686326980591, "num_tokens": 484433448.0, "step": 12762 }, { "epoch": 1.6235847856506807, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.767397403717041, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8777915835380554, "num_tokens": 484469729.0, "step": 12763 }, { "epoch": 1.6237119959292712, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5606706142425537, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8604881167411804, "num_tokens": 484509474.0, "step": 12764 }, { "epoch": 1.6238392062078617, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.758833885192871, "learning_rate": 1e-05, "loss": 0.3879, "mean_token_accuracy": 0.8896233439445496, "num_tokens": 484548999.0, "step": 12765 }, { "epoch": 1.6239664164864522, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5821738243103027, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8565856218338013, "num_tokens": 484580534.0, "step": 12766 }, { "epoch": 1.6240936267650428, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7433133125305176, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.8870736360549927, "num_tokens": 484616183.0, "step": 12767 }, { "epoch": 1.6242208370436333, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5676605701446533, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8707373142242432, "num_tokens": 484658290.0, "step": 12768 }, { "epoch": 1.6243480473222236, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7427656650543213, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.8842051029205322, "num_tokens": 484690226.0, "step": 12769 }, { "epoch": 1.6244752576008141, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5656721591949463, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8683240413665771, "num_tokens": 484731911.0, "step": 12770 }, { "epoch": 1.6246024678794047, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7475671768188477, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8735511302947998, "num_tokens": 484768648.0, "step": 12771 }, { "epoch": 1.6247296781579952, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.568464517593384, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8631758689880371, "num_tokens": 484807266.0, "step": 12772 }, { "epoch": 1.6248568884365857, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.734269857406616, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8611796498298645, "num_tokens": 484850877.0, "step": 12773 }, { "epoch": 1.6249840987151762, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5599379539489746, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.876737117767334, "num_tokens": 484891223.0, "step": 12774 }, { "epoch": 1.6251113089937665, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.752246379852295, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8769299387931824, "num_tokens": 484922081.0, "step": 12775 }, { "epoch": 1.625238519272357, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5659267902374268, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8613492250442505, "num_tokens": 484959016.0, "step": 12776 }, { "epoch": 1.6253657295509476, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.74446177482605, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8719574213027954, "num_tokens": 484998493.0, "step": 12777 }, { "epoch": 1.6254929398295381, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.571946382522583, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8737360239028931, "num_tokens": 485035164.0, "step": 12778 }, { "epoch": 1.6256201501081287, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.724588632583618, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8835204839706421, "num_tokens": 485076347.0, "step": 12779 }, { "epoch": 1.6257473603867192, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5755295753479004, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8747262358665466, "num_tokens": 485107940.0, "step": 12780 }, { "epoch": 1.6258745706653097, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7166311740875244, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8697141408920288, "num_tokens": 485145440.0, "step": 12781 }, { "epoch": 1.6260017809439002, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5738558769226074, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8771008253097534, "num_tokens": 485182948.0, "step": 12782 }, { "epoch": 1.6261289912224908, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7184643745422363, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8816566467285156, "num_tokens": 485220397.0, "step": 12783 }, { "epoch": 1.6262562015010813, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5529263019561768, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8741801977157593, "num_tokens": 485265382.0, "step": 12784 }, { "epoch": 1.6263834117796718, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.723242998123169, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8717337846755981, "num_tokens": 485306558.0, "step": 12785 }, { "epoch": 1.6265106220582624, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5718722343444824, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8696929216384888, "num_tokens": 485341278.0, "step": 12786 }, { "epoch": 1.6266378323368529, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7189464569091797, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.878229022026062, "num_tokens": 485376803.0, "step": 12787 }, { "epoch": 1.6267650426154434, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5791029930114746, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8607789278030396, "num_tokens": 485411168.0, "step": 12788 }, { "epoch": 1.626892252894034, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.702498197555542, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8638356328010559, "num_tokens": 485456320.0, "step": 12789 }, { "epoch": 1.6270194631726245, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.578658103942871, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8639901876449585, "num_tokens": 485491894.0, "step": 12790 }, { "epoch": 1.627146673451215, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7206854820251465, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.859473466873169, "num_tokens": 485526232.0, "step": 12791 }, { "epoch": 1.6272738837298055, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.570521593093872, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8681652545928955, "num_tokens": 485564752.0, "step": 12792 }, { "epoch": 1.6274010940083958, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.712660789489746, "learning_rate": 1e-05, "loss": 0.4156, "mean_token_accuracy": 0.8784664273262024, "num_tokens": 485600023.0, "step": 12793 }, { "epoch": 1.6275283042869864, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5635054111480713, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8732538819313049, "num_tokens": 485639195.0, "step": 12794 }, { "epoch": 1.6276555145655769, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.711947441101074, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.880418598651886, "num_tokens": 485676619.0, "step": 12795 }, { "epoch": 1.6277827248441674, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.569577932357788, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8578324913978577, "num_tokens": 485713739.0, "step": 12796 }, { "epoch": 1.627909935122758, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.718522071838379, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8695691227912903, "num_tokens": 485750010.0, "step": 12797 }, { "epoch": 1.6280371454013485, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5635268688201904, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8562731146812439, "num_tokens": 485790754.0, "step": 12798 }, { "epoch": 1.6281643556799388, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7281980514526367, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8767939209938049, "num_tokens": 485825472.0, "step": 12799 }, { "epoch": 1.6282915659585293, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.549647569656372, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8744564056396484, "num_tokens": 485866349.0, "step": 12800 }, { "epoch": 1.6284187762371198, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.734035015106201, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8731045722961426, "num_tokens": 485908323.0, "step": 12801 }, { "epoch": 1.6285459865157104, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5486011505126953, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8820414543151855, "num_tokens": 485948637.0, "step": 12802 }, { "epoch": 1.628673196794301, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7425742149353027, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8711575269699097, "num_tokens": 485988824.0, "step": 12803 }, { "epoch": 1.6288004070728914, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5493087768554688, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8761333227157593, "num_tokens": 486026364.0, "step": 12804 }, { "epoch": 1.628927617351482, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.741136312484741, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8687015771865845, "num_tokens": 486069578.0, "step": 12805 }, { "epoch": 1.6290548276300725, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5568063259124756, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8688521385192871, "num_tokens": 486105847.0, "step": 12806 }, { "epoch": 1.629182037908663, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7362253665924072, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8856956362724304, "num_tokens": 486146680.0, "step": 12807 }, { "epoch": 1.6293092481872535, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.550614595413208, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8797774910926819, "num_tokens": 486184461.0, "step": 12808 }, { "epoch": 1.629436458465844, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.768171787261963, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8815230131149292, "num_tokens": 486214084.0, "step": 12809 }, { "epoch": 1.6295636687444346, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5557093620300293, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8717876672744751, "num_tokens": 486250932.0, "step": 12810 }, { "epoch": 1.6296908790230251, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.753194570541382, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8788098096847534, "num_tokens": 486287964.0, "step": 12811 }, { "epoch": 1.6298180893016156, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5842292308807373, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8692818880081177, "num_tokens": 486321312.0, "step": 12812 }, { "epoch": 1.6299452995802062, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.732738971710205, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8720114827156067, "num_tokens": 486358992.0, "step": 12813 }, { "epoch": 1.6300725098587967, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.837340831756592, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8595929741859436, "num_tokens": 486399632.0, "step": 12814 }, { "epoch": 1.6301997201373872, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.60390043258667, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8729472756385803, "num_tokens": 486438129.0, "step": 12815 }, { "epoch": 1.6303269304159778, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.67899227142334, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8725563287734985, "num_tokens": 486476167.0, "step": 12816 }, { "epoch": 1.630454140694568, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.790855407714844e-05, "grad_norm": 2.5511043071746826, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8789935111999512, "num_tokens": 486512065.0, "step": 12817 }, { "epoch": 1.6305813509731586, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.702540636062622, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8563395142555237, "num_tokens": 486555723.0, "step": 12818 }, { "epoch": 1.6307085612517491, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.552903413772583, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8602964282035828, "num_tokens": 486598598.0, "step": 12819 }, { "epoch": 1.6308357715303397, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.72345232963562, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8694274425506592, "num_tokens": 486633426.0, "step": 12820 }, { "epoch": 1.6309629818089302, "ewc_loss": 0.064453125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.814697265625e-05, "grad_norm": 2.5310559272766113, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8758562803268433, "num_tokens": 486675584.0, "step": 12821 }, { "epoch": 1.6310901920875207, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7260539531707764, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8750783205032349, "num_tokens": 486722083.0, "step": 12822 }, { "epoch": 1.6312174023661112, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.53085994720459, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8809244632720947, "num_tokens": 486762658.0, "step": 12823 }, { "epoch": 1.6313446126447015, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.755885124206543, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8695837259292603, "num_tokens": 486799441.0, "step": 12824 }, { "epoch": 1.631471822923292, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.541973114013672, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.862280011177063, "num_tokens": 486836742.0, "step": 12825 }, { "epoch": 1.6315990332018826, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.779531955718994, "learning_rate": 1e-05, "loss": 0.514, "mean_token_accuracy": 0.8469317555427551, "num_tokens": 486873460.0, "step": 12826 }, { "epoch": 1.6317262434804731, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5323574542999268, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8727370500564575, "num_tokens": 486909411.0, "step": 12827 }, { "epoch": 1.6318534537590637, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7900872230529785, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8638814687728882, "num_tokens": 486948427.0, "step": 12828 }, { "epoch": 1.6319806640376542, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.54392671585083, "learning_rate": 1e-05, "loss": 0.4946, "mean_token_accuracy": 0.8548612594604492, "num_tokens": 486984594.0, "step": 12829 }, { "epoch": 1.6321078743162447, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7859320640563965, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8788737654685974, "num_tokens": 487028122.0, "step": 12830 }, { "epoch": 1.6322350845948352, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5451908111572266, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8619602918624878, "num_tokens": 487068731.0, "step": 12831 }, { "epoch": 1.6323622948734258, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7979025840759277, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8764445781707764, "num_tokens": 487103993.0, "step": 12832 }, { "epoch": 1.6324895051520163, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5631637573242188, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8704937100410461, "num_tokens": 487139206.0, "step": 12833 }, { "epoch": 1.6326167154306068, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.796341896057129, "learning_rate": 1e-05, "loss": 0.5407, "mean_token_accuracy": 0.8418619632720947, "num_tokens": 487181553.0, "step": 12834 }, { "epoch": 1.6327439257091974, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5642778873443604, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.868209958076477, "num_tokens": 487218931.0, "step": 12835 }, { "epoch": 1.6328711359877879, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7946956157684326, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8653513789176941, "num_tokens": 487252829.0, "step": 12836 }, { "epoch": 1.6329983462663784, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5644779205322266, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8786647915840149, "num_tokens": 487292098.0, "step": 12837 }, { "epoch": 1.633125556544969, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7859504222869873, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8732097744941711, "num_tokens": 487330639.0, "step": 12838 }, { "epoch": 1.6332527668235595, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5656237602233887, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8786566257476807, "num_tokens": 487368499.0, "step": 12839 }, { "epoch": 1.63337997710215, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7798233032226562, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8805024027824402, "num_tokens": 487406626.0, "step": 12840 }, { "epoch": 1.6335071873807405, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.577521562576294, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8741136789321899, "num_tokens": 487441055.0, "step": 12841 }, { "epoch": 1.6336343976593308, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.760502815246582, "learning_rate": 1e-05, "loss": 0.375, "mean_token_accuracy": 0.8930553197860718, "num_tokens": 487475973.0, "step": 12842 }, { "epoch": 1.6337616079379214, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5754551887512207, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8773386478424072, "num_tokens": 487513626.0, "step": 12843 }, { "epoch": 1.6338888182165119, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7610344886779785, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8811815977096558, "num_tokens": 487550053.0, "step": 12844 }, { "epoch": 1.6340160284951024, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.566650867462158, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8807933926582336, "num_tokens": 487586999.0, "step": 12845 }, { "epoch": 1.634143238773693, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7491841316223145, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8825522661209106, "num_tokens": 487631433.0, "step": 12846 }, { "epoch": 1.6342704490522835, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.59045147895813, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8755674362182617, "num_tokens": 487663259.0, "step": 12847 }, { "epoch": 1.6343976593308738, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.753296375274658, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8640841245651245, "num_tokens": 487697616.0, "step": 12848 }, { "epoch": 1.6345248696094643, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5711724758148193, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.882768452167511, "num_tokens": 487735693.0, "step": 12849 }, { "epoch": 1.6346520798880548, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.734365224838257, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8748645782470703, "num_tokens": 487775519.0, "step": 12850 }, { "epoch": 1.6347792901666454, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5709762573242188, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8638291954994202, "num_tokens": 487814058.0, "step": 12851 }, { "epoch": 1.6349065004452359, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7313332557678223, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8823000192642212, "num_tokens": 487850128.0, "step": 12852 }, { "epoch": 1.6350337107238264, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5639569759368896, "learning_rate": 1e-05, "loss": 0.3955, "mean_token_accuracy": 0.8864782452583313, "num_tokens": 487889465.0, "step": 12853 }, { "epoch": 1.635160921002417, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7421035766601562, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8668130040168762, "num_tokens": 487926264.0, "step": 12854 }, { "epoch": 1.6352881312810075, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.576536178588867, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8705572485923767, "num_tokens": 487959872.0, "step": 12855 }, { "epoch": 1.635415341559598, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7411301136016846, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8630796670913696, "num_tokens": 487997554.0, "step": 12856 }, { "epoch": 1.6355425518381885, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.567610502243042, "learning_rate": 1e-05, "loss": 0.391, "mean_token_accuracy": 0.8865005373954773, "num_tokens": 488029172.0, "step": 12857 }, { "epoch": 1.635669762116779, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7527949810028076, "learning_rate": 1e-05, "loss": 0.5055, "mean_token_accuracy": 0.8530281186103821, "num_tokens": 488066651.0, "step": 12858 }, { "epoch": 1.6357969723953696, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.567901372909546, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.862910807132721, "num_tokens": 488102207.0, "step": 12859 }, { "epoch": 1.6359241826739601, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.72745418548584, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8826183676719666, "num_tokens": 488142502.0, "step": 12860 }, { "epoch": 1.6360513929525506, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5759527683258057, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8587973117828369, "num_tokens": 488182334.0, "step": 12861 }, { "epoch": 1.6361786032311412, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.716243267059326, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8763419389724731, "num_tokens": 488219434.0, "step": 12862 }, { "epoch": 1.6363058135097317, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002651214599609375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5713369846343994, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8757244348526001, "num_tokens": 488256961.0, "step": 12863 }, { "epoch": 1.6364330237883222, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7194180488586426, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8659330606460571, "num_tokens": 488297069.0, "step": 12864 }, { "epoch": 1.6365602340669128, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5745654106140137, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8567179441452026, "num_tokens": 488336436.0, "step": 12865 }, { "epoch": 1.636687444345503, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.710263729095459, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8674513101577759, "num_tokens": 488376145.0, "step": 12866 }, { "epoch": 1.6368146546240936, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.566920042037964, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8742274045944214, "num_tokens": 488418319.0, "step": 12867 }, { "epoch": 1.6369418649026841, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7062411308288574, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8726134896278381, "num_tokens": 488460377.0, "step": 12868 }, { "epoch": 1.6370690751812746, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.595470666885376, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8653202056884766, "num_tokens": 488491794.0, "step": 12869 }, { "epoch": 1.6371962854598652, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6930992603302, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8647493124008179, "num_tokens": 488534106.0, "step": 12870 }, { "epoch": 1.6373234957384557, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5929388999938965, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8643717765808105, "num_tokens": 488568198.0, "step": 12871 }, { "epoch": 1.6374507060170462, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6789920330047607, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.868128776550293, "num_tokens": 488606112.0, "step": 12872 }, { "epoch": 1.6375779162956365, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.59114146232605, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8679922819137573, "num_tokens": 488639585.0, "step": 12873 }, { "epoch": 1.637705126574227, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6806726455688477, "learning_rate": 1e-05, "loss": 0.3901, "mean_token_accuracy": 0.8852874040603638, "num_tokens": 488670789.0, "step": 12874 }, { "epoch": 1.6378323368528176, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5788772106170654, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8716107606887817, "num_tokens": 488707837.0, "step": 12875 }, { "epoch": 1.6379595471314081, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6682329177856445, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8706615567207336, "num_tokens": 488747719.0, "step": 12876 }, { "epoch": 1.6380867574099987, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.582858085632324, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8669368624687195, "num_tokens": 488786493.0, "step": 12877 }, { "epoch": 1.6382139676885892, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6629323959350586, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8669683933258057, "num_tokens": 488826390.0, "step": 12878 }, { "epoch": 1.6383411779671797, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.588090658187866, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.869242787361145, "num_tokens": 488861458.0, "step": 12879 }, { "epoch": 1.6384683882457702, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.657261371612549, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8743061423301697, "num_tokens": 488902762.0, "step": 12880 }, { "epoch": 1.6385955985243608, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5837416648864746, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.8838130235671997, "num_tokens": 488938139.0, "step": 12881 }, { "epoch": 1.6387228088029513, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.672729969024658, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8749607801437378, "num_tokens": 488976036.0, "step": 12882 }, { "epoch": 1.6388500190815418, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5865581035614014, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8603662252426147, "num_tokens": 489014333.0, "step": 12883 }, { "epoch": 1.6389772293601323, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6740293502807617, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8746428489685059, "num_tokens": 489050060.0, "step": 12884 }, { "epoch": 1.6391044396387229, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.572314739227295, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8709642887115479, "num_tokens": 489090053.0, "step": 12885 }, { "epoch": 1.6392316499173134, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.682957649230957, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.870002031326294, "num_tokens": 489129081.0, "step": 12886 }, { "epoch": 1.639358860195904, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5600008964538574, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8623008728027344, "num_tokens": 489174811.0, "step": 12887 }, { "epoch": 1.6394860704744945, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6947999000549316, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8715459108352661, "num_tokens": 489214694.0, "step": 12888 }, { "epoch": 1.639613280753085, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.556581497192383, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8755983710289001, "num_tokens": 489255299.0, "step": 12889 }, { "epoch": 1.6397404910316755, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7019097805023193, "learning_rate": 1e-05, "loss": 0.383, "mean_token_accuracy": 0.8889395594596863, "num_tokens": 489292409.0, "step": 12890 }, { "epoch": 1.6398677013102658, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.556497812271118, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8616251349449158, "num_tokens": 489335354.0, "step": 12891 }, { "epoch": 1.6399949115888564, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.742687463760376, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8510172367095947, "num_tokens": 489369253.0, "step": 12892 }, { "epoch": 1.6401221218674469, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5366616249084473, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8718441724777222, "num_tokens": 489408453.0, "step": 12893 }, { "epoch": 1.6402493321460374, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7475554943084717, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.8834735155105591, "num_tokens": 489441609.0, "step": 12894 }, { "epoch": 1.640376542424628, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5365657806396484, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.872830867767334, "num_tokens": 489477228.0, "step": 12895 }, { "epoch": 1.6405037527032185, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.772412061691284, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8594186902046204, "num_tokens": 489518622.0, "step": 12896 }, { "epoch": 1.6406309629818088, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.530073881149292, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8608556985855103, "num_tokens": 489557707.0, "step": 12897 }, { "epoch": 1.6407581732603993, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7957639694213867, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8663750290870667, "num_tokens": 489591180.0, "step": 12898 }, { "epoch": 1.6408853835389898, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5298948287963867, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8711175322532654, "num_tokens": 489631732.0, "step": 12899 }, { "epoch": 1.6410125938175804, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.801754951477051, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8683069348335266, "num_tokens": 489671484.0, "step": 12900 }, { "epoch": 1.6411398040961709, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.556309938430786, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8718402981758118, "num_tokens": 489706929.0, "step": 12901 }, { "epoch": 1.6412670143747614, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8116493225097656, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8704995512962341, "num_tokens": 489741830.0, "step": 12902 }, { "epoch": 1.641394224653352, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5407285690307617, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8821706175804138, "num_tokens": 489782808.0, "step": 12903 }, { "epoch": 1.6415214349319425, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.815112590789795, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8636429905891418, "num_tokens": 489823536.0, "step": 12904 }, { "epoch": 1.641648645210533, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.558701753616333, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.875417947769165, "num_tokens": 489854730.0, "step": 12905 }, { "epoch": 1.6417758554891235, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8217215538024902, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8728175163269043, "num_tokens": 489886836.0, "step": 12906 }, { "epoch": 1.641903065767714, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.555671215057373, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8737844824790955, "num_tokens": 489928227.0, "step": 12907 }, { "epoch": 1.6420302760463046, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8213582038879395, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8582700490951538, "num_tokens": 489965697.0, "step": 12908 }, { "epoch": 1.642157486324895, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.564868927001953, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.878582775592804, "num_tokens": 490000344.0, "step": 12909 }, { "epoch": 1.6422846966034856, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.826629161834717, "learning_rate": 1e-05, "loss": 0.3792, "mean_token_accuracy": 0.892970860004425, "num_tokens": 490030965.0, "step": 12910 }, { "epoch": 1.6424119068820762, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.567248821258545, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8798370957374573, "num_tokens": 490069386.0, "step": 12911 }, { "epoch": 1.6425391171606667, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.816072940826416, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8811078667640686, "num_tokens": 490104040.0, "step": 12912 }, { "epoch": 1.6426663274392572, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5883893966674805, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8635885119438171, "num_tokens": 490147883.0, "step": 12913 }, { "epoch": 1.6427935377178478, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.795123338699341, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8726630806922913, "num_tokens": 490183889.0, "step": 12914 }, { "epoch": 1.642920747996438, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.582916259765625, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8748021125793457, "num_tokens": 490224132.0, "step": 12915 }, { "epoch": 1.6430479582750286, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7866697311401367, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8702273368835449, "num_tokens": 490262745.0, "step": 12916 }, { "epoch": 1.6431751685536191, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5818727016448975, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8671784400939941, "num_tokens": 490301449.0, "step": 12917 }, { "epoch": 1.6433023788322096, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.8037877082824707, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8600447177886963, "num_tokens": 490328873.0, "step": 12918 }, { "epoch": 1.6434295891108002, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5929038524627686, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8754929304122925, "num_tokens": 490361174.0, "step": 12919 }, { "epoch": 1.6435567993893907, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.764873743057251, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8643542528152466, "num_tokens": 490399708.0, "step": 12920 }, { "epoch": 1.6436840096679812, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5828421115875244, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8851603269577026, "num_tokens": 490433159.0, "step": 12921 }, { "epoch": 1.6438112199465715, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.749431848526001, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8740428686141968, "num_tokens": 490473335.0, "step": 12922 }, { "epoch": 1.643938430225162, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.574984073638916, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8634078502655029, "num_tokens": 490517091.0, "step": 12923 }, { "epoch": 1.6440656405037526, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.750662088394165, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8690618276596069, "num_tokens": 490550336.0, "step": 12924 }, { "epoch": 1.6441928507823431, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.561178207397461, "learning_rate": 1e-05, "loss": 0.389, "mean_token_accuracy": 0.8881720900535583, "num_tokens": 490593592.0, "step": 12925 }, { "epoch": 1.6443200610609336, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.751568555831909, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8819567561149597, "num_tokens": 490629354.0, "step": 12926 }, { "epoch": 1.6444472713395242, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.568470001220703, "learning_rate": 1e-05, "loss": 0.3666, "mean_token_accuracy": 0.8948719501495361, "num_tokens": 490668319.0, "step": 12927 }, { "epoch": 1.6445744816181147, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7594854831695557, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8793588280677795, "num_tokens": 490705978.0, "step": 12928 }, { "epoch": 1.6447016918967052, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.568946123123169, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8729007244110107, "num_tokens": 490742455.0, "step": 12929 }, { "epoch": 1.6448289021752958, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.760436773300171, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8575537800788879, "num_tokens": 490781733.0, "step": 12930 }, { "epoch": 1.6449561124538863, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.563002347946167, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8737549781799316, "num_tokens": 490821077.0, "step": 12931 }, { "epoch": 1.6450833227324768, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7377066612243652, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8766180276870728, "num_tokens": 490862725.0, "step": 12932 }, { "epoch": 1.6452105330110673, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.568312644958496, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8588746786117554, "num_tokens": 490909612.0, "step": 12933 }, { "epoch": 1.6453377432896579, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.744392156600952, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8739416599273682, "num_tokens": 490943202.0, "step": 12934 }, { "epoch": 1.6454649535682484, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5777134895324707, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8699908256530762, "num_tokens": 490978481.0, "step": 12935 }, { "epoch": 1.645592163846839, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.728337526321411, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8756438493728638, "num_tokens": 491018672.0, "step": 12936 }, { "epoch": 1.6457193741254295, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5645179748535156, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.876161515712738, "num_tokens": 491055032.0, "step": 12937 }, { "epoch": 1.64584658440402, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7500498294830322, "learning_rate": 1e-05, "loss": 0.5138, "mean_token_accuracy": 0.8465150594711304, "num_tokens": 491089780.0, "step": 12938 }, { "epoch": 1.6459737946826105, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.571486234664917, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8762248158454895, "num_tokens": 491124187.0, "step": 12939 }, { "epoch": 1.6461010049612008, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7222657203674316, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8692961931228638, "num_tokens": 491167558.0, "step": 12940 }, { "epoch": 1.6462282152397913, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.565538167953491, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8677282333374023, "num_tokens": 491210054.0, "step": 12941 }, { "epoch": 1.6463554255183819, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7189483642578125, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8774232864379883, "num_tokens": 491256535.0, "step": 12942 }, { "epoch": 1.6464826357969724, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5563125610351562, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8784428238868713, "num_tokens": 491295909.0, "step": 12943 }, { "epoch": 1.646609846075563, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.737192392349243, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8608430027961731, "num_tokens": 491332547.0, "step": 12944 }, { "epoch": 1.6467370563541535, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5565216541290283, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.8877930641174316, "num_tokens": 491373187.0, "step": 12945 }, { "epoch": 1.6468642666327438, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.745185136795044, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.872052788734436, "num_tokens": 491412476.0, "step": 12946 }, { "epoch": 1.6469914769113343, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5527637004852295, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8790971636772156, "num_tokens": 491456502.0, "step": 12947 }, { "epoch": 1.6471186871899248, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7509381771087646, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8778657913208008, "num_tokens": 491492974.0, "step": 12948 }, { "epoch": 1.6472458974685154, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5671656131744385, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8730714321136475, "num_tokens": 491523621.0, "step": 12949 }, { "epoch": 1.6473731077471059, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7521467208862305, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8609265089035034, "num_tokens": 491563366.0, "step": 12950 }, { "epoch": 1.6475003180256964, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.572585344314575, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8712906837463379, "num_tokens": 491594221.0, "step": 12951 }, { "epoch": 1.647627528304287, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7445051670074463, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8629970550537109, "num_tokens": 491632039.0, "step": 12952 }, { "epoch": 1.6477547385828775, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5551235675811768, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8780251145362854, "num_tokens": 491672282.0, "step": 12953 }, { "epoch": 1.647881948861468, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.755826473236084, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8596099019050598, "num_tokens": 491707969.0, "step": 12954 }, { "epoch": 1.6480091591400585, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.563642740249634, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8775929808616638, "num_tokens": 491743347.0, "step": 12955 }, { "epoch": 1.648136369418649, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.768827438354492, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.870847761631012, "num_tokens": 491777165.0, "step": 12956 }, { "epoch": 1.6482635796972396, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5481650829315186, "learning_rate": 1e-05, "loss": 0.3894, "mean_token_accuracy": 0.8857030868530273, "num_tokens": 491813859.0, "step": 12957 }, { "epoch": 1.64839078997583, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7697885036468506, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8658551573753357, "num_tokens": 491853373.0, "step": 12958 }, { "epoch": 1.6485180002544206, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.54530668258667, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8761550188064575, "num_tokens": 491889401.0, "step": 12959 }, { "epoch": 1.6486452105330112, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7745234966278076, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8677724599838257, "num_tokens": 491933973.0, "step": 12960 }, { "epoch": 1.6487724208116017, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5428669452667236, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8644146919250488, "num_tokens": 491979198.0, "step": 12961 }, { "epoch": 1.6488996310901922, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.810159206390381, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.860236406326294, "num_tokens": 492009127.0, "step": 12962 }, { "epoch": 1.6490268413687827, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.552140235900879, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.872438371181488, "num_tokens": 492045022.0, "step": 12963 }, { "epoch": 1.649154051647373, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7852303981781006, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8699722290039062, "num_tokens": 492092155.0, "step": 12964 }, { "epoch": 1.6492812619259636, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5582313537597656, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8546789884567261, "num_tokens": 492131807.0, "step": 12965 }, { "epoch": 1.649408472204554, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.791835308074951, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8732351064682007, "num_tokens": 492166794.0, "step": 12966 }, { "epoch": 1.6495356824831446, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.554687023162842, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8773034811019897, "num_tokens": 492208375.0, "step": 12967 }, { "epoch": 1.6496628927617352, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.8106088638305664, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8600678443908691, "num_tokens": 492241166.0, "step": 12968 }, { "epoch": 1.6497901030403257, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5520472526550293, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.8859702348709106, "num_tokens": 492278887.0, "step": 12969 }, { "epoch": 1.6499173133189162, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.796135425567627, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8880584836006165, "num_tokens": 492316523.0, "step": 12970 }, { "epoch": 1.6500445235975065, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.566452980041504, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8659850358963013, "num_tokens": 492354094.0, "step": 12971 }, { "epoch": 1.650171733876097, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.812823534011841, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8633838295936584, "num_tokens": 492387250.0, "step": 12972 }, { "epoch": 1.6502989441546876, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.570554494857788, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8689880967140198, "num_tokens": 492423770.0, "step": 12973 }, { "epoch": 1.6504261544332781, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7886478900909424, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8694946765899658, "num_tokens": 492465970.0, "step": 12974 }, { "epoch": 1.6505533647118686, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.582019090652466, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8646020889282227, "num_tokens": 492501668.0, "step": 12975 }, { "epoch": 1.6506805749904592, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7690536975860596, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8678946495056152, "num_tokens": 492541032.0, "step": 12976 }, { "epoch": 1.6508077852690497, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.588135004043579, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8676249980926514, "num_tokens": 492579639.0, "step": 12977 }, { "epoch": 1.6509349955476402, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7710700035095215, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8696497678756714, "num_tokens": 492612324.0, "step": 12978 }, { "epoch": 1.6510622058262308, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5910794734954834, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8646396398544312, "num_tokens": 492648967.0, "step": 12979 }, { "epoch": 1.6511894161048213, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.737670660018921, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8853992223739624, "num_tokens": 492687117.0, "step": 12980 }, { "epoch": 1.6513166263834118, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5972542762756348, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8734921216964722, "num_tokens": 492718760.0, "step": 12981 }, { "epoch": 1.6514438366620023, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7202157974243164, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8749688863754272, "num_tokens": 492760038.0, "step": 12982 }, { "epoch": 1.6515710469405929, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5886428356170654, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8616983294487, "num_tokens": 492798450.0, "step": 12983 }, { "epoch": 1.6516982572191834, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7139930725097656, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8746480941772461, "num_tokens": 492835081.0, "step": 12984 }, { "epoch": 1.651825467497774, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.580763101577759, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.879989743232727, "num_tokens": 492878255.0, "step": 12985 }, { "epoch": 1.6519526777763645, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7030422687530518, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8611448407173157, "num_tokens": 492920800.0, "step": 12986 }, { "epoch": 1.652079888054955, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5914714336395264, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8696930408477783, "num_tokens": 492957170.0, "step": 12987 }, { "epoch": 1.6522070983335455, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.689227342605591, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8743525743484497, "num_tokens": 492995123.0, "step": 12988 }, { "epoch": 1.6523343086121358, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5887575149536133, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8756129741668701, "num_tokens": 493034538.0, "step": 12989 }, { "epoch": 1.6524615188907263, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.686976671218872, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8588906526565552, "num_tokens": 493073218.0, "step": 12990 }, { "epoch": 1.6525887291693169, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.577291488647461, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8689572811126709, "num_tokens": 493110056.0, "step": 12991 }, { "epoch": 1.6527159394479074, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7058398723602295, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8785974979400635, "num_tokens": 493144324.0, "step": 12992 }, { "epoch": 1.652843149726498, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.568955898284912, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8635080456733704, "num_tokens": 493181685.0, "step": 12993 }, { "epoch": 1.6529703600050885, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7025129795074463, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8790889382362366, "num_tokens": 493217712.0, "step": 12994 }, { "epoch": 1.6530975702836788, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5762550830841064, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8577847480773926, "num_tokens": 493253852.0, "step": 12995 }, { "epoch": 1.6532247805622693, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.706902027130127, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8724259734153748, "num_tokens": 493287797.0, "step": 12996 }, { "epoch": 1.6533519908408598, "ewc_loss": 0.06494140625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.559760093688965, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8742067813873291, "num_tokens": 493326072.0, "step": 12997 }, { "epoch": 1.6534792011194503, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7161061763763428, "learning_rate": 1e-05, "loss": 0.4916, "mean_token_accuracy": 0.8545227646827698, "num_tokens": 493365374.0, "step": 12998 }, { "epoch": 1.6536064113980409, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.54390025138855, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8884570002555847, "num_tokens": 493408526.0, "step": 12999 }, { "epoch": 1.6537336216766314, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.749626398086548, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8609820604324341, "num_tokens": 493445689.0, "step": 13000 }, { "epoch": 1.653860831955222, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.559129476547241, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8628138899803162, "num_tokens": 493480884.0, "step": 13001 }, { "epoch": 1.6539880422338125, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7365059852600098, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8725889325141907, "num_tokens": 493523917.0, "step": 13002 }, { "epoch": 1.654115252512403, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.550442695617676, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8703566193580627, "num_tokens": 493560438.0, "step": 13003 }, { "epoch": 1.6542424627909935, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.768857955932617, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8606995940208435, "num_tokens": 493596609.0, "step": 13004 }, { "epoch": 1.654369673069584, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.56064510345459, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8656438589096069, "num_tokens": 493634369.0, "step": 13005 }, { "epoch": 1.6544968833481746, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.764765739440918, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8675037622451782, "num_tokens": 493669568.0, "step": 13006 }, { "epoch": 1.654624093626765, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5431644916534424, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8762034177780151, "num_tokens": 493710136.0, "step": 13007 }, { "epoch": 1.6547513039053556, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7652320861816406, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8681213855743408, "num_tokens": 493750028.0, "step": 13008 }, { "epoch": 1.6548785141839462, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5651307106018066, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8611729145050049, "num_tokens": 493785167.0, "step": 13009 }, { "epoch": 1.6550057244625367, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.766263008117676, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8602076768875122, "num_tokens": 493822724.0, "step": 13010 }, { "epoch": 1.6551329347411272, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5582833290100098, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8630909323692322, "num_tokens": 493863910.0, "step": 13011 }, { "epoch": 1.6552601450197177, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.753305196762085, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8799795508384705, "num_tokens": 493901863.0, "step": 13012 }, { "epoch": 1.655387355298308, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.571269989013672, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8744416236877441, "num_tokens": 493938174.0, "step": 13013 }, { "epoch": 1.6555145655768986, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7599546909332275, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8701145648956299, "num_tokens": 493973483.0, "step": 13014 }, { "epoch": 1.655641775855489, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5545568466186523, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8842767477035522, "num_tokens": 494016481.0, "step": 13015 }, { "epoch": 1.6557689861340796, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.773407220840454, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8732407093048096, "num_tokens": 494047543.0, "step": 13016 }, { "epoch": 1.6558961964126702, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.567279815673828, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8697797060012817, "num_tokens": 494082076.0, "step": 13017 }, { "epoch": 1.6560234066912607, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.761181592941284, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8678429126739502, "num_tokens": 494123822.0, "step": 13018 }, { "epoch": 1.6561506169698512, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5747153759002686, "learning_rate": 1e-05, "loss": 0.3893, "mean_token_accuracy": 0.8860174417495728, "num_tokens": 494152644.0, "step": 13019 }, { "epoch": 1.6562778272484415, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.779135227203369, "learning_rate": 1e-05, "loss": 0.5024, "mean_token_accuracy": 0.8555874824523926, "num_tokens": 494185667.0, "step": 13020 }, { "epoch": 1.656405037527032, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.559307336807251, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8802341222763062, "num_tokens": 494224277.0, "step": 13021 }, { "epoch": 1.6565322478056226, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7879528999328613, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8650332093238831, "num_tokens": 494254475.0, "step": 13022 }, { "epoch": 1.656659458084213, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5592029094696045, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8709375262260437, "num_tokens": 494291891.0, "step": 13023 }, { "epoch": 1.6567866683628036, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7860822677612305, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8643972873687744, "num_tokens": 494327659.0, "step": 13024 }, { "epoch": 1.6569138786413942, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.559920072555542, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.8875163197517395, "num_tokens": 494358149.0, "step": 13025 }, { "epoch": 1.6570410889199847, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7862606048583984, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8773733973503113, "num_tokens": 494393673.0, "step": 13026 }, { "epoch": 1.6571682991985752, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.573493719100952, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8737959861755371, "num_tokens": 494422811.0, "step": 13027 }, { "epoch": 1.6572955094771658, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.768157482147217, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8716219067573547, "num_tokens": 494467758.0, "step": 13028 }, { "epoch": 1.6574227197557563, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5616722106933594, "learning_rate": 1e-05, "loss": 0.3876, "mean_token_accuracy": 0.8862985968589783, "num_tokens": 494504698.0, "step": 13029 }, { "epoch": 1.6575499300343468, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.77536678314209, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8617287874221802, "num_tokens": 494541235.0, "step": 13030 }, { "epoch": 1.6576771403129373, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5530905723571777, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8715381622314453, "num_tokens": 494584499.0, "step": 13031 }, { "epoch": 1.6578043505915279, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7664108276367188, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8700319528579712, "num_tokens": 494625845.0, "step": 13032 }, { "epoch": 1.6579315608701184, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.57541823387146, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8751069903373718, "num_tokens": 494660820.0, "step": 13033 }, { "epoch": 1.658058771148709, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.744983434677124, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8855122327804565, "num_tokens": 494701215.0, "step": 13034 }, { "epoch": 1.6581859814272994, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.589341163635254, "learning_rate": 1e-05, "loss": 0.5284, "mean_token_accuracy": 0.8434354066848755, "num_tokens": 494739062.0, "step": 13035 }, { "epoch": 1.65831319170589, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7226099967956543, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8702371716499329, "num_tokens": 494779467.0, "step": 13036 }, { "epoch": 1.6584404019844805, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.581397294998169, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8714191913604736, "num_tokens": 494817237.0, "step": 13037 }, { "epoch": 1.6585676122630708, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7189152240753174, "learning_rate": 1e-05, "loss": 0.3863, "mean_token_accuracy": 0.888306200504303, "num_tokens": 494853485.0, "step": 13038 }, { "epoch": 1.6586948225416613, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.575150966644287, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8766432404518127, "num_tokens": 494890456.0, "step": 13039 }, { "epoch": 1.6588220328202519, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.721933364868164, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.859721839427948, "num_tokens": 494929441.0, "step": 13040 }, { "epoch": 1.6589492430988424, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5837860107421875, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8629079461097717, "num_tokens": 494963642.0, "step": 13041 }, { "epoch": 1.659076453377433, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7319931983947754, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8624345064163208, "num_tokens": 494998138.0, "step": 13042 }, { "epoch": 1.6592036636560235, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.569173574447632, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8730331063270569, "num_tokens": 495038080.0, "step": 13043 }, { "epoch": 1.6593308739346138, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7171854972839355, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8664908409118652, "num_tokens": 495079103.0, "step": 13044 }, { "epoch": 1.6594580842132043, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5597875118255615, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8727459907531738, "num_tokens": 495119385.0, "step": 13045 }, { "epoch": 1.6595852944917948, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7309467792510986, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8646368384361267, "num_tokens": 495157829.0, "step": 13046 }, { "epoch": 1.6597125047703853, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.564631462097168, "learning_rate": 1e-05, "loss": 0.5217, "mean_token_accuracy": 0.846713662147522, "num_tokens": 495198554.0, "step": 13047 }, { "epoch": 1.6598397150489759, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.714567184448242, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8784295320510864, "num_tokens": 495241389.0, "step": 13048 }, { "epoch": 1.6599669253275664, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5615532398223877, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8829352855682373, "num_tokens": 495274216.0, "step": 13049 }, { "epoch": 1.660094135606157, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7391414642333984, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8647180199623108, "num_tokens": 495308941.0, "step": 13050 }, { "epoch": 1.6602213458847475, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5485169887542725, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8738201856613159, "num_tokens": 495351668.0, "step": 13051 }, { "epoch": 1.660348556163338, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.755080461502075, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8630294799804688, "num_tokens": 495383402.0, "step": 13052 }, { "epoch": 1.6604757664419285, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5438857078552246, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8822488784790039, "num_tokens": 495421371.0, "step": 13053 }, { "epoch": 1.660602976720519, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7592544555664062, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8659584522247314, "num_tokens": 495460692.0, "step": 13054 }, { "epoch": 1.6607301869991096, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.552551507949829, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8754427433013916, "num_tokens": 495493990.0, "step": 13055 }, { "epoch": 1.6608573972777, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7662854194641113, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8687350749969482, "num_tokens": 495531616.0, "step": 13056 }, { "epoch": 1.6609846075562906, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5458014011383057, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8625071048736572, "num_tokens": 495572065.0, "step": 13057 }, { "epoch": 1.6611118178348812, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7729783058166504, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8551872372627258, "num_tokens": 495612582.0, "step": 13058 }, { "epoch": 1.6612390281134717, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.556230068206787, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8688833713531494, "num_tokens": 495649211.0, "step": 13059 }, { "epoch": 1.6613662383920622, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.763592481613159, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8701531887054443, "num_tokens": 495689805.0, "step": 13060 }, { "epoch": 1.6614934486706527, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.6074118614196777, "learning_rate": 1e-05, "loss": 0.5223, "mean_token_accuracy": 0.8472748398780823, "num_tokens": 495727560.0, "step": 13061 }, { "epoch": 1.661620658949243, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7414987087249756, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.871830403804779, "num_tokens": 495759827.0, "step": 13062 }, { "epoch": 1.6617478692278336, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.571938991546631, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8728843927383423, "num_tokens": 495797106.0, "step": 13063 }, { "epoch": 1.661875079506424, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.720614194869995, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8659044504165649, "num_tokens": 495842331.0, "step": 13064 }, { "epoch": 1.6620022897850146, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5732781887054443, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8666119575500488, "num_tokens": 495883491.0, "step": 13065 }, { "epoch": 1.6621295000636052, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7277519702911377, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8595277070999146, "num_tokens": 495924536.0, "step": 13066 }, { "epoch": 1.6622567103421957, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.578387975692749, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8597451448440552, "num_tokens": 495961783.0, "step": 13067 }, { "epoch": 1.662383920620786, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7337584495544434, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8728306889533997, "num_tokens": 495994131.0, "step": 13068 }, { "epoch": 1.6625111308993765, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.570493459701538, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.868432343006134, "num_tokens": 496035981.0, "step": 13069 }, { "epoch": 1.662638341177967, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.740140199661255, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8591404557228088, "num_tokens": 496075622.0, "step": 13070 }, { "epoch": 1.6627655514565576, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.554091691970825, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8697139620780945, "num_tokens": 496117130.0, "step": 13071 }, { "epoch": 1.662892761735148, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.747419595718384, "learning_rate": 1e-05, "loss": 0.3815, "mean_token_accuracy": 0.8934513330459595, "num_tokens": 496153309.0, "step": 13072 }, { "epoch": 1.6630199720137386, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5486037731170654, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.883808970451355, "num_tokens": 496194955.0, "step": 13073 }, { "epoch": 1.6631471822923292, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.747687578201294, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8688000440597534, "num_tokens": 496238787.0, "step": 13074 }, { "epoch": 1.6632743925709197, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5539658069610596, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8610502481460571, "num_tokens": 496279606.0, "step": 13075 }, { "epoch": 1.6634016028495102, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7588088512420654, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8767086267471313, "num_tokens": 496315724.0, "step": 13076 }, { "epoch": 1.6635288131281007, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5459468364715576, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8745489716529846, "num_tokens": 496357254.0, "step": 13077 }, { "epoch": 1.6636560234066913, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.772189140319824, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8740867376327515, "num_tokens": 496394356.0, "step": 13078 }, { "epoch": 1.6637832336852818, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.546994686126709, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8675537705421448, "num_tokens": 496431569.0, "step": 13079 }, { "epoch": 1.6639104439638723, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7931599617004395, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8748014569282532, "num_tokens": 496465854.0, "step": 13080 }, { "epoch": 1.6640376542424629, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5370357036590576, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8733267784118652, "num_tokens": 496506023.0, "step": 13081 }, { "epoch": 1.6641648645210534, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8011927604675293, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.8594728708267212, "num_tokens": 496545950.0, "step": 13082 }, { "epoch": 1.664292074799644, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5475175380706787, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8754373788833618, "num_tokens": 496585728.0, "step": 13083 }, { "epoch": 1.6644192850782344, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8049137592315674, "learning_rate": 1e-05, "loss": 0.5088, "mean_token_accuracy": 0.8531988263130188, "num_tokens": 496628491.0, "step": 13084 }, { "epoch": 1.664546495356825, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5577375888824463, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8634998798370361, "num_tokens": 496665952.0, "step": 13085 }, { "epoch": 1.6646737056354155, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7945187091827393, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.881244957447052, "num_tokens": 496708188.0, "step": 13086 }, { "epoch": 1.6648009159140058, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.569797992706299, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8797506093978882, "num_tokens": 496743079.0, "step": 13087 }, { "epoch": 1.6649281261925963, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.789130449295044, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8702468276023865, "num_tokens": 496781364.0, "step": 13088 }, { "epoch": 1.6650553364711869, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5701470375061035, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8832793235778809, "num_tokens": 496816419.0, "step": 13089 }, { "epoch": 1.6651825467497774, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7760283946990967, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8767292499542236, "num_tokens": 496851941.0, "step": 13090 }, { "epoch": 1.665309757028368, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.00026702880859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5830323696136475, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8760210275650024, "num_tokens": 496884974.0, "step": 13091 }, { "epoch": 1.6654369673069584, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.77881121635437, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8772569298744202, "num_tokens": 496920331.0, "step": 13092 }, { "epoch": 1.6655641775855488, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5764970779418945, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8676433563232422, "num_tokens": 496957599.0, "step": 13093 }, { "epoch": 1.6656913878641393, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.752488851547241, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8837421536445618, "num_tokens": 496992804.0, "step": 13094 }, { "epoch": 1.6658185981427298, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.579397201538086, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8847007751464844, "num_tokens": 497025376.0, "step": 13095 }, { "epoch": 1.6659458084213203, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7408361434936523, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8789238929748535, "num_tokens": 497067923.0, "step": 13096 }, { "epoch": 1.6660730186999109, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.57358980178833, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8616738319396973, "num_tokens": 497113753.0, "step": 13097 }, { "epoch": 1.6662002289785014, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7509069442749023, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8740006685256958, "num_tokens": 497148001.0, "step": 13098 }, { "epoch": 1.666327439257092, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.591139793395996, "learning_rate": 1e-05, "loss": 0.4957, "mean_token_accuracy": 0.8526468276977539, "num_tokens": 497186702.0, "step": 13099 }, { "epoch": 1.6664546495356825, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7320218086242676, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8715652823448181, "num_tokens": 497224550.0, "step": 13100 }, { "epoch": 1.666581859814273, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.593679189682007, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.8526326417922974, "num_tokens": 497261922.0, "step": 13101 }, { "epoch": 1.6667090700928635, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7328004837036133, "learning_rate": 1e-05, "loss": 0.5149, "mean_token_accuracy": 0.8481210470199585, "num_tokens": 497302297.0, "step": 13102 }, { "epoch": 1.666836280371454, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5909194946289062, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8712562322616577, "num_tokens": 497337788.0, "step": 13103 }, { "epoch": 1.6669634906500446, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7193572521209717, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8579014539718628, "num_tokens": 497373455.0, "step": 13104 }, { "epoch": 1.667090700928635, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5759642124176025, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8721485137939453, "num_tokens": 497414743.0, "step": 13105 }, { "epoch": 1.6672179112072256, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.718022108078003, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8756515979766846, "num_tokens": 497446769.0, "step": 13106 }, { "epoch": 1.6673451214858162, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5589535236358643, "learning_rate": 1e-05, "loss": 0.3837, "mean_token_accuracy": 0.8895469307899475, "num_tokens": 497488693.0, "step": 13107 }, { "epoch": 1.6674723317644067, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7285690307617188, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8778841495513916, "num_tokens": 497522258.0, "step": 13108 }, { "epoch": 1.6675995420429972, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5715630054473877, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8664878010749817, "num_tokens": 497557175.0, "step": 13109 }, { "epoch": 1.6677267523215877, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.726330280303955, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8840593099594116, "num_tokens": 497592880.0, "step": 13110 }, { "epoch": 1.667853962600178, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5608129501342773, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8750553131103516, "num_tokens": 497628416.0, "step": 13111 }, { "epoch": 1.6679811728787686, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.731241226196289, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8600242137908936, "num_tokens": 497666739.0, "step": 13112 }, { "epoch": 1.668108383157359, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5569279193878174, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8697370290756226, "num_tokens": 497706307.0, "step": 13113 }, { "epoch": 1.6682355934359496, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.739985704421997, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8576104640960693, "num_tokens": 497747623.0, "step": 13114 }, { "epoch": 1.6683628037145402, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5583980083465576, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8693941235542297, "num_tokens": 497783354.0, "step": 13115 }, { "epoch": 1.6684900139931307, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7517850399017334, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8720660209655762, "num_tokens": 497818178.0, "step": 13116 }, { "epoch": 1.668617224271721, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5580201148986816, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8721864819526672, "num_tokens": 497852899.0, "step": 13117 }, { "epoch": 1.6687444345503115, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7404956817626953, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8729996681213379, "num_tokens": 497896743.0, "step": 13118 }, { "epoch": 1.668871644828902, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5555360317230225, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8755946755409241, "num_tokens": 497938837.0, "step": 13119 }, { "epoch": 1.6689988551074926, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7570438385009766, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8795002698898315, "num_tokens": 497970335.0, "step": 13120 }, { "epoch": 1.669126065386083, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5669472217559814, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8659271001815796, "num_tokens": 498003484.0, "step": 13121 }, { "epoch": 1.6692532756646736, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.738842725753784, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8829378485679626, "num_tokens": 498040932.0, "step": 13122 }, { "epoch": 1.6693804859432642, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.571218729019165, "learning_rate": 1e-05, "loss": 0.5226, "mean_token_accuracy": 0.8477523326873779, "num_tokens": 498078911.0, "step": 13123 }, { "epoch": 1.6695076962218547, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7346818447113037, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8802439570426941, "num_tokens": 498119917.0, "step": 13124 }, { "epoch": 1.6696349065004452, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5539863109588623, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8769271373748779, "num_tokens": 498160359.0, "step": 13125 }, { "epoch": 1.6697621167790357, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.778953790664673, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8651928901672363, "num_tokens": 498197048.0, "step": 13126 }, { "epoch": 1.6698893270576263, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5630204677581787, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8552764058113098, "num_tokens": 498237007.0, "step": 13127 }, { "epoch": 1.6700165373362168, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7556862831115723, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.869999349117279, "num_tokens": 498275324.0, "step": 13128 }, { "epoch": 1.6701437476148073, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.551670551300049, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8540080189704895, "num_tokens": 498317018.0, "step": 13129 }, { "epoch": 1.6702709578933979, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.75710129737854, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8737074732780457, "num_tokens": 498353903.0, "step": 13130 }, { "epoch": 1.6703981681719884, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.559295177459717, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8626145124435425, "num_tokens": 498390959.0, "step": 13131 }, { "epoch": 1.670525378450579, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7530276775360107, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8802022933959961, "num_tokens": 498433067.0, "step": 13132 }, { "epoch": 1.6706525887291694, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.546826124191284, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8655303120613098, "num_tokens": 498476020.0, "step": 13133 }, { "epoch": 1.67077979900776, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7604804039001465, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.874657392501831, "num_tokens": 498515644.0, "step": 13134 }, { "epoch": 1.6709070092863505, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5623552799224854, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8749487400054932, "num_tokens": 498551063.0, "step": 13135 }, { "epoch": 1.6710342195649408, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7703330516815186, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8686500787734985, "num_tokens": 498588849.0, "step": 13136 }, { "epoch": 1.6711614298435313, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.562628746032715, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8654255270957947, "num_tokens": 498626073.0, "step": 13137 }, { "epoch": 1.6712886401221219, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.76774001121521, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8764492869377136, "num_tokens": 498661523.0, "step": 13138 }, { "epoch": 1.6714158504007124, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5573530197143555, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8626492023468018, "num_tokens": 498702075.0, "step": 13139 }, { "epoch": 1.671543060679303, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7786519527435303, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8636768460273743, "num_tokens": 498737425.0, "step": 13140 }, { "epoch": 1.6716702709578934, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.564250946044922, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.850016713142395, "num_tokens": 498779715.0, "step": 13141 }, { "epoch": 1.6717974812364838, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.762338161468506, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8801793456077576, "num_tokens": 498817312.0, "step": 13142 }, { "epoch": 1.6719246915150743, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5705156326293945, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8604245185852051, "num_tokens": 498857990.0, "step": 13143 }, { "epoch": 1.6720519017936648, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7653071880340576, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8681568503379822, "num_tokens": 498896275.0, "step": 13144 }, { "epoch": 1.6721791120722553, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5791518688201904, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8610226511955261, "num_tokens": 498933662.0, "step": 13145 }, { "epoch": 1.6723063223508459, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7481014728546143, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8725975751876831, "num_tokens": 498974216.0, "step": 13146 }, { "epoch": 1.6724335326294364, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5695576667785645, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8691701292991638, "num_tokens": 499015889.0, "step": 13147 }, { "epoch": 1.672560742908027, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.737746477127075, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8809330463409424, "num_tokens": 499056889.0, "step": 13148 }, { "epoch": 1.6726879531866174, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5725245475769043, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8723331689834595, "num_tokens": 499097440.0, "step": 13149 }, { "epoch": 1.672815163465208, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.73443865776062, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8783473968505859, "num_tokens": 499135450.0, "step": 13150 }, { "epoch": 1.6729423737437985, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.58473801612854, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8605916500091553, "num_tokens": 499178487.0, "step": 13151 }, { "epoch": 1.673069584022389, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.728372812271118, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8670761585235596, "num_tokens": 499213824.0, "step": 13152 }, { "epoch": 1.6731967943009796, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.577359914779663, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8696125745773315, "num_tokens": 499252406.0, "step": 13153 }, { "epoch": 1.67332400457957, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.717039108276367, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8634350299835205, "num_tokens": 499295249.0, "step": 13154 }, { "epoch": 1.6734512148581606, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.560930013656616, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.881029486656189, "num_tokens": 499334115.0, "step": 13155 }, { "epoch": 1.6735784251367511, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7261123657226562, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.866671085357666, "num_tokens": 499372181.0, "step": 13156 }, { "epoch": 1.6737056354153417, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5593159198760986, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8621770143508911, "num_tokens": 499413617.0, "step": 13157 }, { "epoch": 1.6738328456939322, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7396345138549805, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8727827072143555, "num_tokens": 499443976.0, "step": 13158 }, { "epoch": 1.6739600559725227, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.565467596054077, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8660871982574463, "num_tokens": 499477973.0, "step": 13159 }, { "epoch": 1.674087266251113, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7297613620758057, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8575068712234497, "num_tokens": 499523163.0, "step": 13160 }, { "epoch": 1.6742144765297036, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.559584856033325, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8554095029830933, "num_tokens": 499567583.0, "step": 13161 }, { "epoch": 1.674341686808294, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7607622146606445, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8634738922119141, "num_tokens": 499603257.0, "step": 13162 }, { "epoch": 1.6744688970868846, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.547619581222534, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.874310314655304, "num_tokens": 499648881.0, "step": 13163 }, { "epoch": 1.6745961073654752, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7569034099578857, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8693605661392212, "num_tokens": 499688632.0, "step": 13164 }, { "epoch": 1.6747233176440657, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.554605722427368, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8714553117752075, "num_tokens": 499723285.0, "step": 13165 }, { "epoch": 1.674850527922656, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7631776332855225, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8531476259231567, "num_tokens": 499761472.0, "step": 13166 }, { "epoch": 1.6749777382012465, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5687599182128906, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8620432615280151, "num_tokens": 499796997.0, "step": 13167 }, { "epoch": 1.675104948479837, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7603328227996826, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8718945980072021, "num_tokens": 499834307.0, "step": 13168 }, { "epoch": 1.6752321587584276, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5509986877441406, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8705542087554932, "num_tokens": 499876497.0, "step": 13169 }, { "epoch": 1.675359369037018, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.766324758529663, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8603785037994385, "num_tokens": 499917620.0, "step": 13170 }, { "epoch": 1.6754865793156086, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.550567388534546, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8772929310798645, "num_tokens": 499957098.0, "step": 13171 }, { "epoch": 1.6756137895941992, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7698323726654053, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.868169367313385, "num_tokens": 500001046.0, "step": 13172 }, { "epoch": 1.6757409998727897, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5547502040863037, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8665529489517212, "num_tokens": 500041443.0, "step": 13173 }, { "epoch": 1.6758682101513802, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.778380870819092, "learning_rate": 1e-05, "loss": 0.3961, "mean_token_accuracy": 0.8871722221374512, "num_tokens": 500076624.0, "step": 13174 }, { "epoch": 1.6759954204299707, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.563777446746826, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8638536930084229, "num_tokens": 500118995.0, "step": 13175 }, { "epoch": 1.6761226307085613, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.781167984008789, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8662679195404053, "num_tokens": 500155134.0, "step": 13176 }, { "epoch": 1.6762498409871518, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.568621873855591, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8770186901092529, "num_tokens": 500189921.0, "step": 13177 }, { "epoch": 1.6763770512657423, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7769198417663574, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8544782400131226, "num_tokens": 500230655.0, "step": 13178 }, { "epoch": 1.6765042615443329, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5651581287384033, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8693197965621948, "num_tokens": 500270827.0, "step": 13179 }, { "epoch": 1.6766314718229234, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.778225898742676, "learning_rate": 1e-05, "loss": 0.4933, "mean_token_accuracy": 0.8593168258666992, "num_tokens": 500312810.0, "step": 13180 }, { "epoch": 1.676758682101514, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5746164321899414, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8618776798248291, "num_tokens": 500345013.0, "step": 13181 }, { "epoch": 1.6768858923801044, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7748825550079346, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8708702325820923, "num_tokens": 500375628.0, "step": 13182 }, { "epoch": 1.677013102658695, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5636987686157227, "learning_rate": 1e-05, "loss": 0.3907, "mean_token_accuracy": 0.8859831094741821, "num_tokens": 500410084.0, "step": 13183 }, { "epoch": 1.6771403129372855, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.762653112411499, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.8880245685577393, "num_tokens": 500452221.0, "step": 13184 }, { "epoch": 1.6772675232158758, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5628974437713623, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.870851457118988, "num_tokens": 500491358.0, "step": 13185 }, { "epoch": 1.6773947334944663, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7652084827423096, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8652723431587219, "num_tokens": 500530363.0, "step": 13186 }, { "epoch": 1.6775219437730569, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5785062313079834, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8782169818878174, "num_tokens": 500568922.0, "step": 13187 }, { "epoch": 1.6776491540516474, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.744800567626953, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8617770671844482, "num_tokens": 500616445.0, "step": 13188 }, { "epoch": 1.677776364330238, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.581611156463623, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.879372775554657, "num_tokens": 500650692.0, "step": 13189 }, { "epoch": 1.6779035746088284, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7419378757476807, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8607387542724609, "num_tokens": 500692465.0, "step": 13190 }, { "epoch": 1.6780307848874187, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.58266544342041, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8597452640533447, "num_tokens": 500728715.0, "step": 13191 }, { "epoch": 1.6781579951660093, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.735044002532959, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8663114309310913, "num_tokens": 500768310.0, "step": 13192 }, { "epoch": 1.6782852054445998, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5761656761169434, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8742721676826477, "num_tokens": 500809777.0, "step": 13193 }, { "epoch": 1.6784124157231903, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7378926277160645, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8612246513366699, "num_tokens": 500851412.0, "step": 13194 }, { "epoch": 1.6785396260017809, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.578803777694702, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8531006574630737, "num_tokens": 500890896.0, "step": 13195 }, { "epoch": 1.6786668362803714, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.725979804992676, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8796903491020203, "num_tokens": 500926826.0, "step": 13196 }, { "epoch": 1.678794046558962, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.593519687652588, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8633779287338257, "num_tokens": 500960755.0, "step": 13197 }, { "epoch": 1.6789212568375524, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7286312580108643, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8703029155731201, "num_tokens": 500993177.0, "step": 13198 }, { "epoch": 1.679048467116143, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5669872760772705, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8704712390899658, "num_tokens": 501031391.0, "step": 13199 }, { "epoch": 1.6791756773947335, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7126829624176025, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8790008425712585, "num_tokens": 501072010.0, "step": 13200 }, { "epoch": 1.679302887673324, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.570638418197632, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.876375675201416, "num_tokens": 501112677.0, "step": 13201 }, { "epoch": 1.6794300979519146, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7148609161376953, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8722595572471619, "num_tokens": 501156634.0, "step": 13202 }, { "epoch": 1.679557308230505, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5901808738708496, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.873043954372406, "num_tokens": 501188194.0, "step": 13203 }, { "epoch": 1.6796845185090956, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7308244705200195, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8667319416999817, "num_tokens": 501216962.0, "step": 13204 }, { "epoch": 1.6798117287876861, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.560988426208496, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8836831450462341, "num_tokens": 501261964.0, "step": 13205 }, { "epoch": 1.6799389390662767, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.714017152786255, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8722853660583496, "num_tokens": 501303701.0, "step": 13206 }, { "epoch": 1.6800661493448672, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.574376106262207, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.879694938659668, "num_tokens": 501338531.0, "step": 13207 }, { "epoch": 1.6801933596234577, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.716269016265869, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8680418729782104, "num_tokens": 501375968.0, "step": 13208 }, { "epoch": 1.680320569902048, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5680408477783203, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8658338785171509, "num_tokens": 501414646.0, "step": 13209 }, { "epoch": 1.6804477801806386, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7171871662139893, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8689765930175781, "num_tokens": 501455035.0, "step": 13210 }, { "epoch": 1.680574990459229, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.56748104095459, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8690404891967773, "num_tokens": 501493198.0, "step": 13211 }, { "epoch": 1.6807022007378196, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.715022087097168, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8747845888137817, "num_tokens": 501534293.0, "step": 13212 }, { "epoch": 1.6808294110164101, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5916824340820312, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8690733909606934, "num_tokens": 501566614.0, "step": 13213 }, { "epoch": 1.6809566212950007, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7060017585754395, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8590502738952637, "num_tokens": 501607535.0, "step": 13214 }, { "epoch": 1.681083831573591, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5799355506896973, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8654478788375854, "num_tokens": 501644208.0, "step": 13215 }, { "epoch": 1.6812110418521815, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7029855251312256, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8588166236877441, "num_tokens": 501684980.0, "step": 13216 }, { "epoch": 1.681338252130772, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5704572200775146, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8722026348114014, "num_tokens": 501722603.0, "step": 13217 }, { "epoch": 1.6814654624093626, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.728428840637207, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8731564283370972, "num_tokens": 501755650.0, "step": 13218 }, { "epoch": 1.681592672687953, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.564777135848999, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8687417507171631, "num_tokens": 501788848.0, "step": 13219 }, { "epoch": 1.6817198829665436, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7385103702545166, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.868046760559082, "num_tokens": 501826396.0, "step": 13220 }, { "epoch": 1.6818470932451342, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5598385334014893, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8591228723526001, "num_tokens": 501863780.0, "step": 13221 }, { "epoch": 1.6819743035237247, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.725724697113037, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8667088150978088, "num_tokens": 501903275.0, "step": 13222 }, { "epoch": 1.6821015138023152, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5552785396575928, "learning_rate": 1e-05, "loss": 0.3653, "mean_token_accuracy": 0.8919069766998291, "num_tokens": 501940397.0, "step": 13223 }, { "epoch": 1.6822287240809057, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.738415479660034, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8708182573318481, "num_tokens": 501979814.0, "step": 13224 }, { "epoch": 1.6823559343594963, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.548170804977417, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8745085597038269, "num_tokens": 502017349.0, "step": 13225 }, { "epoch": 1.6824831446380868, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7496585845947266, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8615787029266357, "num_tokens": 502062441.0, "step": 13226 }, { "epoch": 1.6826103549166773, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5435352325439453, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8695390820503235, "num_tokens": 502099897.0, "step": 13227 }, { "epoch": 1.6827375651952678, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7559587955474854, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8749385476112366, "num_tokens": 502139682.0, "step": 13228 }, { "epoch": 1.6828647754738584, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5551557540893555, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.862132728099823, "num_tokens": 502179534.0, "step": 13229 }, { "epoch": 1.682991985752449, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7599525451660156, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8764359951019287, "num_tokens": 502218031.0, "step": 13230 }, { "epoch": 1.6831191960310394, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5669031143188477, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8770443201065063, "num_tokens": 502251409.0, "step": 13231 }, { "epoch": 1.68324640630963, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7648110389709473, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8747736215591431, "num_tokens": 502285757.0, "step": 13232 }, { "epoch": 1.6833736165882205, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5556578636169434, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8743771314620972, "num_tokens": 502325247.0, "step": 13233 }, { "epoch": 1.6835008268668108, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7643682956695557, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8716810941696167, "num_tokens": 502367691.0, "step": 13234 }, { "epoch": 1.6836280371454013, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5616455078125, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.859359085559845, "num_tokens": 502409084.0, "step": 13235 }, { "epoch": 1.6837552474239919, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7602365016937256, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8730726838111877, "num_tokens": 502448478.0, "step": 13236 }, { "epoch": 1.6838824577025824, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5830330848693848, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8634341359138489, "num_tokens": 502481459.0, "step": 13237 }, { "epoch": 1.684009667981173, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7590789794921875, "learning_rate": 1e-05, "loss": 0.5152, "mean_token_accuracy": 0.8524928092956543, "num_tokens": 502522300.0, "step": 13238 }, { "epoch": 1.6841368782597634, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5655133724212646, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8740174174308777, "num_tokens": 502560257.0, "step": 13239 }, { "epoch": 1.6842640885383537, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.749011993408203, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8608263731002808, "num_tokens": 502602815.0, "step": 13240 }, { "epoch": 1.6843912988169443, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.561330795288086, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8736073970794678, "num_tokens": 502640067.0, "step": 13241 }, { "epoch": 1.6845185090955348, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7551448345184326, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.865892231464386, "num_tokens": 502680351.0, "step": 13242 }, { "epoch": 1.6846457193741253, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.56649112701416, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8778387308120728, "num_tokens": 502715279.0, "step": 13243 }, { "epoch": 1.6847729296527159, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7558038234710693, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8594266772270203, "num_tokens": 502762078.0, "step": 13244 }, { "epoch": 1.6849001399313064, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.557849168777466, "learning_rate": 1e-05, "loss": 0.3948, "mean_token_accuracy": 0.8841607570648193, "num_tokens": 502799886.0, "step": 13245 }, { "epoch": 1.685027350209897, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.768498659133911, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8607136607170105, "num_tokens": 502838414.0, "step": 13246 }, { "epoch": 1.6851545604884874, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.567972421646118, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8623119592666626, "num_tokens": 502879504.0, "step": 13247 }, { "epoch": 1.685281770767078, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.747593879699707, "learning_rate": 1e-05, "loss": 0.3815, "mean_token_accuracy": 0.8901958465576172, "num_tokens": 502919167.0, "step": 13248 }, { "epoch": 1.6854089810456685, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.552828550338745, "learning_rate": 1e-05, "loss": 0.3689, "mean_token_accuracy": 0.8939059972763062, "num_tokens": 502956548.0, "step": 13249 }, { "epoch": 1.685536191324259, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.785024404525757, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8768141269683838, "num_tokens": 502993001.0, "step": 13250 }, { "epoch": 1.6856634016028496, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.55047345161438, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8730117678642273, "num_tokens": 503030690.0, "step": 13251 }, { "epoch": 1.68579061188144, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7728769779205322, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8652927875518799, "num_tokens": 503074010.0, "step": 13252 }, { "epoch": 1.6859178221600306, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.571214437484741, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8724108934402466, "num_tokens": 503111077.0, "step": 13253 }, { "epoch": 1.6860450324386211, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7689273357391357, "learning_rate": 1e-05, "loss": 0.3779, "mean_token_accuracy": 0.894721508026123, "num_tokens": 503146594.0, "step": 13254 }, { "epoch": 1.6861722427172117, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.562077760696411, "learning_rate": 1e-05, "loss": 0.3904, "mean_token_accuracy": 0.8852332830429077, "num_tokens": 503181470.0, "step": 13255 }, { "epoch": 1.6862994529958022, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.773432970046997, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8834555149078369, "num_tokens": 503215839.0, "step": 13256 }, { "epoch": 1.6864266632743927, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.557251453399658, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8791705369949341, "num_tokens": 503253131.0, "step": 13257 }, { "epoch": 1.686553873552983, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7924463748931885, "learning_rate": 1e-05, "loss": 0.5115, "mean_token_accuracy": 0.8483906388282776, "num_tokens": 503290331.0, "step": 13258 }, { "epoch": 1.6866810838315736, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5635836124420166, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.859022855758667, "num_tokens": 503328551.0, "step": 13259 }, { "epoch": 1.686808294110164, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.8082826137542725, "learning_rate": 1e-05, "loss": 0.5807, "mean_token_accuracy": 0.836234450340271, "num_tokens": 503365288.0, "step": 13260 }, { "epoch": 1.6869355043887546, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.554896116256714, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8790292739868164, "num_tokens": 503402941.0, "step": 13261 }, { "epoch": 1.6870627146673451, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.799492835998535, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8602350354194641, "num_tokens": 503443174.0, "step": 13262 }, { "epoch": 1.6871899249459357, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5547525882720947, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8758059740066528, "num_tokens": 503481790.0, "step": 13263 }, { "epoch": 1.687317135224526, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7983274459838867, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.874052107334137, "num_tokens": 503518386.0, "step": 13264 }, { "epoch": 1.6874443455031165, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5581231117248535, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8772449493408203, "num_tokens": 503558985.0, "step": 13265 }, { "epoch": 1.687571555781707, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7942047119140625, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8777363896369934, "num_tokens": 503602423.0, "step": 13266 }, { "epoch": 1.6876987660602976, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5885496139526367, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8785449862480164, "num_tokens": 503634300.0, "step": 13267 }, { "epoch": 1.687825976338888, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.8036959171295166, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8612921237945557, "num_tokens": 503668225.0, "step": 13268 }, { "epoch": 1.6879531866174786, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.582641363143921, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8561530709266663, "num_tokens": 503705158.0, "step": 13269 }, { "epoch": 1.6880803968960691, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7764363288879395, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8726899027824402, "num_tokens": 503741322.0, "step": 13270 }, { "epoch": 1.6882076071746597, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.576226234436035, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8839417695999146, "num_tokens": 503777932.0, "step": 13271 }, { "epoch": 1.6883348174532502, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.769998550415039, "learning_rate": 1e-05, "loss": 0.4013, "mean_token_accuracy": 0.881324291229248, "num_tokens": 503810175.0, "step": 13272 }, { "epoch": 1.6884620277318407, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.592770576477051, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.859707772731781, "num_tokens": 503844498.0, "step": 13273 }, { "epoch": 1.6885892380104313, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.731245279312134, "learning_rate": 1e-05, "loss": 0.3831, "mean_token_accuracy": 0.8896067142486572, "num_tokens": 503878898.0, "step": 13274 }, { "epoch": 1.6887164482890218, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5914082527160645, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8807862997055054, "num_tokens": 503915473.0, "step": 13275 }, { "epoch": 1.6888436585676123, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.723400592803955, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8738365173339844, "num_tokens": 503954234.0, "step": 13276 }, { "epoch": 1.6889708688462028, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6091527938842773, "learning_rate": 1e-05, "loss": 0.5137, "mean_token_accuracy": 0.8510762453079224, "num_tokens": 503988740.0, "step": 13277 }, { "epoch": 1.6890980791247934, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.701738119125366, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8744676113128662, "num_tokens": 504025755.0, "step": 13278 }, { "epoch": 1.689225289403384, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5968611240386963, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8567569255828857, "num_tokens": 504062639.0, "step": 13279 }, { "epoch": 1.6893524996819744, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.707374334335327, "learning_rate": 1e-05, "loss": 0.5001, "mean_token_accuracy": 0.8509026765823364, "num_tokens": 504099645.0, "step": 13280 }, { "epoch": 1.689479709960565, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5758681297302246, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8813642263412476, "num_tokens": 504137678.0, "step": 13281 }, { "epoch": 1.6896069202391555, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7071645259857178, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8722909092903137, "num_tokens": 504174037.0, "step": 13282 }, { "epoch": 1.6897341305177458, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5754435062408447, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8557033538818359, "num_tokens": 504220117.0, "step": 13283 }, { "epoch": 1.6898613407963363, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.714744806289673, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.878900408744812, "num_tokens": 504259104.0, "step": 13284 }, { "epoch": 1.6899885510749268, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.556809425354004, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8723106384277344, "num_tokens": 504302867.0, "step": 13285 }, { "epoch": 1.6901157613535174, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7206850051879883, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.881307065486908, "num_tokens": 504337555.0, "step": 13286 }, { "epoch": 1.690242971632108, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5663108825683594, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8812146186828613, "num_tokens": 504372334.0, "step": 13287 }, { "epoch": 1.6903701819106984, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7228455543518066, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8767749667167664, "num_tokens": 504412894.0, "step": 13288 }, { "epoch": 1.6904973921892887, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5686113834381104, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8741366863250732, "num_tokens": 504452919.0, "step": 13289 }, { "epoch": 1.6906246024678793, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.727713108062744, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8742964267730713, "num_tokens": 504490694.0, "step": 13290 }, { "epoch": 1.6907518127464698, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.564486026763916, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8667657375335693, "num_tokens": 504526606.0, "step": 13291 }, { "epoch": 1.6908790230250603, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.726100444793701, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8763974905014038, "num_tokens": 504568885.0, "step": 13292 }, { "epoch": 1.6910062333036509, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.5602917671203613, "learning_rate": 1e-05, "loss": 0.3994, "mean_token_accuracy": 0.886142909526825, "num_tokens": 504604947.0, "step": 13293 }, { "epoch": 1.6911334435822414, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7316696643829346, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8644592761993408, "num_tokens": 504643937.0, "step": 13294 }, { "epoch": 1.691260653860832, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5644888877868652, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8703631162643433, "num_tokens": 504684298.0, "step": 13295 }, { "epoch": 1.6913878641394224, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7417452335357666, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8572301268577576, "num_tokens": 504723578.0, "step": 13296 }, { "epoch": 1.691515074418013, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5528759956359863, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8634307384490967, "num_tokens": 504768190.0, "step": 13297 }, { "epoch": 1.6916422846966035, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.738239049911499, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8818093538284302, "num_tokens": 504807656.0, "step": 13298 }, { "epoch": 1.691769494975194, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5608651638031006, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8699055910110474, "num_tokens": 504846631.0, "step": 13299 }, { "epoch": 1.6918967052537845, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7355589866638184, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8739999532699585, "num_tokens": 504890451.0, "step": 13300 }, { "epoch": 1.692023915532375, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5526082515716553, "learning_rate": 1e-05, "loss": 0.3956, "mean_token_accuracy": 0.8851003646850586, "num_tokens": 504931863.0, "step": 13301 }, { "epoch": 1.6921511258109656, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.745327949523926, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8778954148292542, "num_tokens": 504970247.0, "step": 13302 }, { "epoch": 1.6922783360895561, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.56081223487854, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8780876398086548, "num_tokens": 505009987.0, "step": 13303 }, { "epoch": 1.6924055463681467, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.743555784225464, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8772284388542175, "num_tokens": 505051933.0, "step": 13304 }, { "epoch": 1.6925327566467372, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.56386137008667, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.874731183052063, "num_tokens": 505086216.0, "step": 13305 }, { "epoch": 1.6926599669253277, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.737304210662842, "learning_rate": 1e-05, "loss": 0.3703, "mean_token_accuracy": 0.8924019932746887, "num_tokens": 505123040.0, "step": 13306 }, { "epoch": 1.692787177203918, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5684378147125244, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8748971223831177, "num_tokens": 505158546.0, "step": 13307 }, { "epoch": 1.6929143874825086, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7480309009552, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.873329222202301, "num_tokens": 505195433.0, "step": 13308 }, { "epoch": 1.693041597761099, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.554330348968506, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8765289187431335, "num_tokens": 505232411.0, "step": 13309 }, { "epoch": 1.6931688080396896, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.767493724822998, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8577670454978943, "num_tokens": 505270590.0, "step": 13310 }, { "epoch": 1.6932960183182801, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5585315227508545, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8722502589225769, "num_tokens": 505307705.0, "step": 13311 }, { "epoch": 1.6934232285968707, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7690696716308594, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8731595277786255, "num_tokens": 505342676.0, "step": 13312 }, { "epoch": 1.693550438875461, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5675153732299805, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8602815866470337, "num_tokens": 505387058.0, "step": 13313 }, { "epoch": 1.6936776491540515, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.758967399597168, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8625460267066956, "num_tokens": 505429584.0, "step": 13314 }, { "epoch": 1.693804859432642, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5567469596862793, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8824026584625244, "num_tokens": 505464619.0, "step": 13315 }, { "epoch": 1.6939320697112326, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.759188413619995, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8660885095596313, "num_tokens": 505507390.0, "step": 13316 }, { "epoch": 1.694059279989823, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5574615001678467, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8709650039672852, "num_tokens": 505543599.0, "step": 13317 }, { "epoch": 1.6941864902684136, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7941107749938965, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8698887825012207, "num_tokens": 505576283.0, "step": 13318 }, { "epoch": 1.6943137005470041, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5520739555358887, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8660625219345093, "num_tokens": 505615244.0, "step": 13319 }, { "epoch": 1.6944409108255947, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7826123237609863, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8625850081443787, "num_tokens": 505653505.0, "step": 13320 }, { "epoch": 1.6945681211041852, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.591127395629883, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8736515641212463, "num_tokens": 505678666.0, "step": 13321 }, { "epoch": 1.6946953313827757, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7621724605560303, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8617762327194214, "num_tokens": 505718191.0, "step": 13322 }, { "epoch": 1.6948225416613663, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.566284418106079, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.885901689529419, "num_tokens": 505759174.0, "step": 13323 }, { "epoch": 1.6949497519399568, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.755136728286743, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8570237159729004, "num_tokens": 505804483.0, "step": 13324 }, { "epoch": 1.6950769622185473, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.577887535095215, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8640612363815308, "num_tokens": 505839767.0, "step": 13325 }, { "epoch": 1.6952041724971378, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.764294147491455, "learning_rate": 1e-05, "loss": 0.506, "mean_token_accuracy": 0.8531801104545593, "num_tokens": 505880207.0, "step": 13326 }, { "epoch": 1.6953313827757284, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.575380563735962, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8762123584747314, "num_tokens": 505914686.0, "step": 13327 }, { "epoch": 1.695458593054319, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.735438823699951, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8736327290534973, "num_tokens": 505959029.0, "step": 13328 }, { "epoch": 1.6955858033329094, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.575270175933838, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8601071834564209, "num_tokens": 505997501.0, "step": 13329 }, { "epoch": 1.6957130136115, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7568559646606445, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8594790697097778, "num_tokens": 506029097.0, "step": 13330 }, { "epoch": 1.6958402238900905, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5711071491241455, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8846032619476318, "num_tokens": 506065912.0, "step": 13331 }, { "epoch": 1.6959674341686808, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7310831546783447, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8763480186462402, "num_tokens": 506108843.0, "step": 13332 }, { "epoch": 1.6960946444472713, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.561447858810425, "learning_rate": 1e-05, "loss": 0.3965, "mean_token_accuracy": 0.8848655223846436, "num_tokens": 506147505.0, "step": 13333 }, { "epoch": 1.6962218547258618, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.736361265182495, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8772265911102295, "num_tokens": 506186491.0, "step": 13334 }, { "epoch": 1.6963490650044524, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5639336109161377, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8783090114593506, "num_tokens": 506223343.0, "step": 13335 }, { "epoch": 1.696476275283043, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.746342897415161, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8776410818099976, "num_tokens": 506262328.0, "step": 13336 }, { "epoch": 1.6966034855616334, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5921480655670166, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8609241843223572, "num_tokens": 506294758.0, "step": 13337 }, { "epoch": 1.6967306958402237, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.73724102973938, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8821730017662048, "num_tokens": 506325423.0, "step": 13338 }, { "epoch": 1.6968579061188143, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.571382522583008, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.859675407409668, "num_tokens": 506367722.0, "step": 13339 }, { "epoch": 1.6969851163974048, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.738093376159668, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8619354963302612, "num_tokens": 506404638.0, "step": 13340 }, { "epoch": 1.6971123266759953, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.574016571044922, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8704363107681274, "num_tokens": 506440976.0, "step": 13341 }, { "epoch": 1.6972395369545858, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7406294345855713, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8667147159576416, "num_tokens": 506475392.0, "step": 13342 }, { "epoch": 1.6973667472331764, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002689361572265625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.559600591659546, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8725315928459167, "num_tokens": 506510933.0, "step": 13343 }, { "epoch": 1.697493957511767, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7424259185791016, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8649790287017822, "num_tokens": 506550217.0, "step": 13344 }, { "epoch": 1.6976211677903574, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.566359043121338, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.87647545337677, "num_tokens": 506583836.0, "step": 13345 }, { "epoch": 1.697748378068948, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7466471195220947, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8783725500106812, "num_tokens": 506616889.0, "step": 13346 }, { "epoch": 1.6978755883475385, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.56007719039917, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8775442838668823, "num_tokens": 506657198.0, "step": 13347 }, { "epoch": 1.698002798626129, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7385778427124023, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8801310658454895, "num_tokens": 506696439.0, "step": 13348 }, { "epoch": 1.6981300089047195, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.570911169052124, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8594895005226135, "num_tokens": 506736140.0, "step": 13349 }, { "epoch": 1.69825721918331, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7464375495910645, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8801697492599487, "num_tokens": 506770976.0, "step": 13350 }, { "epoch": 1.6983844294619006, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.577603578567505, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8737180829048157, "num_tokens": 506804463.0, "step": 13351 }, { "epoch": 1.6985116397404911, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.725493907928467, "learning_rate": 1e-05, "loss": 0.3894, "mean_token_accuracy": 0.8884730339050293, "num_tokens": 506839498.0, "step": 13352 }, { "epoch": 1.6986388500190817, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5704562664031982, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.86725252866745, "num_tokens": 506875899.0, "step": 13353 }, { "epoch": 1.6987660602976722, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7316176891326904, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8646000623703003, "num_tokens": 506914760.0, "step": 13354 }, { "epoch": 1.6988932705762627, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5662319660186768, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8756809234619141, "num_tokens": 506950022.0, "step": 13355 }, { "epoch": 1.699020480854853, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.72312593460083, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.876187264919281, "num_tokens": 506993349.0, "step": 13356 }, { "epoch": 1.6991476911334435, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5649633407592773, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8665087819099426, "num_tokens": 507037519.0, "step": 13357 }, { "epoch": 1.699274901412034, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.754868507385254, "learning_rate": 1e-05, "loss": 0.5285, "mean_token_accuracy": 0.8448131084442139, "num_tokens": 507071932.0, "step": 13358 }, { "epoch": 1.6994021116906246, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5568578243255615, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8716812133789062, "num_tokens": 507108140.0, "step": 13359 }, { "epoch": 1.6995293219692151, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7491841316223145, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8665852546691895, "num_tokens": 507148241.0, "step": 13360 }, { "epoch": 1.6996565322478057, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5644724369049072, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8712440729141235, "num_tokens": 507183200.0, "step": 13361 }, { "epoch": 1.699783742526396, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7467267513275146, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8674800992012024, "num_tokens": 507219036.0, "step": 13362 }, { "epoch": 1.6999109528049865, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.556609869003296, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8838222026824951, "num_tokens": 507254767.0, "step": 13363 }, { "epoch": 1.700038163083577, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.752225160598755, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8851718306541443, "num_tokens": 507289945.0, "step": 13364 }, { "epoch": 1.7001653733621676, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.557400703430176, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8722285032272339, "num_tokens": 507328350.0, "step": 13365 }, { "epoch": 1.700292583640758, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7548751831054688, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8591870069503784, "num_tokens": 507367151.0, "step": 13366 }, { "epoch": 1.7004197939193486, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.560349941253662, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8763819336891174, "num_tokens": 507404181.0, "step": 13367 }, { "epoch": 1.7005470041979391, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.75675106048584, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8825996518135071, "num_tokens": 507438402.0, "step": 13368 }, { "epoch": 1.7006742144765297, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.557852268218994, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8731001615524292, "num_tokens": 507478091.0, "step": 13369 }, { "epoch": 1.7008014247551202, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7566025257110596, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8649172186851501, "num_tokens": 507518285.0, "step": 13370 }, { "epoch": 1.7009286350337107, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5582501888275146, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8720515966415405, "num_tokens": 507555250.0, "step": 13371 }, { "epoch": 1.7010558453123013, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7670440673828125, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8778097629547119, "num_tokens": 507591255.0, "step": 13372 }, { "epoch": 1.7011830555908918, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.571056842803955, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8772971630096436, "num_tokens": 507623805.0, "step": 13373 }, { "epoch": 1.7013102658694823, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.765094518661499, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8621599674224854, "num_tokens": 507660312.0, "step": 13374 }, { "epoch": 1.7014374761480728, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5536704063415527, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8751479387283325, "num_tokens": 507700132.0, "step": 13375 }, { "epoch": 1.7015646864266634, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7709715366363525, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8719527721405029, "num_tokens": 507737521.0, "step": 13376 }, { "epoch": 1.701691896705254, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5406386852264404, "learning_rate": 1e-05, "loss": 0.3782, "mean_token_accuracy": 0.8911697864532471, "num_tokens": 507779327.0, "step": 13377 }, { "epoch": 1.7018191069838444, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.781663179397583, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.87310791015625, "num_tokens": 507821899.0, "step": 13378 }, { "epoch": 1.701946317262435, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5551161766052246, "learning_rate": 1e-05, "loss": 0.3915, "mean_token_accuracy": 0.8863646984100342, "num_tokens": 507862946.0, "step": 13379 }, { "epoch": 1.7020735275410255, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.793382406234741, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8514152765274048, "num_tokens": 507898409.0, "step": 13380 }, { "epoch": 1.7022007378196158, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5511679649353027, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8782733678817749, "num_tokens": 507936670.0, "step": 13381 }, { "epoch": 1.7023279480982063, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7898049354553223, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.87092125415802, "num_tokens": 507976160.0, "step": 13382 }, { "epoch": 1.7024551583767968, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5598020553588867, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8742533922195435, "num_tokens": 508015500.0, "step": 13383 }, { "epoch": 1.7025823686553874, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7875428199768066, "learning_rate": 1e-05, "loss": 0.3968, "mean_token_accuracy": 0.8869048357009888, "num_tokens": 508050860.0, "step": 13384 }, { "epoch": 1.702709578933978, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.574373960494995, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8701152801513672, "num_tokens": 508085794.0, "step": 13385 }, { "epoch": 1.7028367892125684, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7769505977630615, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.8838831782341003, "num_tokens": 508122731.0, "step": 13386 }, { "epoch": 1.7029639994911587, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5614218711853027, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8691882491111755, "num_tokens": 508168513.0, "step": 13387 }, { "epoch": 1.7030912097697493, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.76727294921875, "learning_rate": 1e-05, "loss": 0.3879, "mean_token_accuracy": 0.8884896039962769, "num_tokens": 508209758.0, "step": 13388 }, { "epoch": 1.7032184200483398, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.568160057067871, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8836899995803833, "num_tokens": 508248172.0, "step": 13389 }, { "epoch": 1.7033456303269303, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.770386219024658, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8669614791870117, "num_tokens": 508288694.0, "step": 13390 }, { "epoch": 1.7034728406055208, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5691494941711426, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8654806613922119, "num_tokens": 508329021.0, "step": 13391 }, { "epoch": 1.7036000508841114, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.784987449645996, "learning_rate": 1e-05, "loss": 0.4842, "mean_token_accuracy": 0.8557472229003906, "num_tokens": 508363706.0, "step": 13392 }, { "epoch": 1.703727261162702, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5595083236694336, "learning_rate": 1e-05, "loss": 0.3919, "mean_token_accuracy": 0.8819574117660522, "num_tokens": 508399879.0, "step": 13393 }, { "epoch": 1.7038544714412924, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7748734951019287, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8645057678222656, "num_tokens": 508440172.0, "step": 13394 }, { "epoch": 1.703981681719883, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.578953981399536, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8676199913024902, "num_tokens": 508475531.0, "step": 13395 }, { "epoch": 1.7041088919984735, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7692925930023193, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.880046010017395, "num_tokens": 508509290.0, "step": 13396 }, { "epoch": 1.704236102277064, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.568852424621582, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8748509883880615, "num_tokens": 508545052.0, "step": 13397 }, { "epoch": 1.7043633125556545, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.763080358505249, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8763836622238159, "num_tokens": 508577922.0, "step": 13398 }, { "epoch": 1.704490522834245, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5638256072998047, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.877132773399353, "num_tokens": 508618135.0, "step": 13399 }, { "epoch": 1.7046177331128356, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.773386240005493, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8612760305404663, "num_tokens": 508656665.0, "step": 13400 }, { "epoch": 1.7047449433914261, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.571122407913208, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8728728294372559, "num_tokens": 508693898.0, "step": 13401 }, { "epoch": 1.7048721536700167, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.77724027633667, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.86412513256073, "num_tokens": 508726865.0, "step": 13402 }, { "epoch": 1.7049993639486072, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5765573978424072, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8746179342269897, "num_tokens": 508762040.0, "step": 13403 }, { "epoch": 1.7051265742271977, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7626352310180664, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8702670931816101, "num_tokens": 508803288.0, "step": 13404 }, { "epoch": 1.705253784505788, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.577293872833252, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8782719969749451, "num_tokens": 508840729.0, "step": 13405 }, { "epoch": 1.7053809947843785, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.765368938446045, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8810141086578369, "num_tokens": 508874219.0, "step": 13406 }, { "epoch": 1.705508205062969, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5693533420562744, "learning_rate": 1e-05, "loss": 0.3819, "mean_token_accuracy": 0.8914960026741028, "num_tokens": 508911746.0, "step": 13407 }, { "epoch": 1.7056354153415596, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7707600593566895, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8771176338195801, "num_tokens": 508944996.0, "step": 13408 }, { "epoch": 1.7057626256201501, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5614113807678223, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8648213148117065, "num_tokens": 508986416.0, "step": 13409 }, { "epoch": 1.7058898358987407, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7659971714019775, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8701514005661011, "num_tokens": 509023690.0, "step": 13410 }, { "epoch": 1.706017046177331, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.560899019241333, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8747467398643494, "num_tokens": 509065405.0, "step": 13411 }, { "epoch": 1.7061442564559215, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.781911849975586, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.868300199508667, "num_tokens": 509104871.0, "step": 13412 }, { "epoch": 1.706271466734512, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5771079063415527, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8556283712387085, "num_tokens": 509140938.0, "step": 13413 }, { "epoch": 1.7063986770131025, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.771219253540039, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8761163949966431, "num_tokens": 509175324.0, "step": 13414 }, { "epoch": 1.706525887291693, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5662314891815186, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8739286661148071, "num_tokens": 509214983.0, "step": 13415 }, { "epoch": 1.7066530975702836, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.786245584487915, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8651175498962402, "num_tokens": 509249418.0, "step": 13416 }, { "epoch": 1.7067803078488741, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5712692737579346, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8801733255386353, "num_tokens": 509282884.0, "step": 13417 }, { "epoch": 1.7069075181274647, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7776052951812744, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8770058155059814, "num_tokens": 509317880.0, "step": 13418 }, { "epoch": 1.7070347284060552, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.583292007446289, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8650051355361938, "num_tokens": 509355760.0, "step": 13419 }, { "epoch": 1.7071619386846457, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7561397552490234, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8745673894882202, "num_tokens": 509393088.0, "step": 13420 }, { "epoch": 1.7072891489632362, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.583190441131592, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.866479754447937, "num_tokens": 509431568.0, "step": 13421 }, { "epoch": 1.7074163592418268, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.740332841873169, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8693798780441284, "num_tokens": 509471044.0, "step": 13422 }, { "epoch": 1.7075435695204173, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5796031951904297, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8809207081794739, "num_tokens": 509507746.0, "step": 13423 }, { "epoch": 1.7076707797990078, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.728729248046875, "learning_rate": 1e-05, "loss": 0.4014, "mean_token_accuracy": 0.8846558332443237, "num_tokens": 509545753.0, "step": 13424 }, { "epoch": 1.7077979900775984, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.587071657180786, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8745278716087341, "num_tokens": 509581155.0, "step": 13425 }, { "epoch": 1.7079252003561889, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7325477600097656, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.865550696849823, "num_tokens": 509615540.0, "step": 13426 }, { "epoch": 1.7080524106347794, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5838983058929443, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8693293929100037, "num_tokens": 509651257.0, "step": 13427 }, { "epoch": 1.70817962091337, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7337024211883545, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8554174900054932, "num_tokens": 509688088.0, "step": 13428 }, { "epoch": 1.7083068311919605, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.7358293533325195, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8751546740531921, "num_tokens": 509721122.0, "step": 13429 }, { "epoch": 1.7084340414705508, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7553341388702393, "learning_rate": 1e-05, "loss": 0.4994, "mean_token_accuracy": 0.8534661531448364, "num_tokens": 509756711.0, "step": 13430 }, { "epoch": 1.7085612517491413, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5909502506256104, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8715956807136536, "num_tokens": 509793868.0, "step": 13431 }, { "epoch": 1.7086884620277318, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6783454418182373, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.881952166557312, "num_tokens": 509828741.0, "step": 13432 }, { "epoch": 1.7088156723063224, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000270843505859375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.580922842025757, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8803238272666931, "num_tokens": 509866534.0, "step": 13433 }, { "epoch": 1.708942882584913, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.696246862411499, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8703795075416565, "num_tokens": 509897664.0, "step": 13434 }, { "epoch": 1.7090700928635034, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.607473373413086, "learning_rate": 1e-05, "loss": 0.5248, "mean_token_accuracy": 0.8472820520401001, "num_tokens": 509933349.0, "step": 13435 }, { "epoch": 1.7091973031420937, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6788418292999268, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8680439591407776, "num_tokens": 509970912.0, "step": 13436 }, { "epoch": 1.7093245134206843, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5980417728424072, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.866769552230835, "num_tokens": 510007431.0, "step": 13437 }, { "epoch": 1.7094517236992748, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6809520721435547, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8687707185745239, "num_tokens": 510044109.0, "step": 13438 }, { "epoch": 1.7095789339778653, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.594068765640259, "learning_rate": 1e-05, "loss": 0.4776, "mean_token_accuracy": 0.8554040193557739, "num_tokens": 510084665.0, "step": 13439 }, { "epoch": 1.7097061442564558, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.687631368637085, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.85783451795578, "num_tokens": 510123108.0, "step": 13440 }, { "epoch": 1.7098333545350464, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.57869815826416, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8664385080337524, "num_tokens": 510167918.0, "step": 13441 }, { "epoch": 1.709960564813637, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.691654920578003, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8810456991195679, "num_tokens": 510204555.0, "step": 13442 }, { "epoch": 1.7100877750922274, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.59749436378479, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.85230553150177, "num_tokens": 510239508.0, "step": 13443 }, { "epoch": 1.710214985370818, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.682352304458618, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8736457824707031, "num_tokens": 510275246.0, "step": 13444 }, { "epoch": 1.7103421956494085, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5788357257843018, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.872673749923706, "num_tokens": 510316558.0, "step": 13445 }, { "epoch": 1.710469405927999, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6901841163635254, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8772832751274109, "num_tokens": 510351784.0, "step": 13446 }, { "epoch": 1.7105966162065895, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.572061538696289, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8770223259925842, "num_tokens": 510391044.0, "step": 13447 }, { "epoch": 1.71072382648518, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.698648691177368, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8730482459068298, "num_tokens": 510431199.0, "step": 13448 }, { "epoch": 1.7108510367637706, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.563974618911743, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8755996823310852, "num_tokens": 510471629.0, "step": 13449 }, { "epoch": 1.7109782470423611, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7027695178985596, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.8830446004867554, "num_tokens": 510511546.0, "step": 13450 }, { "epoch": 1.7111054573209517, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.572913885116577, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8644475936889648, "num_tokens": 510548942.0, "step": 13451 }, { "epoch": 1.7112326675995422, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.752610683441162, "learning_rate": 1e-05, "loss": 0.5282, "mean_token_accuracy": 0.8448195457458496, "num_tokens": 510589855.0, "step": 13452 }, { "epoch": 1.7113598778781327, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.545013189315796, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.875188410282135, "num_tokens": 510628960.0, "step": 13453 }, { "epoch": 1.711487088156723, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7471964359283447, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8666739463806152, "num_tokens": 510671107.0, "step": 13454 }, { "epoch": 1.7116142984353135, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.552490234375, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8692301511764526, "num_tokens": 510708957.0, "step": 13455 }, { "epoch": 1.711741508713904, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7612318992614746, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8729053735733032, "num_tokens": 510746150.0, "step": 13456 }, { "epoch": 1.7118687189924946, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5699644088745117, "learning_rate": 1e-05, "loss": 0.4745, "mean_token_accuracy": 0.8618011474609375, "num_tokens": 510775810.0, "step": 13457 }, { "epoch": 1.7119959292710851, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.754755973815918, "learning_rate": 1e-05, "loss": 0.4841, "mean_token_accuracy": 0.8589893579483032, "num_tokens": 510817438.0, "step": 13458 }, { "epoch": 1.7121231395496757, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.556365966796875, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8681182265281677, "num_tokens": 510854529.0, "step": 13459 }, { "epoch": 1.712250349828266, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.76491117477417, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8702652454376221, "num_tokens": 510891568.0, "step": 13460 }, { "epoch": 1.7123775601068565, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.558406114578247, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8749703764915466, "num_tokens": 510928424.0, "step": 13461 }, { "epoch": 1.712504770385447, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7650279998779297, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8661288619041443, "num_tokens": 510971770.0, "step": 13462 }, { "epoch": 1.7126319806640375, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.584179639816284, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8622600436210632, "num_tokens": 511004087.0, "step": 13463 }, { "epoch": 1.712759190942628, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7468008995056152, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8625223636627197, "num_tokens": 511046494.0, "step": 13464 }, { "epoch": 1.7128864012212186, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5669333934783936, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8636630177497864, "num_tokens": 511088729.0, "step": 13465 }, { "epoch": 1.7130136114998091, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.743175506591797, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8748104572296143, "num_tokens": 511127581.0, "step": 13466 }, { "epoch": 1.7131408217783997, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.568387985229492, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8644447922706604, "num_tokens": 511166072.0, "step": 13467 }, { "epoch": 1.7132680320569902, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7435646057128906, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8708492517471313, "num_tokens": 511203529.0, "step": 13468 }, { "epoch": 1.7133952423355807, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5618162155151367, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8703720569610596, "num_tokens": 511245846.0, "step": 13469 }, { "epoch": 1.7135224526141712, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.74223256111145, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8655116558074951, "num_tokens": 511287280.0, "step": 13470 }, { "epoch": 1.7136496628927618, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5486106872558594, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8724862337112427, "num_tokens": 511330905.0, "step": 13471 }, { "epoch": 1.7137768731713523, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.778521776199341, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8591241240501404, "num_tokens": 511369872.0, "step": 13472 }, { "epoch": 1.7139040834499428, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.539032220840454, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.878509521484375, "num_tokens": 511410638.0, "step": 13473 }, { "epoch": 1.7140312937285334, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7960288524627686, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.872722864151001, "num_tokens": 511445517.0, "step": 13474 }, { "epoch": 1.7141585040071239, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.556530237197876, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8566216826438904, "num_tokens": 511482749.0, "step": 13475 }, { "epoch": 1.7142857142857144, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8051939010620117, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.861391544342041, "num_tokens": 511517001.0, "step": 13476 }, { "epoch": 1.714412924564305, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.550870656967163, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8654394149780273, "num_tokens": 511559437.0, "step": 13477 }, { "epoch": 1.7145401348428955, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8198769092559814, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8721017837524414, "num_tokens": 511593713.0, "step": 13478 }, { "epoch": 1.7146673451214858, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.564220428466797, "learning_rate": 1e-05, "loss": 0.4834, "mean_token_accuracy": 0.8591734170913696, "num_tokens": 511633515.0, "step": 13479 }, { "epoch": 1.7147945554000763, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.810638666152954, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8695459365844727, "num_tokens": 511671539.0, "step": 13480 }, { "epoch": 1.7149217656786668, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.578423023223877, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8797154426574707, "num_tokens": 511706972.0, "step": 13481 }, { "epoch": 1.7150489759572574, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8078386783599854, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.865739107131958, "num_tokens": 511740186.0, "step": 13482 }, { "epoch": 1.7151761862358479, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.572199821472168, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8700194358825684, "num_tokens": 511779755.0, "step": 13483 }, { "epoch": 1.7153033965144384, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8034675121307373, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8625900745391846, "num_tokens": 511822531.0, "step": 13484 }, { "epoch": 1.7154306067930287, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5928120613098145, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8681595921516418, "num_tokens": 511852950.0, "step": 13485 }, { "epoch": 1.7155578170716193, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7827866077423096, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.877985954284668, "num_tokens": 511890251.0, "step": 13486 }, { "epoch": 1.7156850273502098, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5851924419403076, "learning_rate": 1e-05, "loss": 0.4928, "mean_token_accuracy": 0.8545634746551514, "num_tokens": 511933103.0, "step": 13487 }, { "epoch": 1.7158122376288003, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.76127290725708, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8779361844062805, "num_tokens": 511975682.0, "step": 13488 }, { "epoch": 1.7159394479073908, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5810556411743164, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8706364035606384, "num_tokens": 512014899.0, "step": 13489 }, { "epoch": 1.7160666581859814, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.758697509765625, "learning_rate": 1e-05, "loss": 0.473, "mean_token_accuracy": 0.8585953712463379, "num_tokens": 512054951.0, "step": 13490 }, { "epoch": 1.716193868464572, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5876948833465576, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8691438436508179, "num_tokens": 512094273.0, "step": 13491 }, { "epoch": 1.7163210787431624, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7563600540161133, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8665452003479004, "num_tokens": 512133493.0, "step": 13492 }, { "epoch": 1.716448289021753, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5829639434814453, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8719820380210876, "num_tokens": 512171527.0, "step": 13493 }, { "epoch": 1.7165754993003435, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7472116947174072, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8744013905525208, "num_tokens": 512211106.0, "step": 13494 }, { "epoch": 1.716702709578934, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5829029083251953, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8747825622558594, "num_tokens": 512247444.0, "step": 13495 }, { "epoch": 1.7168299198575245, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7294387817382812, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8627820014953613, "num_tokens": 512293912.0, "step": 13496 }, { "epoch": 1.716957130136115, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5784120559692383, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.881603479385376, "num_tokens": 512334149.0, "step": 13497 }, { "epoch": 1.7170843404147056, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7300868034362793, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8758314847946167, "num_tokens": 512371161.0, "step": 13498 }, { "epoch": 1.7172115506932961, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5919394493103027, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8579350113868713, "num_tokens": 512407217.0, "step": 13499 }, { "epoch": 1.7173387609718866, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.726658344268799, "learning_rate": 1e-05, "loss": 0.4061, "mean_token_accuracy": 0.8823186755180359, "num_tokens": 512441975.0, "step": 13500 }, { "epoch": 1.7174659712504772, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5846564769744873, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8622899651527405, "num_tokens": 512482956.0, "step": 13501 }, { "epoch": 1.7175931815290677, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7333054542541504, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8689596652984619, "num_tokens": 512518514.0, "step": 13502 }, { "epoch": 1.717720391807658, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5656464099884033, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8621336221694946, "num_tokens": 512561620.0, "step": 13503 }, { "epoch": 1.7178476020862485, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7276248931884766, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8741997480392456, "num_tokens": 512598234.0, "step": 13504 }, { "epoch": 1.717974812364839, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5600409507751465, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8694834113121033, "num_tokens": 512638219.0, "step": 13505 }, { "epoch": 1.7181020226434296, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.747380495071411, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8593891859054565, "num_tokens": 512672277.0, "step": 13506 }, { "epoch": 1.7182292329220201, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5557634830474854, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8756588697433472, "num_tokens": 512709573.0, "step": 13507 }, { "epoch": 1.7183564432006107, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7525784969329834, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8635357022285461, "num_tokens": 512750243.0, "step": 13508 }, { "epoch": 1.718483653479201, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.583584785461426, "learning_rate": 1e-05, "loss": 0.5166, "mean_token_accuracy": 0.848004937171936, "num_tokens": 512786001.0, "step": 13509 }, { "epoch": 1.7186108637577915, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7386422157287598, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8742642998695374, "num_tokens": 512823844.0, "step": 13510 }, { "epoch": 1.718738074036382, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.570434808731079, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8709805011749268, "num_tokens": 512859474.0, "step": 13511 }, { "epoch": 1.7188652843149725, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7514853477478027, "learning_rate": 1e-05, "loss": 0.4003, "mean_token_accuracy": 0.8825256824493408, "num_tokens": 512889670.0, "step": 13512 }, { "epoch": 1.718992494593563, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.571829319000244, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8801403045654297, "num_tokens": 512919749.0, "step": 13513 }, { "epoch": 1.7191197048721536, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.765568256378174, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8687956929206848, "num_tokens": 512953206.0, "step": 13514 }, { "epoch": 1.7192469151507441, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.570910692214966, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8667709827423096, "num_tokens": 512989066.0, "step": 13515 }, { "epoch": 1.7193741254293347, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.74919056892395, "learning_rate": 1e-05, "loss": 0.3881, "mean_token_accuracy": 0.8880345225334167, "num_tokens": 513021975.0, "step": 13516 }, { "epoch": 1.7195013357079252, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5673584938049316, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8711532950401306, "num_tokens": 513061669.0, "step": 13517 }, { "epoch": 1.7196285459865157, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.753230094909668, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.869135856628418, "num_tokens": 513095714.0, "step": 13518 }, { "epoch": 1.7197557562651062, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.558675765991211, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8708093166351318, "num_tokens": 513137235.0, "step": 13519 }, { "epoch": 1.7198829665436968, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7565340995788574, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8721339106559753, "num_tokens": 513176339.0, "step": 13520 }, { "epoch": 1.7200101768222873, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5466864109039307, "learning_rate": 1e-05, "loss": 0.3826, "mean_token_accuracy": 0.8897602558135986, "num_tokens": 513218984.0, "step": 13521 }, { "epoch": 1.7201373871008778, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.770251512527466, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8624376654624939, "num_tokens": 513259487.0, "step": 13522 }, { "epoch": 1.7202645973794684, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.56848406791687, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8643680810928345, "num_tokens": 513296270.0, "step": 13523 }, { "epoch": 1.7203918076580589, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7877888679504395, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8630071878433228, "num_tokens": 513333282.0, "step": 13524 }, { "epoch": 1.7205190179366494, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5748398303985596, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8731991052627563, "num_tokens": 513365679.0, "step": 13525 }, { "epoch": 1.72064622821524, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7723581790924072, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8715353608131409, "num_tokens": 513403387.0, "step": 13526 }, { "epoch": 1.7207734384938305, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.580618143081665, "learning_rate": 1e-05, "loss": 0.519, "mean_token_accuracy": 0.847906231880188, "num_tokens": 513442886.0, "step": 13527 }, { "epoch": 1.7209006487724208, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.753309965133667, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8779771327972412, "num_tokens": 513482401.0, "step": 13528 }, { "epoch": 1.7210278590510113, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5747640132904053, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8707737922668457, "num_tokens": 513517898.0, "step": 13529 }, { "epoch": 1.7211550693296018, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7511556148529053, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8739637136459351, "num_tokens": 513556761.0, "step": 13530 }, { "epoch": 1.7212822796081924, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.588444232940674, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.861260712146759, "num_tokens": 513590803.0, "step": 13531 }, { "epoch": 1.7214094898867829, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.735491991043091, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.8922241926193237, "num_tokens": 513626718.0, "step": 13532 }, { "epoch": 1.7215367001653734, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5997684001922607, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8695904016494751, "num_tokens": 513663439.0, "step": 13533 }, { "epoch": 1.7216639104439637, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7145192623138428, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8731136322021484, "num_tokens": 513699916.0, "step": 13534 }, { "epoch": 1.7217911207225542, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.576878786087036, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8734000325202942, "num_tokens": 513739411.0, "step": 13535 }, { "epoch": 1.7219183310011448, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7150216102600098, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8752219676971436, "num_tokens": 513777497.0, "step": 13536 }, { "epoch": 1.7220455412797353, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5763494968414307, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8661730289459229, "num_tokens": 513817658.0, "step": 13537 }, { "epoch": 1.7221727515583258, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7182443141937256, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8836631774902344, "num_tokens": 513853011.0, "step": 13538 }, { "epoch": 1.7222999618369164, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5598015785217285, "learning_rate": 1e-05, "loss": 0.3761, "mean_token_accuracy": 0.8954432010650635, "num_tokens": 513892933.0, "step": 13539 }, { "epoch": 1.7224271721155069, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.735621690750122, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8584200143814087, "num_tokens": 513932284.0, "step": 13540 }, { "epoch": 1.7225543823940974, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5648186206817627, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8790143728256226, "num_tokens": 513968151.0, "step": 13541 }, { "epoch": 1.722681592672688, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.738013744354248, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8700399398803711, "num_tokens": 514005323.0, "step": 13542 }, { "epoch": 1.7228088029512785, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.56809401512146, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8568923473358154, "num_tokens": 514044158.0, "step": 13543 }, { "epoch": 1.722936013229869, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7416183948516846, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8745477199554443, "num_tokens": 514081076.0, "step": 13544 }, { "epoch": 1.7230632235084595, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5610389709472656, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8588755130767822, "num_tokens": 514122784.0, "step": 13545 }, { "epoch": 1.72319043378705, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.759054183959961, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8699880242347717, "num_tokens": 514158998.0, "step": 13546 }, { "epoch": 1.7233176440656406, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5576324462890625, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8689618110656738, "num_tokens": 514195815.0, "step": 13547 }, { "epoch": 1.7234448543442311, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.781282424926758, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8663578629493713, "num_tokens": 514228353.0, "step": 13548 }, { "epoch": 1.7235720646228216, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5502142906188965, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8658120036125183, "num_tokens": 514267283.0, "step": 13549 }, { "epoch": 1.7236992749014122, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.77544903755188, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8781267404556274, "num_tokens": 514309950.0, "step": 13550 }, { "epoch": 1.7238264851800027, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.557903289794922, "learning_rate": 1e-05, "loss": 0.3924, "mean_token_accuracy": 0.8828321695327759, "num_tokens": 514345043.0, "step": 13551 }, { "epoch": 1.723953695458593, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7914860248565674, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8727640509605408, "num_tokens": 514383692.0, "step": 13552 }, { "epoch": 1.7240809057371835, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5570924282073975, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8580851554870605, "num_tokens": 514423552.0, "step": 13553 }, { "epoch": 1.724208116015774, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7984743118286133, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.868775486946106, "num_tokens": 514460460.0, "step": 13554 }, { "epoch": 1.7243353262943646, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5502946376800537, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8657363057136536, "num_tokens": 514504751.0, "step": 13555 }, { "epoch": 1.7244625365729551, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8090476989746094, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8677384853363037, "num_tokens": 514539967.0, "step": 13556 }, { "epoch": 1.7245897468515456, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5550427436828613, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8743513822555542, "num_tokens": 514583191.0, "step": 13557 }, { "epoch": 1.724716957130136, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7998502254486084, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8727618455886841, "num_tokens": 514622832.0, "step": 13558 }, { "epoch": 1.7248441674087265, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5600621700286865, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8766230344772339, "num_tokens": 514663387.0, "step": 13559 }, { "epoch": 1.724971377687317, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.805598258972168, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8650333881378174, "num_tokens": 514702711.0, "step": 13560 }, { "epoch": 1.7250985879659075, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5791609287261963, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8552005290985107, "num_tokens": 514740341.0, "step": 13561 }, { "epoch": 1.725225798244498, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8072052001953125, "learning_rate": 1e-05, "loss": 0.5058, "mean_token_accuracy": 0.8555463552474976, "num_tokens": 514774834.0, "step": 13562 }, { "epoch": 1.7253530085230886, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.577186107635498, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8568164110183716, "num_tokens": 514814988.0, "step": 13563 }, { "epoch": 1.7254802188016791, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7931439876556396, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8762571811676025, "num_tokens": 514854936.0, "step": 13564 }, { "epoch": 1.7256074290802697, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.589783191680908, "learning_rate": 1e-05, "loss": 0.4973, "mean_token_accuracy": 0.8535231947898865, "num_tokens": 514890528.0, "step": 13565 }, { "epoch": 1.7257346393588602, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.775222063064575, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8669123649597168, "num_tokens": 514935455.0, "step": 13566 }, { "epoch": 1.7258618496374507, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5866806507110596, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8639652729034424, "num_tokens": 514975059.0, "step": 13567 }, { "epoch": 1.7259890599160412, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7742526531219482, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8790231347084045, "num_tokens": 515007820.0, "step": 13568 }, { "epoch": 1.7261162701946318, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6081457138061523, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8617758750915527, "num_tokens": 515051826.0, "step": 13569 }, { "epoch": 1.7262434804732223, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7390975952148438, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8659075498580933, "num_tokens": 515092316.0, "step": 13570 }, { "epoch": 1.7263706907518128, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6002800464630127, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8850198984146118, "num_tokens": 515124314.0, "step": 13571 }, { "epoch": 1.7264979010304033, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7355220317840576, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8698927164077759, "num_tokens": 515159725.0, "step": 13572 }, { "epoch": 1.7266251113089939, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.58994460105896, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8606878519058228, "num_tokens": 515201095.0, "step": 13573 }, { "epoch": 1.7267523215875844, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7167186737060547, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8740853071212769, "num_tokens": 515241761.0, "step": 13574 }, { "epoch": 1.726879531866175, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.589905261993408, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8692957162857056, "num_tokens": 515281983.0, "step": 13575 }, { "epoch": 1.7270067421447655, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.731975555419922, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8646063804626465, "num_tokens": 515315557.0, "step": 13576 }, { "epoch": 1.7271339524233558, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.611604690551758, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8625157475471497, "num_tokens": 515345969.0, "step": 13577 }, { "epoch": 1.7272611627019463, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.707559823989868, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8634272813796997, "num_tokens": 515385520.0, "step": 13578 }, { "epoch": 1.7273883729805368, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5974795818328857, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.861302375793457, "num_tokens": 515424296.0, "step": 13579 }, { "epoch": 1.7275155832591274, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.698587656021118, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8821073770523071, "num_tokens": 515461602.0, "step": 13580 }, { "epoch": 1.7276427935377179, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.584207057952881, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8565868735313416, "num_tokens": 515504123.0, "step": 13581 }, { "epoch": 1.7277700038163084, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6992874145507812, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.872967541217804, "num_tokens": 515543056.0, "step": 13582 }, { "epoch": 1.7278972140948987, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5851101875305176, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8662605285644531, "num_tokens": 515580194.0, "step": 13583 }, { "epoch": 1.7280244243734892, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.703299045562744, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.874972403049469, "num_tokens": 515612493.0, "step": 13584 }, { "epoch": 1.7281516346520798, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.556877374649048, "learning_rate": 1e-05, "loss": 0.3994, "mean_token_accuracy": 0.8841263055801392, "num_tokens": 515656285.0, "step": 13585 }, { "epoch": 1.7282788449306703, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7104859352111816, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8764443397521973, "num_tokens": 515695326.0, "step": 13586 }, { "epoch": 1.7284060552092608, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5762948989868164, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8546820878982544, "num_tokens": 515730865.0, "step": 13587 }, { "epoch": 1.7285332654878514, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7233035564422607, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8628878593444824, "num_tokens": 515766725.0, "step": 13588 }, { "epoch": 1.7286604757664419, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5743892192840576, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8649718761444092, "num_tokens": 515807660.0, "step": 13589 }, { "epoch": 1.7287876860450324, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.725277900695801, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8801077604293823, "num_tokens": 515844833.0, "step": 13590 }, { "epoch": 1.728914896323623, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.57743763923645, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8713927268981934, "num_tokens": 515876538.0, "step": 13591 }, { "epoch": 1.7290421066022135, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.713197946548462, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8698489665985107, "num_tokens": 515916217.0, "step": 13592 }, { "epoch": 1.729169316880804, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5770013332366943, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8498561382293701, "num_tokens": 515955129.0, "step": 13593 }, { "epoch": 1.7292965271593945, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7168025970458984, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8634002208709717, "num_tokens": 515993910.0, "step": 13594 }, { "epoch": 1.729423737437985, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5557570457458496, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8848350644111633, "num_tokens": 516034376.0, "step": 13595 }, { "epoch": 1.7295509477165756, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7481729984283447, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8656072616577148, "num_tokens": 516070383.0, "step": 13596 }, { "epoch": 1.729678157995166, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.543581962585449, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8700187802314758, "num_tokens": 516114527.0, "step": 13597 }, { "epoch": 1.7298053682737566, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7489590644836426, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8807361721992493, "num_tokens": 516156179.0, "step": 13598 }, { "epoch": 1.7299325785523472, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5489490032196045, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.876716673374176, "num_tokens": 516191000.0, "step": 13599 }, { "epoch": 1.7300597888309377, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7772440910339355, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8740724921226501, "num_tokens": 516224194.0, "step": 13600 }, { "epoch": 1.730186999109528, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5425941944122314, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8739833831787109, "num_tokens": 516262210.0, "step": 13601 }, { "epoch": 1.7303142093881185, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7960097789764404, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8653014302253723, "num_tokens": 516298859.0, "step": 13602 }, { "epoch": 1.730441419666709, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.536057949066162, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8628213405609131, "num_tokens": 516344100.0, "step": 13603 }, { "epoch": 1.7305686299452996, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7964165210723877, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8630221486091614, "num_tokens": 516391313.0, "step": 13604 }, { "epoch": 1.7306958402238901, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5607666969299316, "learning_rate": 1e-05, "loss": 0.4832, "mean_token_accuracy": 0.8584328293800354, "num_tokens": 516430814.0, "step": 13605 }, { "epoch": 1.7308230505024806, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7989516258239746, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8724045753479004, "num_tokens": 516471380.0, "step": 13606 }, { "epoch": 1.730950260781071, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.54386043548584, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8712667226791382, "num_tokens": 516518374.0, "step": 13607 }, { "epoch": 1.7310774710596615, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.826752185821533, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8669308423995972, "num_tokens": 516549917.0, "step": 13608 }, { "epoch": 1.731204681338252, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.554544687271118, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8559256792068481, "num_tokens": 516592869.0, "step": 13609 }, { "epoch": 1.7313318916168425, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.816528558731079, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8730676174163818, "num_tokens": 516630550.0, "step": 13610 }, { "epoch": 1.731459101895433, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5639712810516357, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8775816559791565, "num_tokens": 516669285.0, "step": 13611 }, { "epoch": 1.7315863121740236, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8204164505004883, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8840715885162354, "num_tokens": 516704616.0, "step": 13612 }, { "epoch": 1.7317135224526141, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.560192584991455, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8779665231704712, "num_tokens": 516749566.0, "step": 13613 }, { "epoch": 1.7318407327312046, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8271484375, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8741194009780884, "num_tokens": 516786264.0, "step": 13614 }, { "epoch": 1.7319679430097952, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.578678607940674, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8644130229949951, "num_tokens": 516822471.0, "step": 13615 }, { "epoch": 1.7320951532883857, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8234848976135254, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8696972131729126, "num_tokens": 516861840.0, "step": 13616 }, { "epoch": 1.7322223635669762, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5896682739257812, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8689790964126587, "num_tokens": 516895451.0, "step": 13617 }, { "epoch": 1.7323495738455668, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7982065677642822, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8780245780944824, "num_tokens": 516935926.0, "step": 13618 }, { "epoch": 1.7324767841241573, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5875654220581055, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8713116645812988, "num_tokens": 516976268.0, "step": 13619 }, { "epoch": 1.7326039944027478, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7757251262664795, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.870140552520752, "num_tokens": 517021459.0, "step": 13620 }, { "epoch": 1.7327312046813383, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6055262088775635, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8755269646644592, "num_tokens": 517052441.0, "step": 13621 }, { "epoch": 1.7328584149599289, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.759702205657959, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8598306179046631, "num_tokens": 517094575.0, "step": 13622 }, { "epoch": 1.7329856252385194, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.59285831451416, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8687940835952759, "num_tokens": 517135239.0, "step": 13623 }, { "epoch": 1.73311283551711, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.742185115814209, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8684191703796387, "num_tokens": 517174132.0, "step": 13624 }, { "epoch": 1.7332400457957005, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5955307483673096, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.87294602394104, "num_tokens": 517208693.0, "step": 13625 }, { "epoch": 1.7333672560742908, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.73759126663208, "learning_rate": 1e-05, "loss": 0.5081, "mean_token_accuracy": 0.8525316119194031, "num_tokens": 517249441.0, "step": 13626 }, { "epoch": 1.7334944663528813, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.58854079246521, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8660916686058044, "num_tokens": 517291368.0, "step": 13627 }, { "epoch": 1.7336216766314718, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7331671714782715, "learning_rate": 1e-05, "loss": 0.4786, "mean_token_accuracy": 0.8622774481773376, "num_tokens": 517328292.0, "step": 13628 }, { "epoch": 1.7337488869100623, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5953288078308105, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8775969743728638, "num_tokens": 517361049.0, "step": 13629 }, { "epoch": 1.7338760971886529, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7355453968048096, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8694272041320801, "num_tokens": 517395086.0, "step": 13630 }, { "epoch": 1.7340033074672434, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5788321495056152, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8688413500785828, "num_tokens": 517432919.0, "step": 13631 }, { "epoch": 1.7341305177458337, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7307209968566895, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8700464963912964, "num_tokens": 517474676.0, "step": 13632 }, { "epoch": 1.7342577280244242, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5743303298950195, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8783944845199585, "num_tokens": 517514131.0, "step": 13633 }, { "epoch": 1.7343849383030148, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7351176738739014, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8730783462524414, "num_tokens": 517547704.0, "step": 13634 }, { "epoch": 1.7345121485816053, "ewc_loss": 0.0654296875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.838539123535156e-05, "grad_norm": 2.561143159866333, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8682295083999634, "num_tokens": 517586430.0, "step": 13635 }, { "epoch": 1.7346393588601958, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7374236583709717, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8566566109657288, "num_tokens": 517630303.0, "step": 13636 }, { "epoch": 1.7347665691387864, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5670464038848877, "learning_rate": 1e-05, "loss": 0.3744, "mean_token_accuracy": 0.8886314034461975, "num_tokens": 517663276.0, "step": 13637 }, { "epoch": 1.7348937794173769, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7500078678131104, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.866554856300354, "num_tokens": 517697946.0, "step": 13638 }, { "epoch": 1.7350209896959674, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.577599048614502, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8685065507888794, "num_tokens": 517730820.0, "step": 13639 }, { "epoch": 1.735148199974558, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7349131107330322, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8750102519989014, "num_tokens": 517774576.0, "step": 13640 }, { "epoch": 1.7352754102531485, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5683369636535645, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8611568212509155, "num_tokens": 517814748.0, "step": 13641 }, { "epoch": 1.735402620531739, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.758134603500366, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8703783750534058, "num_tokens": 517846582.0, "step": 13642 }, { "epoch": 1.7355298308103295, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5741262435913086, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8560329079627991, "num_tokens": 517883354.0, "step": 13643 }, { "epoch": 1.73565704108892, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7539920806884766, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8540606498718262, "num_tokens": 517920920.0, "step": 13644 }, { "epoch": 1.7357842513675106, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5679612159729004, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8588369488716125, "num_tokens": 517964844.0, "step": 13645 }, { "epoch": 1.735911461646101, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.761942148208618, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.867974042892456, "num_tokens": 517998803.0, "step": 13646 }, { "epoch": 1.7360386719246916, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5500025749206543, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8673979043960571, "num_tokens": 518040012.0, "step": 13647 }, { "epoch": 1.7361658822032822, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7701029777526855, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8616962432861328, "num_tokens": 518082287.0, "step": 13648 }, { "epoch": 1.7362930924818727, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5675230026245117, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8712356090545654, "num_tokens": 518116898.0, "step": 13649 }, { "epoch": 1.736420302760463, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7525250911712646, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8770428895950317, "num_tokens": 518157559.0, "step": 13650 }, { "epoch": 1.7365475130390535, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5601742267608643, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8627030253410339, "num_tokens": 518200691.0, "step": 13651 }, { "epoch": 1.736674723317644, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.764096736907959, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8547048568725586, "num_tokens": 518239188.0, "step": 13652 }, { "epoch": 1.7368019335962346, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.583514451980591, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8613803386688232, "num_tokens": 518274294.0, "step": 13653 }, { "epoch": 1.736929143874825, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7514569759368896, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8816953897476196, "num_tokens": 518310960.0, "step": 13654 }, { "epoch": 1.7370563541534156, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5810587406158447, "learning_rate": 1e-05, "loss": 0.5183, "mean_token_accuracy": 0.8493176102638245, "num_tokens": 518352133.0, "step": 13655 }, { "epoch": 1.737183564432006, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7571463584899902, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8609223365783691, "num_tokens": 518388793.0, "step": 13656 }, { "epoch": 1.7373107747105965, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.566340684890747, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.878762423992157, "num_tokens": 518425581.0, "step": 13657 }, { "epoch": 1.737437984989187, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7521681785583496, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8733189702033997, "num_tokens": 518465477.0, "step": 13658 }, { "epoch": 1.7375651952677775, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.563356637954712, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8619240522384644, "num_tokens": 518506790.0, "step": 13659 }, { "epoch": 1.737692405546368, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7478485107421875, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8706249594688416, "num_tokens": 518548240.0, "step": 13660 }, { "epoch": 1.7378196158249586, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5846846103668213, "learning_rate": 1e-05, "loss": 0.4872, "mean_token_accuracy": 0.8585079908370972, "num_tokens": 518583865.0, "step": 13661 }, { "epoch": 1.7379468261035491, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7376651763916016, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8802973031997681, "num_tokens": 518623303.0, "step": 13662 }, { "epoch": 1.7380740363821396, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.570310592651367, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8683183789253235, "num_tokens": 518662375.0, "step": 13663 }, { "epoch": 1.7382012466607302, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7406232357025146, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8629676699638367, "num_tokens": 518702007.0, "step": 13664 }, { "epoch": 1.7383284569393207, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.559253215789795, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8808051347732544, "num_tokens": 518743676.0, "step": 13665 }, { "epoch": 1.7384556672179112, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7387917041778564, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8765485286712646, "num_tokens": 518783550.0, "step": 13666 }, { "epoch": 1.7385828774965018, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5671002864837646, "learning_rate": 1e-05, "loss": 0.3922, "mean_token_accuracy": 0.8872082829475403, "num_tokens": 518820788.0, "step": 13667 }, { "epoch": 1.7387100877750923, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7413601875305176, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8604030609130859, "num_tokens": 518865195.0, "step": 13668 }, { "epoch": 1.7388372980536828, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5759499073028564, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8672027587890625, "num_tokens": 518903683.0, "step": 13669 }, { "epoch": 1.7389645083322733, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.73453688621521, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8804653882980347, "num_tokens": 518939535.0, "step": 13670 }, { "epoch": 1.7390917186108639, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5759353637695312, "learning_rate": 1e-05, "loss": 0.4938, "mean_token_accuracy": 0.8561595678329468, "num_tokens": 518978753.0, "step": 13671 }, { "epoch": 1.7392189288894544, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.751621961593628, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8576346635818481, "num_tokens": 519014916.0, "step": 13672 }, { "epoch": 1.739346139168045, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5931646823883057, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8661110401153564, "num_tokens": 519043537.0, "step": 13673 }, { "epoch": 1.7394733494466355, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.721416711807251, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8724485635757446, "num_tokens": 519081388.0, "step": 13674 }, { "epoch": 1.7396005597252258, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5784873962402344, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8806585073471069, "num_tokens": 519119531.0, "step": 13675 }, { "epoch": 1.7397277700038163, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.72066068649292, "learning_rate": 1e-05, "loss": 0.5011, "mean_token_accuracy": 0.8549164533615112, "num_tokens": 519157982.0, "step": 13676 }, { "epoch": 1.7398549802824068, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5759077072143555, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8644169569015503, "num_tokens": 519198928.0, "step": 13677 }, { "epoch": 1.7399821905609973, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7327985763549805, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.864284873008728, "num_tokens": 519237760.0, "step": 13678 }, { "epoch": 1.7401094008395879, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5633225440979004, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8714447021484375, "num_tokens": 519273557.0, "step": 13679 }, { "epoch": 1.7402366111181784, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.725945472717285, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8658692240715027, "num_tokens": 519316082.0, "step": 13680 }, { "epoch": 1.7403638213967687, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5578970909118652, "learning_rate": 1e-05, "loss": 0.3983, "mean_token_accuracy": 0.8841007947921753, "num_tokens": 519352528.0, "step": 13681 }, { "epoch": 1.7404910316753592, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.739919424057007, "learning_rate": 1e-05, "loss": 0.3842, "mean_token_accuracy": 0.8894089460372925, "num_tokens": 519384931.0, "step": 13682 }, { "epoch": 1.7406182419539498, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5621585845947266, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8705992102622986, "num_tokens": 519421974.0, "step": 13683 }, { "epoch": 1.7407454522325403, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7378623485565186, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.877119243144989, "num_tokens": 519460623.0, "step": 13684 }, { "epoch": 1.7408726625111308, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5503976345062256, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.882323682308197, "num_tokens": 519497079.0, "step": 13685 }, { "epoch": 1.7409998727897213, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7477850914001465, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8544780015945435, "num_tokens": 519542680.0, "step": 13686 }, { "epoch": 1.7411270830683119, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5684094429016113, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.867594301700592, "num_tokens": 519577025.0, "step": 13687 }, { "epoch": 1.7412542933469024, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.749559164047241, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8672620058059692, "num_tokens": 519618616.0, "step": 13688 }, { "epoch": 1.741381503625493, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5581369400024414, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8661766648292542, "num_tokens": 519658672.0, "step": 13689 }, { "epoch": 1.7415087139040835, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.745131015777588, "learning_rate": 1e-05, "loss": 0.3806, "mean_token_accuracy": 0.8921664357185364, "num_tokens": 519692830.0, "step": 13690 }, { "epoch": 1.741635924182674, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.566619634628296, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8726701736450195, "num_tokens": 519729949.0, "step": 13691 }, { "epoch": 1.7417631344612645, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7582242488861084, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8681581616401672, "num_tokens": 519763556.0, "step": 13692 }, { "epoch": 1.741890344739855, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.556450366973877, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8612167835235596, "num_tokens": 519804556.0, "step": 13693 }, { "epoch": 1.7420175550184456, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.751476526260376, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8750014901161194, "num_tokens": 519845846.0, "step": 13694 }, { "epoch": 1.742144765297036, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5655276775360107, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8632859587669373, "num_tokens": 519883333.0, "step": 13695 }, { "epoch": 1.7422719755756266, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.764259099960327, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8806406259536743, "num_tokens": 519915927.0, "step": 13696 }, { "epoch": 1.7423991858542172, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5666568279266357, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8712088465690613, "num_tokens": 519951555.0, "step": 13697 }, { "epoch": 1.7425263961328077, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.745129346847534, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8814365863800049, "num_tokens": 519996616.0, "step": 13698 }, { "epoch": 1.742653606411398, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5836169719696045, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8548998832702637, "num_tokens": 520032467.0, "step": 13699 }, { "epoch": 1.7427808166899885, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.749187707901001, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8727604150772095, "num_tokens": 520068893.0, "step": 13700 }, { "epoch": 1.742908026968579, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5869619846343994, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8706135153770447, "num_tokens": 520104745.0, "step": 13701 }, { "epoch": 1.7430352372471696, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.73701810836792, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8566575646400452, "num_tokens": 520146994.0, "step": 13702 }, { "epoch": 1.74316244752576, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.563131093978882, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8797513246536255, "num_tokens": 520188220.0, "step": 13703 }, { "epoch": 1.7432896578043506, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7396345138549805, "learning_rate": 1e-05, "loss": 0.3911, "mean_token_accuracy": 0.8875000476837158, "num_tokens": 520223410.0, "step": 13704 }, { "epoch": 1.743416868082941, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.566303253173828, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8686904311180115, "num_tokens": 520260017.0, "step": 13705 }, { "epoch": 1.7435440783615315, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.754937171936035, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8772232532501221, "num_tokens": 520296232.0, "step": 13706 }, { "epoch": 1.743671288640122, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5796473026275635, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8597018718719482, "num_tokens": 520330720.0, "step": 13707 }, { "epoch": 1.7437984989187125, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7525737285614014, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8779809474945068, "num_tokens": 520362438.0, "step": 13708 }, { "epoch": 1.743925709197303, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.560368061065674, "learning_rate": 1e-05, "loss": 0.3928, "mean_token_accuracy": 0.8852216601371765, "num_tokens": 520398170.0, "step": 13709 }, { "epoch": 1.7440529194758936, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.75431752204895, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8543094396591187, "num_tokens": 520438219.0, "step": 13710 }, { "epoch": 1.744180129754484, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5514097213745117, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8752824068069458, "num_tokens": 520481642.0, "step": 13711 }, { "epoch": 1.7443073400330746, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7752654552459717, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8726037740707397, "num_tokens": 520515945.0, "step": 13712 }, { "epoch": 1.7444345503116652, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5617997646331787, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8631252646446228, "num_tokens": 520553259.0, "step": 13713 }, { "epoch": 1.7445617605902557, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.770775556564331, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.88177889585495, "num_tokens": 520590371.0, "step": 13714 }, { "epoch": 1.7446889708688462, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.568476915359497, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.871584415435791, "num_tokens": 520624738.0, "step": 13715 }, { "epoch": 1.7448161811474368, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.768789291381836, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.883867621421814, "num_tokens": 520666631.0, "step": 13716 }, { "epoch": 1.7449433914260273, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.582148551940918, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8779835104942322, "num_tokens": 520696567.0, "step": 13717 }, { "epoch": 1.7450706017046178, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.755277156829834, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8671001195907593, "num_tokens": 520737975.0, "step": 13718 }, { "epoch": 1.7451978119832083, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.560231924057007, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8835057020187378, "num_tokens": 520776884.0, "step": 13719 }, { "epoch": 1.7453250222617989, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7714476585388184, "learning_rate": 1e-05, "loss": 0.491, "mean_token_accuracy": 0.8589336276054382, "num_tokens": 520817770.0, "step": 13720 }, { "epoch": 1.7454522325403894, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5599889755249023, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8722110390663147, "num_tokens": 520859381.0, "step": 13721 }, { "epoch": 1.74557944281898, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.75498628616333, "learning_rate": 1e-05, "loss": 0.3775, "mean_token_accuracy": 0.8931381702423096, "num_tokens": 520900007.0, "step": 13722 }, { "epoch": 1.7457066530975704, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5650129318237305, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8779640197753906, "num_tokens": 520940707.0, "step": 13723 }, { "epoch": 1.7458338633761608, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.782940149307251, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8642442226409912, "num_tokens": 520972565.0, "step": 13724 }, { "epoch": 1.7459610736547513, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5606937408447266, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8651615381240845, "num_tokens": 521012364.0, "step": 13725 }, { "epoch": 1.7460882839333418, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7610280513763428, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8840806484222412, "num_tokens": 521058761.0, "step": 13726 }, { "epoch": 1.7462154942119323, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5753865242004395, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.871056079864502, "num_tokens": 521096620.0, "step": 13727 }, { "epoch": 1.7463427044905229, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7747135162353516, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8672685623168945, "num_tokens": 521133223.0, "step": 13728 }, { "epoch": 1.7464699147691134, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.564582109451294, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8841015100479126, "num_tokens": 521170334.0, "step": 13729 }, { "epoch": 1.7465971250477037, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7651562690734863, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8778421878814697, "num_tokens": 521207458.0, "step": 13730 }, { "epoch": 1.7467243353262942, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.572225570678711, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8732208013534546, "num_tokens": 521245017.0, "step": 13731 }, { "epoch": 1.7468515456048848, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.759920120239258, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8671300411224365, "num_tokens": 521285792.0, "step": 13732 }, { "epoch": 1.7469787558834753, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5563855171203613, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8818731307983398, "num_tokens": 521327107.0, "step": 13733 }, { "epoch": 1.7471059661620658, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.765031576156616, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8817107677459717, "num_tokens": 521364720.0, "step": 13734 }, { "epoch": 1.7472331764406563, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.571105718612671, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.8797045946121216, "num_tokens": 521398183.0, "step": 13735 }, { "epoch": 1.7473603867192469, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.77132511138916, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8704245090484619, "num_tokens": 521436837.0, "step": 13736 }, { "epoch": 1.7474875969978374, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.554457426071167, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8713076710700989, "num_tokens": 521477722.0, "step": 13737 }, { "epoch": 1.747614807276428, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.777992010116577, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8777571320533752, "num_tokens": 521513647.0, "step": 13738 }, { "epoch": 1.7477420175550185, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5795350074768066, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8721128702163696, "num_tokens": 521545480.0, "step": 13739 }, { "epoch": 1.747869227833609, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7588131427764893, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8602123856544495, "num_tokens": 521587363.0, "step": 13740 }, { "epoch": 1.7479964381121995, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.573728561401367, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8655847311019897, "num_tokens": 521629399.0, "step": 13741 }, { "epoch": 1.74812364839079, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7585675716400146, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8693017959594727, "num_tokens": 521667065.0, "step": 13742 }, { "epoch": 1.7482508586693806, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5869855880737305, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.877655565738678, "num_tokens": 521696317.0, "step": 13743 }, { "epoch": 1.748378068947971, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.741908550262451, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8769222497940063, "num_tokens": 521736948.0, "step": 13744 }, { "epoch": 1.7485052792265616, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5789577960968018, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.8836257457733154, "num_tokens": 521770320.0, "step": 13745 }, { "epoch": 1.7486324895051522, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7317280769348145, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.8851428627967834, "num_tokens": 521803836.0, "step": 13746 }, { "epoch": 1.7487596997837427, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.600153923034668, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.859610915184021, "num_tokens": 521838066.0, "step": 13747 }, { "epoch": 1.748886910062333, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7074828147888184, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8758207559585571, "num_tokens": 521876806.0, "step": 13748 }, { "epoch": 1.7490141203409235, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5951554775238037, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8646823763847351, "num_tokens": 521916241.0, "step": 13749 }, { "epoch": 1.749141330619514, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7268574237823486, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.87677001953125, "num_tokens": 521949369.0, "step": 13750 }, { "epoch": 1.7492685408981046, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5889480113983154, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8599863052368164, "num_tokens": 521989251.0, "step": 13751 }, { "epoch": 1.749395751176695, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7243592739105225, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8777886033058167, "num_tokens": 522019221.0, "step": 13752 }, { "epoch": 1.7495229614552856, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.569793701171875, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8680992126464844, "num_tokens": 522058858.0, "step": 13753 }, { "epoch": 1.749650171733876, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7426066398620605, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.872300386428833, "num_tokens": 522089834.0, "step": 13754 }, { "epoch": 1.7497773820124665, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.565600872039795, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8798397779464722, "num_tokens": 522124474.0, "step": 13755 }, { "epoch": 1.749904592291057, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7470808029174805, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8644041419029236, "num_tokens": 522159046.0, "step": 13756 }, { "epoch": 1.7500318025696475, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.555931806564331, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8703079223632812, "num_tokens": 522200698.0, "step": 13757 }, { "epoch": 1.750159012848238, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.778017997741699, "learning_rate": 1e-05, "loss": 0.479, "mean_token_accuracy": 0.8574506044387817, "num_tokens": 522233533.0, "step": 13758 }, { "epoch": 1.7502862231268286, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.559403896331787, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.861933708190918, "num_tokens": 522274234.0, "step": 13759 }, { "epoch": 1.750413433405419, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.773808240890503, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8735125064849854, "num_tokens": 522313055.0, "step": 13760 }, { "epoch": 1.7505406436840096, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.561251401901245, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8732796907424927, "num_tokens": 522346372.0, "step": 13761 }, { "epoch": 1.7506678539626002, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.783245086669922, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8736509084701538, "num_tokens": 522381199.0, "step": 13762 }, { "epoch": 1.7507950642411907, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5679714679718018, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8538955450057983, "num_tokens": 522419509.0, "step": 13763 }, { "epoch": 1.7509222745197812, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.76969313621521, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8721609115600586, "num_tokens": 522458433.0, "step": 13764 }, { "epoch": 1.7510494847983717, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5723824501037598, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8714014887809753, "num_tokens": 522492010.0, "step": 13765 }, { "epoch": 1.7511766950769623, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7872135639190674, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8741584420204163, "num_tokens": 522524190.0, "step": 13766 }, { "epoch": 1.7513039053555528, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5824904441833496, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8587339520454407, "num_tokens": 522558218.0, "step": 13767 }, { "epoch": 1.7514311156341433, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.787222146987915, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8745606541633606, "num_tokens": 522594683.0, "step": 13768 }, { "epoch": 1.7515583259127339, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.564023017883301, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8769879341125488, "num_tokens": 522633409.0, "step": 13769 }, { "epoch": 1.7516855361913244, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7796266078948975, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.86860191822052, "num_tokens": 522672469.0, "step": 13770 }, { "epoch": 1.751812746469915, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5588715076446533, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8749340772628784, "num_tokens": 522714676.0, "step": 13771 }, { "epoch": 1.7519399567485054, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7837836742401123, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.8831166625022888, "num_tokens": 522748726.0, "step": 13772 }, { "epoch": 1.7520671670270958, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6003003120422363, "learning_rate": 1e-05, "loss": 0.513, "mean_token_accuracy": 0.8522855043411255, "num_tokens": 522783761.0, "step": 13773 }, { "epoch": 1.7521943773056863, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7588095664978027, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8790884017944336, "num_tokens": 522822027.0, "step": 13774 }, { "epoch": 1.7523215875842768, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5784285068511963, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8777934312820435, "num_tokens": 522856952.0, "step": 13775 }, { "epoch": 1.7524487978628673, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.76322603225708, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8650186657905579, "num_tokens": 522897450.0, "step": 13776 }, { "epoch": 1.7525760081414579, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5899345874786377, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8763383626937866, "num_tokens": 522930547.0, "step": 13777 }, { "epoch": 1.7527032184200484, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.745600461959839, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8694530129432678, "num_tokens": 522970145.0, "step": 13778 }, { "epoch": 1.7528304286986387, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5785515308380127, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.8862769603729248, "num_tokens": 523006492.0, "step": 13779 }, { "epoch": 1.7529576389772292, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7426395416259766, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8627784252166748, "num_tokens": 523045524.0, "step": 13780 }, { "epoch": 1.7530848492558198, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5791025161743164, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8745242357254028, "num_tokens": 523086091.0, "step": 13781 }, { "epoch": 1.7532120595344103, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.728353977203369, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8774291276931763, "num_tokens": 523127351.0, "step": 13782 }, { "epoch": 1.7533392698130008, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5930564403533936, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8693137168884277, "num_tokens": 523161775.0, "step": 13783 }, { "epoch": 1.7534664800915913, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7322723865509033, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8674639463424683, "num_tokens": 523198977.0, "step": 13784 }, { "epoch": 1.7535936903701819, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.588803768157959, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8661563396453857, "num_tokens": 523233917.0, "step": 13785 }, { "epoch": 1.7537209006487724, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7293899059295654, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8762166500091553, "num_tokens": 523267679.0, "step": 13786 }, { "epoch": 1.753848110927363, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5927600860595703, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8602227568626404, "num_tokens": 523301512.0, "step": 13787 }, { "epoch": 1.7539753212059535, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7022438049316406, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.878941535949707, "num_tokens": 523342618.0, "step": 13788 }, { "epoch": 1.754102531484544, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6009275913238525, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8725707530975342, "num_tokens": 523379555.0, "step": 13789 }, { "epoch": 1.7542297417631345, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6943953037261963, "learning_rate": 1e-05, "loss": 0.3826, "mean_token_accuracy": 0.8913456201553345, "num_tokens": 523414621.0, "step": 13790 }, { "epoch": 1.754356952041725, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.572573184967041, "learning_rate": 1e-05, "loss": 0.3737, "mean_token_accuracy": 0.8937115669250488, "num_tokens": 523459022.0, "step": 13791 }, { "epoch": 1.7544841623203156, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6941351890563965, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.8857372999191284, "num_tokens": 523498718.0, "step": 13792 }, { "epoch": 1.754611372598906, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5879740715026855, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8838833570480347, "num_tokens": 523531926.0, "step": 13793 }, { "epoch": 1.7547385828774966, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.69696044921875, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8731249570846558, "num_tokens": 523572457.0, "step": 13794 }, { "epoch": 1.7548657931560872, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5952517986297607, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8642792701721191, "num_tokens": 523607317.0, "step": 13795 }, { "epoch": 1.7549930034346777, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7052249908447266, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8658113479614258, "num_tokens": 523638979.0, "step": 13796 }, { "epoch": 1.755120213713268, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.566391706466675, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8737651109695435, "num_tokens": 523681839.0, "step": 13797 }, { "epoch": 1.7552474239918585, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7051234245300293, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8780709505081177, "num_tokens": 523719797.0, "step": 13798 }, { "epoch": 1.755374634270449, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5552244186401367, "learning_rate": 1e-05, "loss": 0.4044, "mean_token_accuracy": 0.8783758282661438, "num_tokens": 523761774.0, "step": 13799 }, { "epoch": 1.7555018445490396, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7252020835876465, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8733598589897156, "num_tokens": 523799522.0, "step": 13800 }, { "epoch": 1.75562905482763, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.568803071975708, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.874962568283081, "num_tokens": 523831927.0, "step": 13801 }, { "epoch": 1.7557562651062206, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.759211778640747, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8638952374458313, "num_tokens": 523863058.0, "step": 13802 }, { "epoch": 1.755883475384811, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.55064058303833, "learning_rate": 1e-05, "loss": 0.4958, "mean_token_accuracy": 0.8511672019958496, "num_tokens": 523907017.0, "step": 13803 }, { "epoch": 1.7560106856634015, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7640490531921387, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8686412572860718, "num_tokens": 523943376.0, "step": 13804 }, { "epoch": 1.756137895941992, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5424115657806396, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8805180788040161, "num_tokens": 523982902.0, "step": 13805 }, { "epoch": 1.7562651062205825, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7922778129577637, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8654236793518066, "num_tokens": 524016423.0, "step": 13806 }, { "epoch": 1.756392316499173, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5461175441741943, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8565412163734436, "num_tokens": 524056686.0, "step": 13807 }, { "epoch": 1.7565195267777636, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7956998348236084, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.87127685546875, "num_tokens": 524093731.0, "step": 13808 }, { "epoch": 1.756646737056354, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5413055419921875, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8704028129577637, "num_tokens": 524139398.0, "step": 13809 }, { "epoch": 1.7567739473349446, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8170061111450195, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.863976001739502, "num_tokens": 524177012.0, "step": 13810 }, { "epoch": 1.7569011576135352, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5513012409210205, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8862895369529724, "num_tokens": 524211785.0, "step": 13811 }, { "epoch": 1.7570283678921257, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8206539154052734, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8845934867858887, "num_tokens": 524252263.0, "step": 13812 }, { "epoch": 1.7571555781707162, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.579131603240967, "learning_rate": 1e-05, "loss": 0.3801, "mean_token_accuracy": 0.8894342184066772, "num_tokens": 524286561.0, "step": 13813 }, { "epoch": 1.7572827884493067, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.823035717010498, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8766652941703796, "num_tokens": 524320910.0, "step": 13814 }, { "epoch": 1.7574099987278973, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.561455726623535, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.880285382270813, "num_tokens": 524362416.0, "step": 13815 }, { "epoch": 1.7575372090064878, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.808964729309082, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8796771764755249, "num_tokens": 524405432.0, "step": 13816 }, { "epoch": 1.7576644192850783, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5797159671783447, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8753023743629456, "num_tokens": 524440605.0, "step": 13817 }, { "epoch": 1.7577916295636689, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.810432195663452, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8700529336929321, "num_tokens": 524483126.0, "step": 13818 }, { "epoch": 1.7579188398422594, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5948679447174072, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8783507347106934, "num_tokens": 524516675.0, "step": 13819 }, { "epoch": 1.75804605012085, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7798898220062256, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8756903409957886, "num_tokens": 524554394.0, "step": 13820 }, { "epoch": 1.7581732603994404, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.587707042694092, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8693106174468994, "num_tokens": 524591195.0, "step": 13821 }, { "epoch": 1.7583004706780307, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.775994062423706, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8639355897903442, "num_tokens": 524629077.0, "step": 13822 }, { "epoch": 1.7584276809566213, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.581310987472534, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8553541898727417, "num_tokens": 524672072.0, "step": 13823 }, { "epoch": 1.7585548912352118, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.781121253967285, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.874036431312561, "num_tokens": 524706735.0, "step": 13824 }, { "epoch": 1.7586821015138023, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.588731288909912, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8771833181381226, "num_tokens": 524741750.0, "step": 13825 }, { "epoch": 1.7588093117923929, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7600111961364746, "learning_rate": 1e-05, "loss": 0.39, "mean_token_accuracy": 0.888750433921814, "num_tokens": 524774565.0, "step": 13826 }, { "epoch": 1.7589365220709834, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5910072326660156, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8824158310890198, "num_tokens": 524803731.0, "step": 13827 }, { "epoch": 1.7590637323495737, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.755422353744507, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8695401549339294, "num_tokens": 524842570.0, "step": 13828 }, { "epoch": 1.7591909426281642, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.585895538330078, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8728858828544617, "num_tokens": 524880695.0, "step": 13829 }, { "epoch": 1.7593181529067548, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.734665632247925, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8748003244400024, "num_tokens": 524921331.0, "step": 13830 }, { "epoch": 1.7594453631853453, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5927515029907227, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8658371567726135, "num_tokens": 524955769.0, "step": 13831 }, { "epoch": 1.7595725734639358, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.723515272140503, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8588281869888306, "num_tokens": 524998299.0, "step": 13832 }, { "epoch": 1.7596997837425263, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5937585830688477, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8604861497879028, "num_tokens": 525038590.0, "step": 13833 }, { "epoch": 1.7598269940211169, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7216854095458984, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8719128966331482, "num_tokens": 525073688.0, "step": 13834 }, { "epoch": 1.7599542042997074, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.572718620300293, "learning_rate": 1e-05, "loss": 0.377, "mean_token_accuracy": 0.8916476368904114, "num_tokens": 525112109.0, "step": 13835 }, { "epoch": 1.760081414578298, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7344248294830322, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8769994974136353, "num_tokens": 525147554.0, "step": 13836 }, { "epoch": 1.7602086248568884, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5705015659332275, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8605167865753174, "num_tokens": 525187233.0, "step": 13837 }, { "epoch": 1.760335835135479, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7558677196502686, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8632062077522278, "num_tokens": 525218104.0, "step": 13838 }, { "epoch": 1.7604630454140695, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5602681636810303, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.877577543258667, "num_tokens": 525255690.0, "step": 13839 }, { "epoch": 1.76059025569266, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7601919174194336, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8724547624588013, "num_tokens": 525287972.0, "step": 13840 }, { "epoch": 1.7607174659712506, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5576183795928955, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8646177053451538, "num_tokens": 525328003.0, "step": 13841 }, { "epoch": 1.760844676249841, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7533135414123535, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8786271810531616, "num_tokens": 525370642.0, "step": 13842 }, { "epoch": 1.7609718865284316, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.567150831222534, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.859462559223175, "num_tokens": 525408668.0, "step": 13843 }, { "epoch": 1.7610990968070221, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.757572650909424, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8722939491271973, "num_tokens": 525447566.0, "step": 13844 }, { "epoch": 1.7612263070856127, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5643210411071777, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8654876351356506, "num_tokens": 525485456.0, "step": 13845 }, { "epoch": 1.761353517364203, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7733912467956543, "learning_rate": 1e-05, "loss": 0.4909, "mean_token_accuracy": 0.8571975231170654, "num_tokens": 525521190.0, "step": 13846 }, { "epoch": 1.7614807276427935, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5584113597869873, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8730407953262329, "num_tokens": 525559331.0, "step": 13847 }, { "epoch": 1.761607937921384, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7756426334381104, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8638622760772705, "num_tokens": 525598380.0, "step": 13848 }, { "epoch": 1.7617351481999746, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.553459882736206, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8634101152420044, "num_tokens": 525637901.0, "step": 13849 }, { "epoch": 1.761862358478565, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.794682502746582, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8673382997512817, "num_tokens": 525676072.0, "step": 13850 }, { "epoch": 1.7619895687571556, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5554661750793457, "learning_rate": 1e-05, "loss": 0.383, "mean_token_accuracy": 0.8885869979858398, "num_tokens": 525710655.0, "step": 13851 }, { "epoch": 1.762116779035746, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.795889139175415, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8772746324539185, "num_tokens": 525745714.0, "step": 13852 }, { "epoch": 1.7622439893143365, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.550328016281128, "learning_rate": 1e-05, "loss": 0.3871, "mean_token_accuracy": 0.8869520425796509, "num_tokens": 525782246.0, "step": 13853 }, { "epoch": 1.762371199592927, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.816533327102661, "learning_rate": 1e-05, "loss": 0.3926, "mean_token_accuracy": 0.8870323896408081, "num_tokens": 525814232.0, "step": 13854 }, { "epoch": 1.7624984098715175, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.571119546890259, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8715547323226929, "num_tokens": 525847775.0, "step": 13855 }, { "epoch": 1.762625620150108, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.797705888748169, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8744088411331177, "num_tokens": 525890760.0, "step": 13856 }, { "epoch": 1.7627528304286986, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5733652114868164, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.8822678923606873, "num_tokens": 525926893.0, "step": 13857 }, { "epoch": 1.762880040707289, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7948813438415527, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8635869026184082, "num_tokens": 525968907.0, "step": 13858 }, { "epoch": 1.7630072509858796, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5781824588775635, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8766734600067139, "num_tokens": 526003988.0, "step": 13859 }, { "epoch": 1.7631344612644702, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7810730934143066, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8756403923034668, "num_tokens": 526044575.0, "step": 13860 }, { "epoch": 1.7632616715430607, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.572916269302368, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.873833417892456, "num_tokens": 526087803.0, "step": 13861 }, { "epoch": 1.7633888818216512, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7898552417755127, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8572763204574585, "num_tokens": 526123432.0, "step": 13862 }, { "epoch": 1.7635160921002417, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.573845148086548, "learning_rate": 1e-05, "loss": 0.4008, "mean_token_accuracy": 0.8837555646896362, "num_tokens": 526161285.0, "step": 13863 }, { "epoch": 1.7636433023788323, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.789562940597534, "learning_rate": 1e-05, "loss": 0.5032, "mean_token_accuracy": 0.8565633296966553, "num_tokens": 526203096.0, "step": 13864 }, { "epoch": 1.7637705126574228, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5772864818573, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8843337893486023, "num_tokens": 526237741.0, "step": 13865 }, { "epoch": 1.7638977229360133, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.774641513824463, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.873682975769043, "num_tokens": 526279704.0, "step": 13866 }, { "epoch": 1.7640249332146039, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5843377113342285, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8708964586257935, "num_tokens": 526315708.0, "step": 13867 }, { "epoch": 1.7641521434931944, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.763950824737549, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8698849678039551, "num_tokens": 526360725.0, "step": 13868 }, { "epoch": 1.764279353771785, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5870273113250732, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8668234348297119, "num_tokens": 526398920.0, "step": 13869 }, { "epoch": 1.7644065640503754, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.750401020050049, "learning_rate": 1e-05, "loss": 0.3856, "mean_token_accuracy": 0.887191653251648, "num_tokens": 526433453.0, "step": 13870 }, { "epoch": 1.7645337743289657, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.571516990661621, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.8806292414665222, "num_tokens": 526476256.0, "step": 13871 }, { "epoch": 1.7646609846075563, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.753621816635132, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8745189905166626, "num_tokens": 526517707.0, "step": 13872 }, { "epoch": 1.7647881948861468, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.589275360107422, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8746013641357422, "num_tokens": 526551696.0, "step": 13873 }, { "epoch": 1.7649154051647373, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.741821527481079, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8827886581420898, "num_tokens": 526587884.0, "step": 13874 }, { "epoch": 1.7650426154433279, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.600680351257324, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8688256740570068, "num_tokens": 526625398.0, "step": 13875 }, { "epoch": 1.7651698257219184, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.718198299407959, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8702954649925232, "num_tokens": 526662481.0, "step": 13876 }, { "epoch": 1.7652970360005087, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5966875553131104, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.862481951713562, "num_tokens": 526700719.0, "step": 13877 }, { "epoch": 1.7654242462790992, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.723628282546997, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.873204231262207, "num_tokens": 526732941.0, "step": 13878 }, { "epoch": 1.7655514565576897, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5682880878448486, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8741756677627563, "num_tokens": 526777825.0, "step": 13879 }, { "epoch": 1.7656786668362803, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.739298105239868, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8659989237785339, "num_tokens": 526811331.0, "step": 13880 }, { "epoch": 1.7658058771148708, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5798866748809814, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8729913830757141, "num_tokens": 526846788.0, "step": 13881 }, { "epoch": 1.7659330873934613, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7369742393493652, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8803466558456421, "num_tokens": 526880127.0, "step": 13882 }, { "epoch": 1.7660602976720519, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5831356048583984, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.864011287689209, "num_tokens": 526914099.0, "step": 13883 }, { "epoch": 1.7661875079506424, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7221779823303223, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8649818897247314, "num_tokens": 526956409.0, "step": 13884 }, { "epoch": 1.766314718229233, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5801453590393066, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8715301752090454, "num_tokens": 526995175.0, "step": 13885 }, { "epoch": 1.7664419285078234, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.720313549041748, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8721540570259094, "num_tokens": 527035833.0, "step": 13886 }, { "epoch": 1.766569138786414, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.569952964782715, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8727608919143677, "num_tokens": 527078780.0, "step": 13887 }, { "epoch": 1.7666963490650045, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7314836978912354, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8797627091407776, "num_tokens": 527119494.0, "step": 13888 }, { "epoch": 1.766823559343595, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.57331919670105, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8721736073493958, "num_tokens": 527158112.0, "step": 13889 }, { "epoch": 1.7669507696221856, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7305686473846436, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8754420280456543, "num_tokens": 527195885.0, "step": 13890 }, { "epoch": 1.767077979900776, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5786547660827637, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8729649782180786, "num_tokens": 527229681.0, "step": 13891 }, { "epoch": 1.7672051901793666, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7390031814575195, "learning_rate": 1e-05, "loss": 0.4989, "mean_token_accuracy": 0.8483877778053284, "num_tokens": 527267248.0, "step": 13892 }, { "epoch": 1.7673324004579571, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.680535078048706, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8737542629241943, "num_tokens": 527299695.0, "step": 13893 }, { "epoch": 1.7674596107365477, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.682960271835327, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8704113960266113, "num_tokens": 527334398.0, "step": 13894 }, { "epoch": 1.767586821015138, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6047558784484863, "learning_rate": 1e-05, "loss": 0.4858, "mean_token_accuracy": 0.8587802052497864, "num_tokens": 527376954.0, "step": 13895 }, { "epoch": 1.7677140312937285, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6497554779052734, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8708934187889099, "num_tokens": 527419737.0, "step": 13896 }, { "epoch": 1.767841241572319, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.621309757232666, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8527706265449524, "num_tokens": 527455874.0, "step": 13897 }, { "epoch": 1.7679684518509096, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6570708751678467, "learning_rate": 1e-05, "loss": 0.5131, "mean_token_accuracy": 0.8463919162750244, "num_tokens": 527494796.0, "step": 13898 }, { "epoch": 1.7680956621295, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6146926879882812, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8544730544090271, "num_tokens": 527536955.0, "step": 13899 }, { "epoch": 1.7682228724080906, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6660265922546387, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8697375059127808, "num_tokens": 527569472.0, "step": 13900 }, { "epoch": 1.768350082686681, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6059679985046387, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8740482926368713, "num_tokens": 527603641.0, "step": 13901 }, { "epoch": 1.7684772929652715, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6595892906188965, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8646971583366394, "num_tokens": 527642942.0, "step": 13902 }, { "epoch": 1.768604503243862, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.605358123779297, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8659078478813171, "num_tokens": 527682687.0, "step": 13903 }, { "epoch": 1.7687317135224525, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6507010459899902, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8837891221046448, "num_tokens": 527721355.0, "step": 13904 }, { "epoch": 1.768858923801043, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6036620140075684, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8728629350662231, "num_tokens": 527760887.0, "step": 13905 }, { "epoch": 1.7689861340796336, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6547670364379883, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8615213632583618, "num_tokens": 527808669.0, "step": 13906 }, { "epoch": 1.769113344358224, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6267926692962646, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8591912984848022, "num_tokens": 527840262.0, "step": 13907 }, { "epoch": 1.7692405546368146, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6547696590423584, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8621370792388916, "num_tokens": 527880822.0, "step": 13908 }, { "epoch": 1.7693677649154052, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.593620777130127, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8765795230865479, "num_tokens": 527923677.0, "step": 13909 }, { "epoch": 1.7694949751939957, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.653402328491211, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8587082624435425, "num_tokens": 527966004.0, "step": 13910 }, { "epoch": 1.7696221854725862, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.602067708969116, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8722519874572754, "num_tokens": 528004725.0, "step": 13911 }, { "epoch": 1.7697493957511767, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.652930736541748, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8704839944839478, "num_tokens": 528043262.0, "step": 13912 }, { "epoch": 1.7698766060297673, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.605172634124756, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8666939735412598, "num_tokens": 528082810.0, "step": 13913 }, { "epoch": 1.7700038163083578, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6688108444213867, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8652097582817078, "num_tokens": 528117672.0, "step": 13914 }, { "epoch": 1.7701310265869483, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5981929302215576, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8714849948883057, "num_tokens": 528155756.0, "step": 13915 }, { "epoch": 1.7702582368655388, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6579790115356445, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8792681694030762, "num_tokens": 528194664.0, "step": 13916 }, { "epoch": 1.7703854471441294, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5864579677581787, "learning_rate": 1e-05, "loss": 0.3986, "mean_token_accuracy": 0.8832083940505981, "num_tokens": 528234334.0, "step": 13917 }, { "epoch": 1.77051265742272, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6664953231811523, "learning_rate": 1e-05, "loss": 0.3909, "mean_token_accuracy": 0.8876699209213257, "num_tokens": 528270543.0, "step": 13918 }, { "epoch": 1.7706398677013102, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5874521732330322, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8647251725196838, "num_tokens": 528309698.0, "step": 13919 }, { "epoch": 1.7707670779799007, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6677706241607666, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8660648465156555, "num_tokens": 528356396.0, "step": 13920 }, { "epoch": 1.7708942882584913, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.597844123840332, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8641525506973267, "num_tokens": 528394542.0, "step": 13921 }, { "epoch": 1.7710214985370818, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6672115325927734, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8838284015655518, "num_tokens": 528435433.0, "step": 13922 }, { "epoch": 1.7711487088156723, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5940988063812256, "learning_rate": 1e-05, "loss": 0.3917, "mean_token_accuracy": 0.8866897225379944, "num_tokens": 528471152.0, "step": 13923 }, { "epoch": 1.7712759190942629, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6657090187072754, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8803298473358154, "num_tokens": 528505084.0, "step": 13924 }, { "epoch": 1.7714031293728534, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.610508680343628, "learning_rate": 1e-05, "loss": 0.527, "mean_token_accuracy": 0.8443882465362549, "num_tokens": 528545234.0, "step": 13925 }, { "epoch": 1.7715303396514437, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.659665107727051, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.875993549823761, "num_tokens": 528584099.0, "step": 13926 }, { "epoch": 1.7716575499300342, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.593583345413208, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8632121682167053, "num_tokens": 528624143.0, "step": 13927 }, { "epoch": 1.7717847602086247, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6606385707855225, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8710297346115112, "num_tokens": 528661521.0, "step": 13928 }, { "epoch": 1.7719119704872153, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.594524621963501, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8821111917495728, "num_tokens": 528699317.0, "step": 13929 }, { "epoch": 1.7720391807658058, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6727170944213867, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8581502437591553, "num_tokens": 528737277.0, "step": 13930 }, { "epoch": 1.7721663910443963, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5889878273010254, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8599979877471924, "num_tokens": 528777076.0, "step": 13931 }, { "epoch": 1.7722936013229869, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.669156312942505, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8726120591163635, "num_tokens": 528813940.0, "step": 13932 }, { "epoch": 1.7724208116015774, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5762810707092285, "learning_rate": 1e-05, "loss": 0.3801, "mean_token_accuracy": 0.8905689120292664, "num_tokens": 528852883.0, "step": 13933 }, { "epoch": 1.772548021880168, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6870675086975098, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8610748648643494, "num_tokens": 528891561.0, "step": 13934 }, { "epoch": 1.7726752321587584, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.57724928855896, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8678682446479797, "num_tokens": 528930755.0, "step": 13935 }, { "epoch": 1.772802442437349, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6996548175811768, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.867273211479187, "num_tokens": 528969230.0, "step": 13936 }, { "epoch": 1.7729296527159395, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5687921047210693, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.877447247505188, "num_tokens": 529005157.0, "step": 13937 }, { "epoch": 1.77305686299453, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7007346153259277, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8783766627311707, "num_tokens": 529042934.0, "step": 13938 }, { "epoch": 1.7731840732731206, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5601537227630615, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8743305206298828, "num_tokens": 529087811.0, "step": 13939 }, { "epoch": 1.773311283551711, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.719787120819092, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.861877977848053, "num_tokens": 529133143.0, "step": 13940 }, { "epoch": 1.7734384938303016, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5500566959381104, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8842625021934509, "num_tokens": 529172737.0, "step": 13941 }, { "epoch": 1.7735657041088921, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.737729787826538, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8731398582458496, "num_tokens": 529208913.0, "step": 13942 }, { "epoch": 1.7736929143874827, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.550431251525879, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8681281805038452, "num_tokens": 529250905.0, "step": 13943 }, { "epoch": 1.773820124666073, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7413594722747803, "learning_rate": 1e-05, "loss": 0.3891, "mean_token_accuracy": 0.8882200717926025, "num_tokens": 529289542.0, "step": 13944 }, { "epoch": 1.7739473349446635, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5557680130004883, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8794689178466797, "num_tokens": 529324759.0, "step": 13945 }, { "epoch": 1.774074545223254, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.759089231491089, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8712325096130371, "num_tokens": 529361845.0, "step": 13946 }, { "epoch": 1.7742017555018446, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.8623809814453125e-05, "grad_norm": 2.5435597896575928, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.871894896030426, "num_tokens": 529402220.0, "step": 13947 }, { "epoch": 1.774328965780435, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.784857988357544, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8589474558830261, "num_tokens": 529440532.0, "step": 13948 }, { "epoch": 1.7744561760590256, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.530057907104492, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.876623809337616, "num_tokens": 529480357.0, "step": 13949 }, { "epoch": 1.774583386337616, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8217406272888184, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8658144474029541, "num_tokens": 529514075.0, "step": 13950 }, { "epoch": 1.7747105966162064, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.551210403442383, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.8581499457359314, "num_tokens": 529550738.0, "step": 13951 }, { "epoch": 1.774837806894797, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.816472053527832, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8643838167190552, "num_tokens": 529594151.0, "step": 13952 }, { "epoch": 1.7749650171733875, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.543797492980957, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8751736283302307, "num_tokens": 529633631.0, "step": 13953 }, { "epoch": 1.775092227451978, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.825277328491211, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8686063289642334, "num_tokens": 529674377.0, "step": 13954 }, { "epoch": 1.7752194377305686, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5713860988616943, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8610268831253052, "num_tokens": 529707209.0, "step": 13955 }, { "epoch": 1.775346648009159, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8452956676483154, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8687780499458313, "num_tokens": 529737546.0, "step": 13956 }, { "epoch": 1.7754738582877496, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5715038776397705, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8723496198654175, "num_tokens": 529775532.0, "step": 13957 }, { "epoch": 1.7756010685663401, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.820713758468628, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8778645992279053, "num_tokens": 529814705.0, "step": 13958 }, { "epoch": 1.7757282788449307, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5806710720062256, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8713839054107666, "num_tokens": 529850912.0, "step": 13959 }, { "epoch": 1.7758554891235212, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8048510551452637, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8791346549987793, "num_tokens": 529889042.0, "step": 13960 }, { "epoch": 1.7759826994021117, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.585904598236084, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8737156391143799, "num_tokens": 529926749.0, "step": 13961 }, { "epoch": 1.7761099096807023, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.826592206954956, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8745468258857727, "num_tokens": 529951775.0, "step": 13962 }, { "epoch": 1.7762371199592928, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5631392002105713, "learning_rate": 1e-05, "loss": 0.3998, "mean_token_accuracy": 0.8858282566070557, "num_tokens": 529996172.0, "step": 13963 }, { "epoch": 1.7763643302378833, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.807337760925293, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8539942502975464, "num_tokens": 530039577.0, "step": 13964 }, { "epoch": 1.7764915405164738, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5890731811523438, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8815950155258179, "num_tokens": 530076516.0, "step": 13965 }, { "epoch": 1.7766187507950644, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.782240629196167, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8821104168891907, "num_tokens": 530118552.0, "step": 13966 }, { "epoch": 1.776745961073655, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5945842266082764, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8655362129211426, "num_tokens": 530154971.0, "step": 13967 }, { "epoch": 1.7768731713522452, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7689619064331055, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8842591047286987, "num_tokens": 530190509.0, "step": 13968 }, { "epoch": 1.7770003816308357, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.587533950805664, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8781324028968811, "num_tokens": 530229814.0, "step": 13969 }, { "epoch": 1.7771275919094263, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.765026569366455, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8661255240440369, "num_tokens": 530269426.0, "step": 13970 }, { "epoch": 1.7772548021880168, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5951528549194336, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8652662634849548, "num_tokens": 530305036.0, "step": 13971 }, { "epoch": 1.7773820124666073, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.744196653366089, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8695275783538818, "num_tokens": 530344995.0, "step": 13972 }, { "epoch": 1.7775092227451978, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6122965812683105, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8779287338256836, "num_tokens": 530378283.0, "step": 13973 }, { "epoch": 1.7776364330237884, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.732391119003296, "learning_rate": 1e-05, "loss": 0.495, "mean_token_accuracy": 0.8550474047660828, "num_tokens": 530416147.0, "step": 13974 }, { "epoch": 1.7777636433023787, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.591350793838501, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8700932264328003, "num_tokens": 530452096.0, "step": 13975 }, { "epoch": 1.7778908535809692, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7261362075805664, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8599322438240051, "num_tokens": 530489347.0, "step": 13976 }, { "epoch": 1.7780180638595597, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.587554693222046, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8672682046890259, "num_tokens": 530528265.0, "step": 13977 }, { "epoch": 1.7781452741381503, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7142741680145264, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8829237222671509, "num_tokens": 530567501.0, "step": 13978 }, { "epoch": 1.7782724844167408, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5798935890197754, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8701283931732178, "num_tokens": 530610038.0, "step": 13979 }, { "epoch": 1.7783996946953313, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.72525691986084, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8675196170806885, "num_tokens": 530649039.0, "step": 13980 }, { "epoch": 1.7785269049739219, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5836682319641113, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8760680556297302, "num_tokens": 530687142.0, "step": 13981 }, { "epoch": 1.7786541152525124, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7131705284118652, "learning_rate": 1e-05, "loss": 0.4017, "mean_token_accuracy": 0.8807575702667236, "num_tokens": 530724399.0, "step": 13982 }, { "epoch": 1.778781325531103, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.573352813720703, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8712506294250488, "num_tokens": 530764552.0, "step": 13983 }, { "epoch": 1.7789085358096934, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7108566761016846, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8667309284210205, "num_tokens": 530808587.0, "step": 13984 }, { "epoch": 1.779035746088284, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.588261604309082, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8674138784408569, "num_tokens": 530845887.0, "step": 13985 }, { "epoch": 1.7791629563668745, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.721639394760132, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.860149085521698, "num_tokens": 530887398.0, "step": 13986 }, { "epoch": 1.779290166645465, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.601962089538574, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8816918134689331, "num_tokens": 530916377.0, "step": 13987 }, { "epoch": 1.7794173769240555, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.713038206100464, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8606328368186951, "num_tokens": 530954651.0, "step": 13988 }, { "epoch": 1.779544587202646, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5795438289642334, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8612611889839172, "num_tokens": 530999406.0, "step": 13989 }, { "epoch": 1.7796717974812366, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7000019550323486, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8776001930236816, "num_tokens": 531036925.0, "step": 13990 }, { "epoch": 1.7797990077598271, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5725088119506836, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8732677698135376, "num_tokens": 531076168.0, "step": 13991 }, { "epoch": 1.7799262180384177, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7189018726348877, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8742557764053345, "num_tokens": 531113217.0, "step": 13992 }, { "epoch": 1.780053428317008, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5614495277404785, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8706233501434326, "num_tokens": 531155938.0, "step": 13993 }, { "epoch": 1.7801806385955985, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.73055100440979, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8739833235740662, "num_tokens": 531194556.0, "step": 13994 }, { "epoch": 1.780307848874189, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.572953939437866, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.859775185585022, "num_tokens": 531230704.0, "step": 13995 }, { "epoch": 1.7804350591527796, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7261576652526855, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8744604587554932, "num_tokens": 531270672.0, "step": 13996 }, { "epoch": 1.78056226943137, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5720009803771973, "learning_rate": 1e-05, "loss": 0.3917, "mean_token_accuracy": 0.8817346096038818, "num_tokens": 531302043.0, "step": 13997 }, { "epoch": 1.7806894797099606, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.732032537460327, "learning_rate": 1e-05, "loss": 0.3956, "mean_token_accuracy": 0.8858551979064941, "num_tokens": 531337014.0, "step": 13998 }, { "epoch": 1.780816689988551, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.568673849105835, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8761985301971436, "num_tokens": 531371537.0, "step": 13999 }, { "epoch": 1.7809439002671414, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.734416961669922, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8817621469497681, "num_tokens": 531406171.0, "step": 14000 }, { "epoch": 1.781071110545732, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.570223569869995, "learning_rate": 1e-05, "loss": 0.4875, "mean_token_accuracy": 0.8590105772018433, "num_tokens": 531445625.0, "step": 14001 }, { "epoch": 1.7811983208243225, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7274606227874756, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8813479542732239, "num_tokens": 531490748.0, "step": 14002 }, { "epoch": 1.781325531102913, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5736167430877686, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8717317581176758, "num_tokens": 531526142.0, "step": 14003 }, { "epoch": 1.7814527413815036, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7334446907043457, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8756873607635498, "num_tokens": 531568325.0, "step": 14004 }, { "epoch": 1.781579951660094, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5681276321411133, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8689901828765869, "num_tokens": 531608898.0, "step": 14005 }, { "epoch": 1.7817071619386846, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7399535179138184, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8725207448005676, "num_tokens": 531640860.0, "step": 14006 }, { "epoch": 1.7818343722172751, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5644338130950928, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.875340461730957, "num_tokens": 531680121.0, "step": 14007 }, { "epoch": 1.7819615824958657, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7699134349823, "learning_rate": 1e-05, "loss": 0.4948, "mean_token_accuracy": 0.8536834716796875, "num_tokens": 531713496.0, "step": 14008 }, { "epoch": 1.7820887927744562, "ewc_loss": 0.06591796875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5600452423095703, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8635495901107788, "num_tokens": 531749989.0, "step": 14009 }, { "epoch": 1.7822160030530467, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7730746269226074, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8692681789398193, "num_tokens": 531788032.0, "step": 14010 }, { "epoch": 1.7823432133316373, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.574841022491455, "learning_rate": 1e-05, "loss": 0.4963, "mean_token_accuracy": 0.8494331240653992, "num_tokens": 531822178.0, "step": 14011 }, { "epoch": 1.7824704236102278, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7785801887512207, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8626266121864319, "num_tokens": 531855713.0, "step": 14012 }, { "epoch": 1.7825976338888183, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5555920600891113, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8603734374046326, "num_tokens": 531898714.0, "step": 14013 }, { "epoch": 1.7827248441674088, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.787893295288086, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.867303729057312, "num_tokens": 531933968.0, "step": 14014 }, { "epoch": 1.7828520544459994, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.547187566757202, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8721673488616943, "num_tokens": 531980672.0, "step": 14015 }, { "epoch": 1.78297926472459, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.794461250305176, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8740900754928589, "num_tokens": 532022461.0, "step": 14016 }, { "epoch": 1.7831064750031802, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.548372983932495, "learning_rate": 1e-05, "loss": 0.3856, "mean_token_accuracy": 0.8882372379302979, "num_tokens": 532059090.0, "step": 14017 }, { "epoch": 1.7832336852817707, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.816429853439331, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8619734644889832, "num_tokens": 532095926.0, "step": 14018 }, { "epoch": 1.7833608955603613, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5655674934387207, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8726646900177002, "num_tokens": 532131623.0, "step": 14019 }, { "epoch": 1.7834881058389518, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.810134172439575, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8551392555236816, "num_tokens": 532171594.0, "step": 14020 }, { "epoch": 1.7836153161175423, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.566155195236206, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8802732825279236, "num_tokens": 532209884.0, "step": 14021 }, { "epoch": 1.7837425263961328, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.820828437805176, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8624438643455505, "num_tokens": 532246182.0, "step": 14022 }, { "epoch": 1.7838697366747234, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5763485431671143, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8721556663513184, "num_tokens": 532281133.0, "step": 14023 }, { "epoch": 1.7839969469533137, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8065121173858643, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8677065372467041, "num_tokens": 532318593.0, "step": 14024 }, { "epoch": 1.7841241572319042, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.595763683319092, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8616646528244019, "num_tokens": 532353473.0, "step": 14025 }, { "epoch": 1.7842513675104947, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7835371494293213, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.8579153418540955, "num_tokens": 532395462.0, "step": 14026 }, { "epoch": 1.7843785777890853, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5802979469299316, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8772650957107544, "num_tokens": 532434677.0, "step": 14027 }, { "epoch": 1.7845057880676758, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7912721633911133, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8671674728393555, "num_tokens": 532470585.0, "step": 14028 }, { "epoch": 1.7846329983462663, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.598532199859619, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8636370301246643, "num_tokens": 532502657.0, "step": 14029 }, { "epoch": 1.7847602086248568, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7566380500793457, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8629759550094604, "num_tokens": 532548372.0, "step": 14030 }, { "epoch": 1.7848874189034474, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6053478717803955, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8621368408203125, "num_tokens": 532584707.0, "step": 14031 }, { "epoch": 1.785014629182038, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.731924295425415, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8706343770027161, "num_tokens": 532623179.0, "step": 14032 }, { "epoch": 1.7851418394606284, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5931382179260254, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8739979267120361, "num_tokens": 532663054.0, "step": 14033 }, { "epoch": 1.785269049739219, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.737577438354492, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8687683343887329, "num_tokens": 532696456.0, "step": 14034 }, { "epoch": 1.7853962600178095, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5798444747924805, "learning_rate": 1e-05, "loss": 0.3866, "mean_token_accuracy": 0.8843755722045898, "num_tokens": 532730726.0, "step": 14035 }, { "epoch": 1.7855234702964, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.736557960510254, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8688167333602905, "num_tokens": 532765561.0, "step": 14036 }, { "epoch": 1.7856506805749905, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5796597003936768, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8791056871414185, "num_tokens": 532803225.0, "step": 14037 }, { "epoch": 1.785777890853581, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.735644817352295, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8698734045028687, "num_tokens": 532843284.0, "step": 14038 }, { "epoch": 1.7859051011321716, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.598480224609375, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8675761222839355, "num_tokens": 532877637.0, "step": 14039 }, { "epoch": 1.7860323114107621, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.728602170944214, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.871801495552063, "num_tokens": 532914678.0, "step": 14040 }, { "epoch": 1.7861595216893527, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5815012454986572, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8706381320953369, "num_tokens": 532952309.0, "step": 14041 }, { "epoch": 1.786286731967943, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7301063537597656, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8773934841156006, "num_tokens": 532986440.0, "step": 14042 }, { "epoch": 1.7864139422465335, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6110174655914307, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8642109036445618, "num_tokens": 533020488.0, "step": 14043 }, { "epoch": 1.786541152525124, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.712249994277954, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8743859529495239, "num_tokens": 533052535.0, "step": 14044 }, { "epoch": 1.7866683628037145, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5826282501220703, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8786592483520508, "num_tokens": 533092924.0, "step": 14045 }, { "epoch": 1.786795573082305, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7208926677703857, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8680069446563721, "num_tokens": 533130497.0, "step": 14046 }, { "epoch": 1.7869227833608956, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5943312644958496, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8508298993110657, "num_tokens": 533170012.0, "step": 14047 }, { "epoch": 1.787049993639486, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.714007616043091, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8696423768997192, "num_tokens": 533207433.0, "step": 14048 }, { "epoch": 1.7871772039180764, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.571255922317505, "learning_rate": 1e-05, "loss": 0.3994, "mean_token_accuracy": 0.884957492351532, "num_tokens": 533248809.0, "step": 14049 }, { "epoch": 1.787304414196667, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.723966598510742, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8789722323417664, "num_tokens": 533283381.0, "step": 14050 }, { "epoch": 1.7874316244752575, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.565786838531494, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8811098337173462, "num_tokens": 533320951.0, "step": 14051 }, { "epoch": 1.787558834753848, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7246007919311523, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8733700513839722, "num_tokens": 533361080.0, "step": 14052 }, { "epoch": 1.7876860450324386, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.570574998855591, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8840999007225037, "num_tokens": 533396448.0, "step": 14053 }, { "epoch": 1.787813255311029, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7349750995635986, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8726863861083984, "num_tokens": 533434924.0, "step": 14054 }, { "epoch": 1.7879404655896196, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.569870948791504, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8735423684120178, "num_tokens": 533471248.0, "step": 14055 }, { "epoch": 1.7880676758682101, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7368433475494385, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.880282998085022, "num_tokens": 533512521.0, "step": 14056 }, { "epoch": 1.7881948861468007, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.562502861022949, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8786251544952393, "num_tokens": 533550385.0, "step": 14057 }, { "epoch": 1.7883220964253912, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7645604610443115, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8724353909492493, "num_tokens": 533580048.0, "step": 14058 }, { "epoch": 1.7884493067039817, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5702719688415527, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.882808268070221, "num_tokens": 533615115.0, "step": 14059 }, { "epoch": 1.7885765169825723, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7485387325286865, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8732314109802246, "num_tokens": 533653774.0, "step": 14060 }, { "epoch": 1.7887037272611628, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.555488109588623, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8832546472549438, "num_tokens": 533692628.0, "step": 14061 }, { "epoch": 1.7888309375397533, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7498373985290527, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8895300626754761, "num_tokens": 533734935.0, "step": 14062 }, { "epoch": 1.7889581478183438, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.555908203125, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8666919469833374, "num_tokens": 533772541.0, "step": 14063 }, { "epoch": 1.7890853580969344, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7563090324401855, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.876765251159668, "num_tokens": 533816036.0, "step": 14064 }, { "epoch": 1.789212568375525, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5727810859680176, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8718942403793335, "num_tokens": 533848174.0, "step": 14065 }, { "epoch": 1.7893397786541152, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.752180337905884, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8598195910453796, "num_tokens": 533890011.0, "step": 14066 }, { "epoch": 1.7894669889327057, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5789270401000977, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8621776103973389, "num_tokens": 533927650.0, "step": 14067 }, { "epoch": 1.7895941992112963, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.762078285217285, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8658195734024048, "num_tokens": 533959117.0, "step": 14068 }, { "epoch": 1.7897214094898868, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.616828441619873, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8606411218643188, "num_tokens": 533995396.0, "step": 14069 }, { "epoch": 1.7898486197684773, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002727508544921875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.729053497314453, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.882016658782959, "num_tokens": 534032498.0, "step": 14070 }, { "epoch": 1.7899758300470678, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.593553304672241, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.862804651260376, "num_tokens": 534071790.0, "step": 14071 }, { "epoch": 1.7901030403256584, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7079946994781494, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8638841509819031, "num_tokens": 534112778.0, "step": 14072 }, { "epoch": 1.7902302506042487, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5884056091308594, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8766441345214844, "num_tokens": 534150815.0, "step": 14073 }, { "epoch": 1.7903574608828392, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7191858291625977, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8682738542556763, "num_tokens": 534186148.0, "step": 14074 }, { "epoch": 1.7904846711614297, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5885870456695557, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8662723302841187, "num_tokens": 534219496.0, "step": 14075 }, { "epoch": 1.7906118814400203, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.704333543777466, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8778854608535767, "num_tokens": 534263093.0, "step": 14076 }, { "epoch": 1.7907390917186108, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.598593235015869, "learning_rate": 1e-05, "loss": 0.4025, "mean_token_accuracy": 0.8830414414405823, "num_tokens": 534298515.0, "step": 14077 }, { "epoch": 1.7908663019972013, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.702691078186035, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.875417947769165, "num_tokens": 534337210.0, "step": 14078 }, { "epoch": 1.7909935122757918, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6018612384796143, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8618043661117554, "num_tokens": 534373440.0, "step": 14079 }, { "epoch": 1.7911207225543824, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.691591501235962, "learning_rate": 1e-05, "loss": 0.3722, "mean_token_accuracy": 0.8948808908462524, "num_tokens": 534405560.0, "step": 14080 }, { "epoch": 1.791247932832973, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5894923210144043, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8743031024932861, "num_tokens": 534442824.0, "step": 14081 }, { "epoch": 1.7913751431115634, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.698075771331787, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8723717927932739, "num_tokens": 534482569.0, "step": 14082 }, { "epoch": 1.791502353390154, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5952768325805664, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8650836944580078, "num_tokens": 534518481.0, "step": 14083 }, { "epoch": 1.7916295636687445, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7314207553863525, "learning_rate": 1e-05, "loss": 0.5004, "mean_token_accuracy": 0.8507879972457886, "num_tokens": 534553273.0, "step": 14084 }, { "epoch": 1.791756773947335, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5743768215179443, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8662452697753906, "num_tokens": 534587322.0, "step": 14085 }, { "epoch": 1.7918839842259255, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7192366123199463, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.8819627165794373, "num_tokens": 534622999.0, "step": 14086 }, { "epoch": 1.792011194504516, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5802013874053955, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8633594512939453, "num_tokens": 534655953.0, "step": 14087 }, { "epoch": 1.7921384047831066, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.711146831512451, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.865395188331604, "num_tokens": 534699495.0, "step": 14088 }, { "epoch": 1.7922656150616971, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5716543197631836, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.879810094833374, "num_tokens": 534737650.0, "step": 14089 }, { "epoch": 1.7923928253402877, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7268834114074707, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.874599277973175, "num_tokens": 534771169.0, "step": 14090 }, { "epoch": 1.792520035618878, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5770294666290283, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.881508469581604, "num_tokens": 534802968.0, "step": 14091 }, { "epoch": 1.7926472458974685, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7196733951568604, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8751505017280579, "num_tokens": 534846975.0, "step": 14092 }, { "epoch": 1.792774456176059, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5766093730926514, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8750605583190918, "num_tokens": 534880969.0, "step": 14093 }, { "epoch": 1.7929016664546495, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7333486080169678, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8661067485809326, "num_tokens": 534918423.0, "step": 14094 }, { "epoch": 1.79302887673324, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.575286626815796, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8795748949050903, "num_tokens": 534952733.0, "step": 14095 }, { "epoch": 1.7931560870118306, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.748117208480835, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8782253265380859, "num_tokens": 534981868.0, "step": 14096 }, { "epoch": 1.793283297290421, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5599710941314697, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8693742752075195, "num_tokens": 535020940.0, "step": 14097 }, { "epoch": 1.7934105075690114, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.758867025375366, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8761318922042847, "num_tokens": 535053073.0, "step": 14098 }, { "epoch": 1.793537717847602, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.554290294647217, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.870561957359314, "num_tokens": 535097731.0, "step": 14099 }, { "epoch": 1.7936649281261925, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7642812728881836, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8646787405014038, "num_tokens": 535145571.0, "step": 14100 }, { "epoch": 1.793792138404783, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5570874214172363, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8685277104377747, "num_tokens": 535184226.0, "step": 14101 }, { "epoch": 1.7939193486833735, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.795792579650879, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8527594208717346, "num_tokens": 535221990.0, "step": 14102 }, { "epoch": 1.794046558961964, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.556251049041748, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8640843629837036, "num_tokens": 535261588.0, "step": 14103 }, { "epoch": 1.7941737692405546, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8045730590820312, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8533825874328613, "num_tokens": 535299023.0, "step": 14104 }, { "epoch": 1.7943009795191451, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5698130130767822, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.866288423538208, "num_tokens": 535332174.0, "step": 14105 }, { "epoch": 1.7944281897977357, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.799586057662964, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8760705590248108, "num_tokens": 535367242.0, "step": 14106 }, { "epoch": 1.7945554000763262, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5715157985687256, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8731298446655273, "num_tokens": 535401962.0, "step": 14107 }, { "epoch": 1.7946826103549167, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7949745655059814, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8621575832366943, "num_tokens": 535441481.0, "step": 14108 }, { "epoch": 1.7948098206335072, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5774056911468506, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8775326013565063, "num_tokens": 535473619.0, "step": 14109 }, { "epoch": 1.7949370309120978, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.780836343765259, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8839677572250366, "num_tokens": 535515340.0, "step": 14110 }, { "epoch": 1.7950642411906883, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5869104862213135, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8559239506721497, "num_tokens": 535555090.0, "step": 14111 }, { "epoch": 1.7951914514692788, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7762229442596436, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.87083899974823, "num_tokens": 535595484.0, "step": 14112 }, { "epoch": 1.7953186617478694, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5801947116851807, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.869493842124939, "num_tokens": 535634170.0, "step": 14113 }, { "epoch": 1.7954458720264599, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.76765513420105, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8749514222145081, "num_tokens": 535671771.0, "step": 14114 }, { "epoch": 1.7955730823050502, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.575057029724121, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8794726133346558, "num_tokens": 535710280.0, "step": 14115 }, { "epoch": 1.7957002925836407, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.775348663330078, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8677219152450562, "num_tokens": 535749509.0, "step": 14116 }, { "epoch": 1.7958275028622313, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5598487854003906, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8786197900772095, "num_tokens": 535794184.0, "step": 14117 }, { "epoch": 1.7959547131408218, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8006207942962646, "learning_rate": 1e-05, "loss": 0.4952, "mean_token_accuracy": 0.8526390194892883, "num_tokens": 535828818.0, "step": 14118 }, { "epoch": 1.7960819234194123, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.565192461013794, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8646738529205322, "num_tokens": 535869461.0, "step": 14119 }, { "epoch": 1.7962091336980028, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7944419384002686, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8726032972335815, "num_tokens": 535907287.0, "step": 14120 }, { "epoch": 1.7963363439765934, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5690722465515137, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8794251680374146, "num_tokens": 535949063.0, "step": 14121 }, { "epoch": 1.7964635542551837, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7923896312713623, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8733687400817871, "num_tokens": 535987202.0, "step": 14122 }, { "epoch": 1.7965907645337742, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5820038318634033, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.881234884262085, "num_tokens": 536021790.0, "step": 14123 }, { "epoch": 1.7967179748123647, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7908167839050293, "learning_rate": 1e-05, "loss": 0.5418, "mean_token_accuracy": 0.8438788652420044, "num_tokens": 536069932.0, "step": 14124 }, { "epoch": 1.7968451850909553, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5702602863311768, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8796207308769226, "num_tokens": 536110428.0, "step": 14125 }, { "epoch": 1.7969723953695458, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7883880138397217, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8684534430503845, "num_tokens": 536147711.0, "step": 14126 }, { "epoch": 1.7970996056481363, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.580089569091797, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8777564764022827, "num_tokens": 536184422.0, "step": 14127 }, { "epoch": 1.7972268159267268, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7667722702026367, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8838119506835938, "num_tokens": 536224802.0, "step": 14128 }, { "epoch": 1.7973540262053174, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5760498046875, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8676517605781555, "num_tokens": 536267596.0, "step": 14129 }, { "epoch": 1.797481236483908, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.802379608154297, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8754143714904785, "num_tokens": 536306130.0, "step": 14130 }, { "epoch": 1.7976084467624984, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.575073003768921, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8700562119483948, "num_tokens": 536345451.0, "step": 14131 }, { "epoch": 1.797735657041089, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.795846939086914, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8654440641403198, "num_tokens": 536377953.0, "step": 14132 }, { "epoch": 1.7978628673196795, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.576742172241211, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8810877799987793, "num_tokens": 536410879.0, "step": 14133 }, { "epoch": 1.79799007759827, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.787414073944092, "learning_rate": 1e-05, "loss": 0.485, "mean_token_accuracy": 0.8582455515861511, "num_tokens": 536448351.0, "step": 14134 }, { "epoch": 1.7981172878768605, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5712697505950928, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.86687171459198, "num_tokens": 536488968.0, "step": 14135 }, { "epoch": 1.798244498155451, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.771315097808838, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8807367086410522, "num_tokens": 536529189.0, "step": 14136 }, { "epoch": 1.7983717084340416, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.570310592651367, "learning_rate": 1e-05, "loss": 0.3918, "mean_token_accuracy": 0.8871655464172363, "num_tokens": 536567778.0, "step": 14137 }, { "epoch": 1.7984989187126321, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7740373611450195, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.870811402797699, "num_tokens": 536609193.0, "step": 14138 }, { "epoch": 1.7986261289912227, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5843358039855957, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8550189733505249, "num_tokens": 536646122.0, "step": 14139 }, { "epoch": 1.798753339269813, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7737770080566406, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8693574666976929, "num_tokens": 536682146.0, "step": 14140 }, { "epoch": 1.7988805495484035, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5934953689575195, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8622934818267822, "num_tokens": 536715449.0, "step": 14141 }, { "epoch": 1.799007759826994, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.757408618927002, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8671875596046448, "num_tokens": 536752430.0, "step": 14142 }, { "epoch": 1.7991349701055845, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.599799394607544, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.858203649520874, "num_tokens": 536785577.0, "step": 14143 }, { "epoch": 1.799262180384175, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.735491991043091, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8825088739395142, "num_tokens": 536824074.0, "step": 14144 }, { "epoch": 1.7993893906627656, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5778260231018066, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8872992396354675, "num_tokens": 536865819.0, "step": 14145 }, { "epoch": 1.799516600941356, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.731494188308716, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8725132942199707, "num_tokens": 536905538.0, "step": 14146 }, { "epoch": 1.7996438112199464, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5888772010803223, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8799180388450623, "num_tokens": 536944895.0, "step": 14147 }, { "epoch": 1.799771021498537, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7322943210601807, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.868450403213501, "num_tokens": 536981094.0, "step": 14148 }, { "epoch": 1.7998982317771275, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5901503562927246, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8588656783103943, "num_tokens": 537020242.0, "step": 14149 }, { "epoch": 1.800025442055718, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7134687900543213, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8607494831085205, "num_tokens": 537062475.0, "step": 14150 }, { "epoch": 1.8001526523343085, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.591322183609009, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8658860921859741, "num_tokens": 537100969.0, "step": 14151 }, { "epoch": 1.800279862612899, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.714195966720581, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8700858354568481, "num_tokens": 537141828.0, "step": 14152 }, { "epoch": 1.8004070728914896, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5827107429504395, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8776621222496033, "num_tokens": 537181324.0, "step": 14153 }, { "epoch": 1.8005342831700801, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.713728904724121, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.876808762550354, "num_tokens": 537224593.0, "step": 14154 }, { "epoch": 1.8006614934486707, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5890510082244873, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.866547167301178, "num_tokens": 537262835.0, "step": 14155 }, { "epoch": 1.8007887037272612, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.712912082672119, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8685176372528076, "num_tokens": 537299722.0, "step": 14156 }, { "epoch": 1.8009159140058517, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5745766162872314, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8748427629470825, "num_tokens": 537345445.0, "step": 14157 }, { "epoch": 1.8010431242844422, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7228899002075195, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8732032179832458, "num_tokens": 537382327.0, "step": 14158 }, { "epoch": 1.8011703345630328, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5841197967529297, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8713781833648682, "num_tokens": 537419013.0, "step": 14159 }, { "epoch": 1.8012975448416233, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.716668128967285, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8702132701873779, "num_tokens": 537459841.0, "step": 14160 }, { "epoch": 1.8014247551202138, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5762922763824463, "learning_rate": 1e-05, "loss": 0.3964, "mean_token_accuracy": 0.8830995559692383, "num_tokens": 537494124.0, "step": 14161 }, { "epoch": 1.8015519653988044, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7175769805908203, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8772262930870056, "num_tokens": 537532798.0, "step": 14162 }, { "epoch": 1.8016791756773949, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.578028917312622, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8707882165908813, "num_tokens": 537568548.0, "step": 14163 }, { "epoch": 1.8018063859559852, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.717336893081665, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.8859578371047974, "num_tokens": 537604413.0, "step": 14164 }, { "epoch": 1.8019335962345757, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5717477798461914, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8726977705955505, "num_tokens": 537640588.0, "step": 14165 }, { "epoch": 1.8020608065131662, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.722330331802368, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8628524541854858, "num_tokens": 537682664.0, "step": 14166 }, { "epoch": 1.8021880167917568, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.591005563735962, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8550719618797302, "num_tokens": 537716020.0, "step": 14167 }, { "epoch": 1.8023152270703473, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7048165798187256, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.876784086227417, "num_tokens": 537756895.0, "step": 14168 }, { "epoch": 1.8024424373489378, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.590620756149292, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8668206930160522, "num_tokens": 537794130.0, "step": 14169 }, { "epoch": 1.8025696476275284, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.728752851486206, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8841580748558044, "num_tokens": 537822865.0, "step": 14170 }, { "epoch": 1.8026968579061187, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.583585500717163, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.877413809299469, "num_tokens": 537854594.0, "step": 14171 }, { "epoch": 1.8028240681847092, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.724555492401123, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8713948726654053, "num_tokens": 537889667.0, "step": 14172 }, { "epoch": 1.8029512784632997, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.5820260047912598, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8599058389663696, "num_tokens": 537924322.0, "step": 14173 }, { "epoch": 1.8030784887418903, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.722632884979248, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8816725611686707, "num_tokens": 537962196.0, "step": 14174 }, { "epoch": 1.8032056990204808, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.57814884185791, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8792357444763184, "num_tokens": 537999916.0, "step": 14175 }, { "epoch": 1.8033329092990713, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.724343776702881, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8824383020401001, "num_tokens": 538035823.0, "step": 14176 }, { "epoch": 1.8034601195776618, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5742948055267334, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8614935874938965, "num_tokens": 538080664.0, "step": 14177 }, { "epoch": 1.8035873298562524, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7245585918426514, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8649550080299377, "num_tokens": 538123178.0, "step": 14178 }, { "epoch": 1.803714540134843, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.580214262008667, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8759092092514038, "num_tokens": 538158757.0, "step": 14179 }, { "epoch": 1.8038417504134334, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.727288246154785, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8668718338012695, "num_tokens": 538199019.0, "step": 14180 }, { "epoch": 1.803968960692024, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5617740154266357, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8712700009346008, "num_tokens": 538239839.0, "step": 14181 }, { "epoch": 1.8040961709706145, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7355592250823975, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8737941384315491, "num_tokens": 538287677.0, "step": 14182 }, { "epoch": 1.804223381249205, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.571356773376465, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.873200535774231, "num_tokens": 538322304.0, "step": 14183 }, { "epoch": 1.8043505915277955, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.726749897003174, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8765410780906677, "num_tokens": 538364087.0, "step": 14184 }, { "epoch": 1.804477801806386, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.571737051010132, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8699862957000732, "num_tokens": 538401507.0, "step": 14185 }, { "epoch": 1.8046050120849766, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.725466251373291, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8759044408798218, "num_tokens": 538440666.0, "step": 14186 }, { "epoch": 1.8047322223635671, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5812714099884033, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8636553287506104, "num_tokens": 538478915.0, "step": 14187 }, { "epoch": 1.8048594326421576, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7302441596984863, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8742722272872925, "num_tokens": 538517770.0, "step": 14188 }, { "epoch": 1.804986642920748, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.582291603088379, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8596466779708862, "num_tokens": 538555812.0, "step": 14189 }, { "epoch": 1.8051138531993385, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7187023162841797, "learning_rate": 1e-05, "loss": 0.3904, "mean_token_accuracy": 0.8863475322723389, "num_tokens": 538596023.0, "step": 14190 }, { "epoch": 1.805241063477929, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.577803373336792, "learning_rate": 1e-05, "loss": 0.3909, "mean_token_accuracy": 0.886143684387207, "num_tokens": 538630840.0, "step": 14191 }, { "epoch": 1.8053682737565195, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.737581968307495, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.863304615020752, "num_tokens": 538668515.0, "step": 14192 }, { "epoch": 1.80549548403511, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5818252563476562, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8628334403038025, "num_tokens": 538703996.0, "step": 14193 }, { "epoch": 1.8056226943137006, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.728057384490967, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8721569776535034, "num_tokens": 538742819.0, "step": 14194 }, { "epoch": 1.805749904592291, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5769498348236084, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.8826266527175903, "num_tokens": 538776218.0, "step": 14195 }, { "epoch": 1.8058771148708814, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7324135303497314, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8729700446128845, "num_tokens": 538813377.0, "step": 14196 }, { "epoch": 1.806004325149472, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.564145088195801, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8687149882316589, "num_tokens": 538856755.0, "step": 14197 }, { "epoch": 1.8061315354280625, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.750265598297119, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8709562420845032, "num_tokens": 538898175.0, "step": 14198 }, { "epoch": 1.806258745706653, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5694544315338135, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8650897741317749, "num_tokens": 538935539.0, "step": 14199 }, { "epoch": 1.8063859559852435, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.748723268508911, "learning_rate": 1e-05, "loss": 0.4968, "mean_token_accuracy": 0.8556064963340759, "num_tokens": 538978999.0, "step": 14200 }, { "epoch": 1.806513166263834, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5702757835388184, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8802158832550049, "num_tokens": 539013480.0, "step": 14201 }, { "epoch": 1.8066403765424246, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7637832164764404, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8630210161209106, "num_tokens": 539047120.0, "step": 14202 }, { "epoch": 1.8067675868210151, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5613064765930176, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8817997574806213, "num_tokens": 539088243.0, "step": 14203 }, { "epoch": 1.8068947970996057, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.776564836502075, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8622250556945801, "num_tokens": 539128240.0, "step": 14204 }, { "epoch": 1.8070220073781962, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5558979511260986, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.870464026927948, "num_tokens": 539168792.0, "step": 14205 }, { "epoch": 1.8071492176567867, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.778014898300171, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8765667676925659, "num_tokens": 539210510.0, "step": 14206 }, { "epoch": 1.8072764279353772, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5637223720550537, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8743792772293091, "num_tokens": 539248926.0, "step": 14207 }, { "epoch": 1.8074036382139678, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.788233518600464, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8673456311225891, "num_tokens": 539285795.0, "step": 14208 }, { "epoch": 1.8075308484925583, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5526857376098633, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8785268068313599, "num_tokens": 539324891.0, "step": 14209 }, { "epoch": 1.8076580587711488, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.801374673843384, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8626468777656555, "num_tokens": 539364684.0, "step": 14210 }, { "epoch": 1.8077852690497394, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5481045246124268, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8789162039756775, "num_tokens": 539403312.0, "step": 14211 }, { "epoch": 1.8079124793283299, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8232853412628174, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.884100079536438, "num_tokens": 539443567.0, "step": 14212 }, { "epoch": 1.8080396896069202, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5629916191101074, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8651484847068787, "num_tokens": 539480969.0, "step": 14213 }, { "epoch": 1.8081668998855107, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8230791091918945, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8708792924880981, "num_tokens": 539518105.0, "step": 14214 }, { "epoch": 1.8082941101641012, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5582494735717773, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8577568531036377, "num_tokens": 539561364.0, "step": 14215 }, { "epoch": 1.8084213204426918, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8368897438049316, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8654335141181946, "num_tokens": 539594602.0, "step": 14216 }, { "epoch": 1.8085485307212823, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5718460083007812, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8576943874359131, "num_tokens": 539635985.0, "step": 14217 }, { "epoch": 1.8086757409998728, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8276562690734863, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8739303350448608, "num_tokens": 539675201.0, "step": 14218 }, { "epoch": 1.8088029512784631, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.577071189880371, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8802295923233032, "num_tokens": 539712661.0, "step": 14219 }, { "epoch": 1.8089301615570537, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8294193744659424, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8589165806770325, "num_tokens": 539749941.0, "step": 14220 }, { "epoch": 1.8090573718356442, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5935163497924805, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8669461607933044, "num_tokens": 539786906.0, "step": 14221 }, { "epoch": 1.8091845821142347, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8168740272521973, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8761287927627563, "num_tokens": 539824235.0, "step": 14222 }, { "epoch": 1.8093117923928252, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5892930030822754, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8609616756439209, "num_tokens": 539864326.0, "step": 14223 }, { "epoch": 1.8094390026714158, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7920281887054443, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8577960133552551, "num_tokens": 539909105.0, "step": 14224 }, { "epoch": 1.8095662129500063, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6080713272094727, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8712589740753174, "num_tokens": 539944846.0, "step": 14225 }, { "epoch": 1.8096934232285968, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.770082473754883, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8739746809005737, "num_tokens": 539984720.0, "step": 14226 }, { "epoch": 1.8098206335071874, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6001551151275635, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8629035949707031, "num_tokens": 540025002.0, "step": 14227 }, { "epoch": 1.8099478437857779, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.771902561187744, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8622164726257324, "num_tokens": 540060256.0, "step": 14228 }, { "epoch": 1.8100750540643684, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5981411933898926, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8700004816055298, "num_tokens": 540098073.0, "step": 14229 }, { "epoch": 1.810202264342959, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7710626125335693, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8728171586990356, "num_tokens": 540128704.0, "step": 14230 }, { "epoch": 1.8103294746215495, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.592473268508911, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.8822671175003052, "num_tokens": 540161293.0, "step": 14231 }, { "epoch": 1.81045668490014, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.763950824737549, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8833847045898438, "num_tokens": 540193685.0, "step": 14232 }, { "epoch": 1.8105838951787305, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.565837860107422, "learning_rate": 1e-05, "loss": 0.3954, "mean_token_accuracy": 0.8831763863563538, "num_tokens": 540232217.0, "step": 14233 }, { "epoch": 1.810711105457321, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7792389392852783, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8679442405700684, "num_tokens": 540267561.0, "step": 14234 }, { "epoch": 1.8108383157359116, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5656778812408447, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8788143396377563, "num_tokens": 540305788.0, "step": 14235 }, { "epoch": 1.8109655260145021, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.794875383377075, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8662821650505066, "num_tokens": 540341105.0, "step": 14236 }, { "epoch": 1.8110927362930926, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.581524610519409, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8675141334533691, "num_tokens": 540380502.0, "step": 14237 }, { "epoch": 1.811219946571683, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.783461093902588, "learning_rate": 1e-05, "loss": 0.5038, "mean_token_accuracy": 0.8512859344482422, "num_tokens": 540420561.0, "step": 14238 }, { "epoch": 1.8113471568502735, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5724990367889404, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8665449619293213, "num_tokens": 540460561.0, "step": 14239 }, { "epoch": 1.811474367128864, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7856805324554443, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8687841296195984, "num_tokens": 540498939.0, "step": 14240 }, { "epoch": 1.8116015774074545, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.580045461654663, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8616548180580139, "num_tokens": 540536127.0, "step": 14241 }, { "epoch": 1.811728787686045, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.790963649749756, "learning_rate": 1e-05, "loss": 0.4025, "mean_token_accuracy": 0.8831899762153625, "num_tokens": 540567812.0, "step": 14242 }, { "epoch": 1.8118559979646356, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5706429481506348, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8827126026153564, "num_tokens": 540604176.0, "step": 14243 }, { "epoch": 1.811983208243226, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7918286323547363, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8556019067764282, "num_tokens": 540640989.0, "step": 14244 }, { "epoch": 1.8121104185218164, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.57643461227417, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8763853907585144, "num_tokens": 540680590.0, "step": 14245 }, { "epoch": 1.812237628800407, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7741191387176514, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8740760684013367, "num_tokens": 540723739.0, "step": 14246 }, { "epoch": 1.8123648390789975, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.585134267807007, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.872222900390625, "num_tokens": 540759686.0, "step": 14247 }, { "epoch": 1.812492049357588, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7657577991485596, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8862305879592896, "num_tokens": 540795710.0, "step": 14248 }, { "epoch": 1.8126192596361785, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.605478048324585, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8708422780036926, "num_tokens": 540838558.0, "step": 14249 }, { "epoch": 1.812746469914769, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.742931842803955, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8843183517456055, "num_tokens": 540874543.0, "step": 14250 }, { "epoch": 1.8128736801933596, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.596581220626831, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8730411529541016, "num_tokens": 540909903.0, "step": 14251 }, { "epoch": 1.8130008904719501, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7330141067504883, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8713645935058594, "num_tokens": 540947977.0, "step": 14252 }, { "epoch": 1.8131281007505406, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5886826515197754, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8614182472229004, "num_tokens": 540986288.0, "step": 14253 }, { "epoch": 1.8132553110291312, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7287490367889404, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.872782826423645, "num_tokens": 541027607.0, "step": 14254 }, { "epoch": 1.8133825213077217, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5834403038024902, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8759630918502808, "num_tokens": 541064424.0, "step": 14255 }, { "epoch": 1.8135097315863122, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7400872707366943, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8661721348762512, "num_tokens": 541097815.0, "step": 14256 }, { "epoch": 1.8136369418649028, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5883941650390625, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8643356561660767, "num_tokens": 541136567.0, "step": 14257 }, { "epoch": 1.8137641521434933, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7319118976593018, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8753706216812134, "num_tokens": 541174305.0, "step": 14258 }, { "epoch": 1.8138913624220838, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5916452407836914, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8713240623474121, "num_tokens": 541207118.0, "step": 14259 }, { "epoch": 1.8140185727006743, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7274205684661865, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8825697302818298, "num_tokens": 541242208.0, "step": 14260 }, { "epoch": 1.8141457829792649, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5810065269470215, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8744149208068848, "num_tokens": 541278553.0, "step": 14261 }, { "epoch": 1.8142729932578552, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7315611839294434, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8666471242904663, "num_tokens": 541318329.0, "step": 14262 }, { "epoch": 1.8144002035364457, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5752851963043213, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8791071772575378, "num_tokens": 541353406.0, "step": 14263 }, { "epoch": 1.8145274138150362, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.750577449798584, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8681999444961548, "num_tokens": 541385504.0, "step": 14264 }, { "epoch": 1.8146546240936268, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.56475830078125, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8653271198272705, "num_tokens": 541426172.0, "step": 14265 }, { "epoch": 1.8147818343722173, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7554852962493896, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8756124973297119, "num_tokens": 541464785.0, "step": 14266 }, { "epoch": 1.8149090446508078, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5716612339019775, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8646426200866699, "num_tokens": 541500271.0, "step": 14267 }, { "epoch": 1.8150362549293981, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7499380111694336, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8677347302436829, "num_tokens": 541537482.0, "step": 14268 }, { "epoch": 1.8151634652079887, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5634140968322754, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8744336366653442, "num_tokens": 541576669.0, "step": 14269 }, { "epoch": 1.8152906754865792, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7649381160736084, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8717971444129944, "num_tokens": 541616040.0, "step": 14270 }, { "epoch": 1.8154178857651697, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5749433040618896, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8560632467269897, "num_tokens": 541649994.0, "step": 14271 }, { "epoch": 1.8155450960437602, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.771909475326538, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8581496477127075, "num_tokens": 541688468.0, "step": 14272 }, { "epoch": 1.8156723063223508, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.592906951904297, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8691166639328003, "num_tokens": 541725173.0, "step": 14273 }, { "epoch": 1.8157995166009413, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7546591758728027, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8685082793235779, "num_tokens": 541762523.0, "step": 14274 }, { "epoch": 1.8159267268795318, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5766332149505615, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8695505261421204, "num_tokens": 541799731.0, "step": 14275 }, { "epoch": 1.8160539371581224, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7366256713867188, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8810184597969055, "num_tokens": 541844172.0, "step": 14276 }, { "epoch": 1.8161811474367129, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5602121353149414, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.872983992099762, "num_tokens": 541887526.0, "step": 14277 }, { "epoch": 1.8163083577153034, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7388455867767334, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8725367784500122, "num_tokens": 541932712.0, "step": 14278 }, { "epoch": 1.816435567993894, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.577334403991699, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8717274069786072, "num_tokens": 541968915.0, "step": 14279 }, { "epoch": 1.8165627782724845, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7456274032592773, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8658721446990967, "num_tokens": 542010296.0, "step": 14280 }, { "epoch": 1.816689988551075, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5563912391662598, "learning_rate": 1e-05, "loss": 0.3872, "mean_token_accuracy": 0.8875904083251953, "num_tokens": 542050096.0, "step": 14281 }, { "epoch": 1.8168171988296655, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7712104320526123, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8670003414154053, "num_tokens": 542086746.0, "step": 14282 }, { "epoch": 1.816944409108256, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.562251567840576, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8762419819831848, "num_tokens": 542123095.0, "step": 14283 }, { "epoch": 1.8170716193868466, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.772395133972168, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8762356638908386, "num_tokens": 542157797.0, "step": 14284 }, { "epoch": 1.817198829665437, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.5587971210479736, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8642565011978149, "num_tokens": 542196091.0, "step": 14285 }, { "epoch": 1.8173260399440276, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.784003257751465, "learning_rate": 1e-05, "loss": 0.4846, "mean_token_accuracy": 0.8574692606925964, "num_tokens": 542234242.0, "step": 14286 }, { "epoch": 1.817453250222618, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5502469539642334, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8768839836120605, "num_tokens": 542273187.0, "step": 14287 }, { "epoch": 1.8175804605012085, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7960712909698486, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8786274194717407, "num_tokens": 542308116.0, "step": 14288 }, { "epoch": 1.817707670779799, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.572201728820801, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8777801990509033, "num_tokens": 542339964.0, "step": 14289 }, { "epoch": 1.8178348810583895, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7826569080352783, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8661078810691833, "num_tokens": 542381923.0, "step": 14290 }, { "epoch": 1.81796209133698, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5656862258911133, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.880661129951477, "num_tokens": 542418468.0, "step": 14291 }, { "epoch": 1.8180893016155706, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.817209243774414, "learning_rate": 1e-05, "loss": 0.5273, "mean_token_accuracy": 0.8515159487724304, "num_tokens": 542448806.0, "step": 14292 }, { "epoch": 1.818216511894161, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5724828243255615, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.872988224029541, "num_tokens": 542483991.0, "step": 14293 }, { "epoch": 1.8183437221727514, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7927138805389404, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8747154474258423, "num_tokens": 542528334.0, "step": 14294 }, { "epoch": 1.818470932451342, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.570432424545288, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8842709064483643, "num_tokens": 542562903.0, "step": 14295 }, { "epoch": 1.8185981427299325, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7907400131225586, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8754147291183472, "num_tokens": 542600150.0, "step": 14296 }, { "epoch": 1.818725353008523, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.579627513885498, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8794493675231934, "num_tokens": 542637894.0, "step": 14297 }, { "epoch": 1.8188525632871135, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8134677410125732, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8569628000259399, "num_tokens": 542667729.0, "step": 14298 }, { "epoch": 1.818979773565704, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5894412994384766, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.874687671661377, "num_tokens": 542699994.0, "step": 14299 }, { "epoch": 1.8191069838442946, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7689435482025146, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8782070875167847, "num_tokens": 542743849.0, "step": 14300 }, { "epoch": 1.8192341941228851, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5822441577911377, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.860858678817749, "num_tokens": 542784788.0, "step": 14301 }, { "epoch": 1.8193614044014756, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7795934677124023, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8690237402915955, "num_tokens": 542820942.0, "step": 14302 }, { "epoch": 1.8194886146800662, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5888283252716064, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8819989562034607, "num_tokens": 542857345.0, "step": 14303 }, { "epoch": 1.8196158249586567, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7684524059295654, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8806379437446594, "num_tokens": 542890952.0, "step": 14304 }, { "epoch": 1.8197430352372472, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.604599952697754, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8580069541931152, "num_tokens": 542927487.0, "step": 14305 }, { "epoch": 1.8198702455158378, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7454848289489746, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8700910806655884, "num_tokens": 542967554.0, "step": 14306 }, { "epoch": 1.8199974557944283, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5978946685791016, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8599101305007935, "num_tokens": 543003768.0, "step": 14307 }, { "epoch": 1.8201246660730188, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7280044555664062, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8757302165031433, "num_tokens": 543041937.0, "step": 14308 }, { "epoch": 1.8202518763516093, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5752627849578857, "learning_rate": 1e-05, "loss": 0.3923, "mean_token_accuracy": 0.8882502317428589, "num_tokens": 543084358.0, "step": 14309 }, { "epoch": 1.8203790866301999, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7374348640441895, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8756523728370667, "num_tokens": 543121749.0, "step": 14310 }, { "epoch": 1.8205062969087902, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5899717807769775, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8601006269454956, "num_tokens": 543161160.0, "step": 14311 }, { "epoch": 1.8206335071873807, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.734281301498413, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8697781562805176, "num_tokens": 543197364.0, "step": 14312 }, { "epoch": 1.8207607174659712, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5870139598846436, "learning_rate": 1e-05, "loss": 0.3994, "mean_token_accuracy": 0.8861029148101807, "num_tokens": 543233381.0, "step": 14313 }, { "epoch": 1.8208879277445618, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7512199878692627, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8632603883743286, "num_tokens": 543266139.0, "step": 14314 }, { "epoch": 1.8210151380231523, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5866048336029053, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8572304844856262, "num_tokens": 543307681.0, "step": 14315 }, { "epoch": 1.8211423483017428, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7442245483398438, "learning_rate": 1e-05, "loss": 0.4932, "mean_token_accuracy": 0.860873818397522, "num_tokens": 543347916.0, "step": 14316 }, { "epoch": 1.8212695585803331, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5673906803131104, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8810888528823853, "num_tokens": 543386326.0, "step": 14317 }, { "epoch": 1.8213967688589237, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7537841796875, "learning_rate": 1e-05, "loss": 0.4612, "mean_token_accuracy": 0.8666918277740479, "num_tokens": 543423220.0, "step": 14318 }, { "epoch": 1.8215239791375142, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.568603754043579, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8780542612075806, "num_tokens": 543461672.0, "step": 14319 }, { "epoch": 1.8216511894161047, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7601888179779053, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8740313053131104, "num_tokens": 543498775.0, "step": 14320 }, { "epoch": 1.8217783996946952, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5543577671051025, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8821086287498474, "num_tokens": 543539966.0, "step": 14321 }, { "epoch": 1.8219056099732858, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.768327236175537, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8729780912399292, "num_tokens": 543577176.0, "step": 14322 }, { "epoch": 1.8220328202518763, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.555891513824463, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8610041737556458, "num_tokens": 543621328.0, "step": 14323 }, { "epoch": 1.8221600305304668, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.784209966659546, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8597358465194702, "num_tokens": 543662446.0, "step": 14324 }, { "epoch": 1.8222872408090574, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.570296049118042, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8808847069740295, "num_tokens": 543695739.0, "step": 14325 }, { "epoch": 1.8224144510876479, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.76979660987854, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8771378993988037, "num_tokens": 543735823.0, "step": 14326 }, { "epoch": 1.8225416613662384, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.575726270675659, "learning_rate": 1e-05, "loss": 0.4903, "mean_token_accuracy": 0.8541516661643982, "num_tokens": 543771800.0, "step": 14327 }, { "epoch": 1.822668871644829, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.767549514770508, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8633238673210144, "num_tokens": 543812425.0, "step": 14328 }, { "epoch": 1.8227960819234195, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5673608779907227, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8751763105392456, "num_tokens": 543849867.0, "step": 14329 }, { "epoch": 1.82292329220201, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7605679035186768, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.88258957862854, "num_tokens": 543887289.0, "step": 14330 }, { "epoch": 1.8230505024806005, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5763890743255615, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8769088983535767, "num_tokens": 543923432.0, "step": 14331 }, { "epoch": 1.823177712759191, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.749495267868042, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8567495346069336, "num_tokens": 543966655.0, "step": 14332 }, { "epoch": 1.8233049230377816, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5705649852752686, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8778135776519775, "num_tokens": 544006920.0, "step": 14333 }, { "epoch": 1.823432133316372, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.757986545562744, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8720970153808594, "num_tokens": 544044066.0, "step": 14334 }, { "epoch": 1.8235593435949626, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5964560508728027, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8658536672592163, "num_tokens": 544073795.0, "step": 14335 }, { "epoch": 1.823686553873553, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.741945505142212, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8768242597579956, "num_tokens": 544107378.0, "step": 14336 }, { "epoch": 1.8238137641521435, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5829808712005615, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8738837838172913, "num_tokens": 544144377.0, "step": 14337 }, { "epoch": 1.823940974430734, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7474730014801025, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8675657510757446, "num_tokens": 544176484.0, "step": 14338 }, { "epoch": 1.8240681847093245, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5871098041534424, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8779458403587341, "num_tokens": 544209152.0, "step": 14339 }, { "epoch": 1.824195394987915, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.737164258956909, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8669205904006958, "num_tokens": 544249197.0, "step": 14340 }, { "epoch": 1.8243226052665056, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5702052116394043, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.8861943483352661, "num_tokens": 544290225.0, "step": 14341 }, { "epoch": 1.8244498155450959, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7292590141296387, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8685941100120544, "num_tokens": 544333015.0, "step": 14342 }, { "epoch": 1.8245770258236864, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5819082260131836, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.866047739982605, "num_tokens": 544370165.0, "step": 14343 }, { "epoch": 1.824704236102277, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.730072259902954, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8733596205711365, "num_tokens": 544407284.0, "step": 14344 }, { "epoch": 1.8248314463808675, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.597540855407715, "learning_rate": 1e-05, "loss": 0.4874, "mean_token_accuracy": 0.8603063821792603, "num_tokens": 544441558.0, "step": 14345 }, { "epoch": 1.824958656659458, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7337117195129395, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8463538885116577, "num_tokens": 544478468.0, "step": 14346 }, { "epoch": 1.8250858669380485, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.574190139770508, "learning_rate": 1e-05, "loss": 0.3926, "mean_token_accuracy": 0.8880417346954346, "num_tokens": 544516271.0, "step": 14347 }, { "epoch": 1.825213077216639, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.730839490890503, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8815906643867493, "num_tokens": 544553198.0, "step": 14348 }, { "epoch": 1.8253402874952296, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6039860248565674, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8780241012573242, "num_tokens": 544586381.0, "step": 14349 }, { "epoch": 1.8254674977738201, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7189714908599854, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8689069747924805, "num_tokens": 544628342.0, "step": 14350 }, { "epoch": 1.8255947080524106, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.584993839263916, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8667579889297485, "num_tokens": 544667096.0, "step": 14351 }, { "epoch": 1.8257219183310012, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7110424041748047, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8772575259208679, "num_tokens": 544711467.0, "step": 14352 }, { "epoch": 1.8258491286095917, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.6042325496673584, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.873859167098999, "num_tokens": 544751672.0, "step": 14353 }, { "epoch": 1.8259763388881822, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6934664249420166, "learning_rate": 1e-05, "loss": 0.3732, "mean_token_accuracy": 0.894701361656189, "num_tokens": 544785479.0, "step": 14354 }, { "epoch": 1.8261035491667728, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.593256711959839, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8754082918167114, "num_tokens": 544823122.0, "step": 14355 }, { "epoch": 1.8262307594453633, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.695282459259033, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8812592625617981, "num_tokens": 544862544.0, "step": 14356 }, { "epoch": 1.8263579697239538, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.595757484436035, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8657167553901672, "num_tokens": 544899538.0, "step": 14357 }, { "epoch": 1.8264851800025443, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7002906799316406, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8792337775230408, "num_tokens": 544934544.0, "step": 14358 }, { "epoch": 1.8266123902811349, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5971741676330566, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8639000058174133, "num_tokens": 544973304.0, "step": 14359 }, { "epoch": 1.8267396005597252, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6843109130859375, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.8797167539596558, "num_tokens": 545010748.0, "step": 14360 }, { "epoch": 1.8268668108383157, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5852503776550293, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.866920530796051, "num_tokens": 545050742.0, "step": 14361 }, { "epoch": 1.8269940211169062, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.70955753326416, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8685246706008911, "num_tokens": 545083904.0, "step": 14362 }, { "epoch": 1.8271212313954968, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5697035789489746, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8812959790229797, "num_tokens": 545123885.0, "step": 14363 }, { "epoch": 1.8272484416740873, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7195756435394287, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8667479753494263, "num_tokens": 545160295.0, "step": 14364 }, { "epoch": 1.8273756519526778, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5717804431915283, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8599851131439209, "num_tokens": 545203469.0, "step": 14365 }, { "epoch": 1.8275028622312681, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7296764850616455, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8783631324768066, "num_tokens": 545242173.0, "step": 14366 }, { "epoch": 1.8276300725098586, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5591585636138916, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8655611872673035, "num_tokens": 545284968.0, "step": 14367 }, { "epoch": 1.8277572827884492, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7589643001556396, "learning_rate": 1e-05, "loss": 0.4953, "mean_token_accuracy": 0.856582760810852, "num_tokens": 545324764.0, "step": 14368 }, { "epoch": 1.8278844930670397, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.553306818008423, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8664751052856445, "num_tokens": 545362783.0, "step": 14369 }, { "epoch": 1.8280117033456302, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.779883861541748, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8811513185501099, "num_tokens": 545395757.0, "step": 14370 }, { "epoch": 1.8281389136242208, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5430901050567627, "learning_rate": 1e-05, "loss": 0.3955, "mean_token_accuracy": 0.8858916163444519, "num_tokens": 545432194.0, "step": 14371 }, { "epoch": 1.8282661239028113, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7967426776885986, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.876907467842102, "num_tokens": 545469743.0, "step": 14372 }, { "epoch": 1.8283933341814018, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5499625205993652, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8656381368637085, "num_tokens": 545508976.0, "step": 14373 }, { "epoch": 1.8285205444599923, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8050577640533447, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8729463815689087, "num_tokens": 545548223.0, "step": 14374 }, { "epoch": 1.8286477547385829, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5661823749542236, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8560092449188232, "num_tokens": 545586110.0, "step": 14375 }, { "epoch": 1.8287749650171734, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.804736375808716, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8691872358322144, "num_tokens": 545625779.0, "step": 14376 }, { "epoch": 1.828902175295764, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.570617198944092, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.8512631058692932, "num_tokens": 545669464.0, "step": 14377 }, { "epoch": 1.8290293855743545, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.807234287261963, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8802347183227539, "num_tokens": 545702674.0, "step": 14378 }, { "epoch": 1.829156595852945, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5618855953216553, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8765602111816406, "num_tokens": 545742410.0, "step": 14379 }, { "epoch": 1.8292838061315355, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8463516235351562, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8650844693183899, "num_tokens": 545779748.0, "step": 14380 }, { "epoch": 1.829411016410126, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5792236328125, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.865405797958374, "num_tokens": 545811401.0, "step": 14381 }, { "epoch": 1.8295382266887166, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.813849449157715, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8791744709014893, "num_tokens": 545848565.0, "step": 14382 }, { "epoch": 1.829665436967307, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5639541149139404, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.868310809135437, "num_tokens": 545891245.0, "step": 14383 }, { "epoch": 1.8297926472458976, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8209900856018066, "learning_rate": 1e-05, "loss": 0.3819, "mean_token_accuracy": 0.8907393217086792, "num_tokens": 545924860.0, "step": 14384 }, { "epoch": 1.829919857524488, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5822317600250244, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.881920337677002, "num_tokens": 545958119.0, "step": 14385 }, { "epoch": 1.8300470678030785, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8236751556396484, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8657491207122803, "num_tokens": 545994195.0, "step": 14386 }, { "epoch": 1.830174278081669, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.585035562515259, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8721239566802979, "num_tokens": 546031998.0, "step": 14387 }, { "epoch": 1.8303014883602595, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8135874271392822, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8719800114631653, "num_tokens": 546067152.0, "step": 14388 }, { "epoch": 1.83042869863885, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.582106590270996, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8856064081192017, "num_tokens": 546104207.0, "step": 14389 }, { "epoch": 1.8305559089174406, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.796689748764038, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.860034704208374, "num_tokens": 546150659.0, "step": 14390 }, { "epoch": 1.8306831191960309, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.591946601867676, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8781216144561768, "num_tokens": 546188427.0, "step": 14391 }, { "epoch": 1.8308103294746214, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7836050987243652, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.864008367061615, "num_tokens": 546227249.0, "step": 14392 }, { "epoch": 1.830937539753212, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6144216060638428, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8615632057189941, "num_tokens": 546263469.0, "step": 14393 }, { "epoch": 1.8310647500318025, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.771049976348877, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.869255781173706, "num_tokens": 546300820.0, "step": 14394 }, { "epoch": 1.831191960310393, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.593240261077881, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8772532939910889, "num_tokens": 546336639.0, "step": 14395 }, { "epoch": 1.8313191705889835, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7566745281219482, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8653851747512817, "num_tokens": 546377769.0, "step": 14396 }, { "epoch": 1.831446380867574, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.580035448074341, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8800094127655029, "num_tokens": 546419286.0, "step": 14397 }, { "epoch": 1.8315735911461646, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.762338876724243, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.859896719455719, "num_tokens": 546458907.0, "step": 14398 }, { "epoch": 1.831700801424755, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5923242568969727, "learning_rate": 1e-05, "loss": 0.3752, "mean_token_accuracy": 0.8904194831848145, "num_tokens": 546491682.0, "step": 14399 }, { "epoch": 1.8318280117033456, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7505030632019043, "learning_rate": 1e-05, "loss": 0.5066, "mean_token_accuracy": 0.8521659970283508, "num_tokens": 546537141.0, "step": 14400 }, { "epoch": 1.8319552219819362, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.582029104232788, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8686699867248535, "num_tokens": 546574318.0, "step": 14401 }, { "epoch": 1.8320824322605267, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7508373260498047, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8620786666870117, "num_tokens": 546615938.0, "step": 14402 }, { "epoch": 1.8322096425391172, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5962727069854736, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8738886117935181, "num_tokens": 546652608.0, "step": 14403 }, { "epoch": 1.8323368528177078, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.724799394607544, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.8844090700149536, "num_tokens": 546695363.0, "step": 14404 }, { "epoch": 1.8324640630962983, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5936760902404785, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8574407696723938, "num_tokens": 546741174.0, "step": 14405 }, { "epoch": 1.8325912733748888, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7208971977233887, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8794176578521729, "num_tokens": 546777827.0, "step": 14406 }, { "epoch": 1.8327184836534793, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.577413320541382, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8786753416061401, "num_tokens": 546821199.0, "step": 14407 }, { "epoch": 1.8328456939320699, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7308268547058105, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8718470335006714, "num_tokens": 546860151.0, "step": 14408 }, { "epoch": 1.8329729042106602, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5886118412017822, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8809086680412292, "num_tokens": 546892456.0, "step": 14409 }, { "epoch": 1.8331001144892507, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.737426519393921, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8645608425140381, "num_tokens": 546931611.0, "step": 14410 }, { "epoch": 1.8332273247678412, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5788328647613525, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8774623870849609, "num_tokens": 546968408.0, "step": 14411 }, { "epoch": 1.8333545350464318, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.740959644317627, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8594412207603455, "num_tokens": 547007025.0, "step": 14412 }, { "epoch": 1.8334817453250223, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.564771890640259, "learning_rate": 1e-05, "loss": 0.4809, "mean_token_accuracy": 0.8585246801376343, "num_tokens": 547051371.0, "step": 14413 }, { "epoch": 1.8336089556036128, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.757627487182617, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.864800214767456, "num_tokens": 547087476.0, "step": 14414 }, { "epoch": 1.8337361658822031, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.566990375518799, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8755767345428467, "num_tokens": 547126922.0, "step": 14415 }, { "epoch": 1.8338633761607936, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7633159160614014, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8774725198745728, "num_tokens": 547164678.0, "step": 14416 }, { "epoch": 1.8339905864393842, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.573042154312134, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.86974036693573, "num_tokens": 547200039.0, "step": 14417 }, { "epoch": 1.8341177967179747, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.767411947250366, "learning_rate": 1e-05, "loss": 0.3803, "mean_token_accuracy": 0.8910709023475647, "num_tokens": 547233325.0, "step": 14418 }, { "epoch": 1.8342450069965652, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5679194927215576, "learning_rate": 1e-05, "loss": 0.3965, "mean_token_accuracy": 0.8853849172592163, "num_tokens": 547266658.0, "step": 14419 }, { "epoch": 1.8343722172751558, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7678072452545166, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8642752766609192, "num_tokens": 547305876.0, "step": 14420 }, { "epoch": 1.8344994275537463, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.554173707962036, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8768195509910583, "num_tokens": 547346847.0, "step": 14421 }, { "epoch": 1.8346266378323368, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.781489372253418, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8833765387535095, "num_tokens": 547379896.0, "step": 14422 }, { "epoch": 1.8347538481109273, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5498597621917725, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8667542934417725, "num_tokens": 547425666.0, "step": 14423 }, { "epoch": 1.8348810583895179, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.784181833267212, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8783373236656189, "num_tokens": 547467007.0, "step": 14424 }, { "epoch": 1.8350082686681084, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5654327869415283, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8528720140457153, "num_tokens": 547507718.0, "step": 14425 }, { "epoch": 1.835135478946699, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7961299419403076, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8765945434570312, "num_tokens": 547543445.0, "step": 14426 }, { "epoch": 1.8352626892252895, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5793566703796387, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8742994070053101, "num_tokens": 547576000.0, "step": 14427 }, { "epoch": 1.83538989950388, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7952911853790283, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8774989247322083, "num_tokens": 547609950.0, "step": 14428 }, { "epoch": 1.8355171097824705, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5623881816864014, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8778104782104492, "num_tokens": 547650133.0, "step": 14429 }, { "epoch": 1.835644320061061, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.798153877258301, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8673427700996399, "num_tokens": 547687754.0, "step": 14430 }, { "epoch": 1.8357715303396516, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5646443367004395, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8754750490188599, "num_tokens": 547732119.0, "step": 14431 }, { "epoch": 1.835898740618242, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.799670696258545, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8777536749839783, "num_tokens": 547770779.0, "step": 14432 }, { "epoch": 1.8360259508968326, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.565859079360962, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8787610530853271, "num_tokens": 547812524.0, "step": 14433 }, { "epoch": 1.836153161175423, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8067901134490967, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8649669885635376, "num_tokens": 547850955.0, "step": 14434 }, { "epoch": 1.8362803714540135, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5863053798675537, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8744972348213196, "num_tokens": 547891390.0, "step": 14435 }, { "epoch": 1.836407581732604, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.800840377807617, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8660620450973511, "num_tokens": 547930306.0, "step": 14436 }, { "epoch": 1.8365347920111945, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5833957195281982, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8640503883361816, "num_tokens": 547970940.0, "step": 14437 }, { "epoch": 1.836662002289785, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7859952449798584, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8800821900367737, "num_tokens": 548007908.0, "step": 14438 }, { "epoch": 1.8367892125683756, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.59858775138855, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8679880499839783, "num_tokens": 548045040.0, "step": 14439 }, { "epoch": 1.8369164228469659, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7617340087890625, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8734060525894165, "num_tokens": 548088501.0, "step": 14440 }, { "epoch": 1.8370436331255564, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.610567331314087, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8601851463317871, "num_tokens": 548121888.0, "step": 14441 }, { "epoch": 1.837170843404147, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7425718307495117, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.87424635887146, "num_tokens": 548161117.0, "step": 14442 }, { "epoch": 1.8372980536827375, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.590918779373169, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8826770782470703, "num_tokens": 548201709.0, "step": 14443 }, { "epoch": 1.837425263961328, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7377827167510986, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8714690208435059, "num_tokens": 548241843.0, "step": 14444 }, { "epoch": 1.8375524742399185, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5919992923736572, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.878636360168457, "num_tokens": 548278785.0, "step": 14445 }, { "epoch": 1.837679684518509, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.740034580230713, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8768882751464844, "num_tokens": 548315766.0, "step": 14446 }, { "epoch": 1.8378068947970996, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.594093084335327, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8685001730918884, "num_tokens": 548353971.0, "step": 14447 }, { "epoch": 1.83793410507569, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7408034801483154, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8722598552703857, "num_tokens": 548391261.0, "step": 14448 }, { "epoch": 1.8380613153542806, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.581490993499756, "learning_rate": 1e-05, "loss": 0.398, "mean_token_accuracy": 0.8833028078079224, "num_tokens": 548424773.0, "step": 14449 }, { "epoch": 1.8381885256328712, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7295608520507812, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.888140082359314, "num_tokens": 548462404.0, "step": 14450 }, { "epoch": 1.8383157359114617, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.594479560852051, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8738702535629272, "num_tokens": 548504595.0, "step": 14451 }, { "epoch": 1.8384429461900522, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7214295864105225, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.880336582660675, "num_tokens": 548547945.0, "step": 14452 }, { "epoch": 1.8385701564686427, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.577367067337036, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8620074987411499, "num_tokens": 548590302.0, "step": 14453 }, { "epoch": 1.8386973667472333, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7241251468658447, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8827795386314392, "num_tokens": 548629688.0, "step": 14454 }, { "epoch": 1.8388245770258238, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5913479328155518, "learning_rate": 1e-05, "loss": 0.5009, "mean_token_accuracy": 0.8526136875152588, "num_tokens": 548666294.0, "step": 14455 }, { "epoch": 1.8389517873044143, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7148571014404297, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8594881296157837, "num_tokens": 548707536.0, "step": 14456 }, { "epoch": 1.8390789975830049, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5766758918762207, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8783800005912781, "num_tokens": 548751443.0, "step": 14457 }, { "epoch": 1.8392062078615952, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7172441482543945, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8699828386306763, "num_tokens": 548789259.0, "step": 14458 }, { "epoch": 1.8393334181401857, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5836174488067627, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8755738735198975, "num_tokens": 548822045.0, "step": 14459 }, { "epoch": 1.8394606284187762, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7356865406036377, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8598384857177734, "num_tokens": 548854722.0, "step": 14460 }, { "epoch": 1.8395878386973668, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5850987434387207, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.867042064666748, "num_tokens": 548888903.0, "step": 14461 }, { "epoch": 1.8397150489759573, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.726992607116699, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8764468431472778, "num_tokens": 548925611.0, "step": 14462 }, { "epoch": 1.8398422592545478, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.566488265991211, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8794145584106445, "num_tokens": 548967239.0, "step": 14463 }, { "epoch": 1.8399694695331381, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.726608991622925, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8775292634963989, "num_tokens": 549005453.0, "step": 14464 }, { "epoch": 1.8400966798117286, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5917563438415527, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8696464896202087, "num_tokens": 549041418.0, "step": 14465 }, { "epoch": 1.8402238900903192, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7274649143218994, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8805248737335205, "num_tokens": 549079223.0, "step": 14466 }, { "epoch": 1.8403511003689097, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.575711250305176, "learning_rate": 1e-05, "loss": 0.5132, "mean_token_accuracy": 0.8507812023162842, "num_tokens": 549124542.0, "step": 14467 }, { "epoch": 1.8404783106475002, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.753215789794922, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8621987104415894, "num_tokens": 549165374.0, "step": 14468 }, { "epoch": 1.8406055209260908, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.58223295211792, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8602883219718933, "num_tokens": 549203887.0, "step": 14469 }, { "epoch": 1.8407327312046813, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7272026538848877, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8704404234886169, "num_tokens": 549247583.0, "step": 14470 }, { "epoch": 1.8408599414832718, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5667121410369873, "learning_rate": 1e-05, "loss": 0.3844, "mean_token_accuracy": 0.8889333009719849, "num_tokens": 549282874.0, "step": 14471 }, { "epoch": 1.8409871517618623, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7313168048858643, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8793354034423828, "num_tokens": 549323133.0, "step": 14472 }, { "epoch": 1.8411143620404529, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 7.891940116882324, "learning_rate": 1e-05, "loss": 0.5318, "mean_token_accuracy": 0.8567990660667419, "num_tokens": 549359794.0, "step": 14473 }, { "epoch": 1.8412415723190434, "ewc_loss": 0.1064453125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 7.867813110351562e-05, "grad_norm": 3.728137731552124, "learning_rate": 1e-05, "loss": 0.5384, "mean_token_accuracy": 0.8540048599243164, "num_tokens": 549393281.0, "step": 14474 }, { "epoch": 1.841368782597634, "ewc_loss": 0.06298828125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 3.504753112792969e-05, "grad_norm": 2.2490649223327637, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8676046133041382, "num_tokens": 549428181.0, "step": 14475 }, { "epoch": 1.8414959928762245, "ewc_loss": 0.09326171875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 6.532669067382812e-05, "grad_norm": 3.302921772003174, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8783787488937378, "num_tokens": 549465521.0, "step": 14476 }, { "epoch": 1.841623203154815, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 5.841255187988281e-05, "grad_norm": 2.733832359313965, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8713794946670532, "num_tokens": 549503800.0, "step": 14477 }, { "epoch": 1.8417504134334055, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 6.008148193359375e-05, "grad_norm": 3.1108596324920654, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8829941153526306, "num_tokens": 549536725.0, "step": 14478 }, { "epoch": 1.841877623711996, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 5.602836608886719e-05, "grad_norm": 2.786959409713745, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8726656436920166, "num_tokens": 549575774.0, "step": 14479 }, { "epoch": 1.8420048339905866, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 5.888938903808594e-05, "grad_norm": 3.0172154903411865, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8735820055007935, "num_tokens": 549616715.0, "step": 14480 }, { "epoch": 1.842132044269177, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 5.2928924560546875e-05, "grad_norm": 2.7855112552642822, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8715598583221436, "num_tokens": 549652005.0, "step": 14481 }, { "epoch": 1.8422592545477676, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 5.507469177246094e-05, "grad_norm": 2.9393248558044434, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8700538873672485, "num_tokens": 549683708.0, "step": 14482 }, { "epoch": 1.842386464826358, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 2.7309093475341797, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8860281109809875, "num_tokens": 549722961.0, "step": 14483 }, { "epoch": 1.8425136751049485, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 5.316734313964844e-05, "grad_norm": 2.8856747150421143, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8642675280570984, "num_tokens": 549758857.0, "step": 14484 }, { "epoch": 1.842640885383539, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.709184169769287, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8726643323898315, "num_tokens": 549793835.0, "step": 14485 }, { "epoch": 1.8427680956621295, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.982948303222656e-05, "grad_norm": 2.8258872032165527, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8738983869552612, "num_tokens": 549832691.0, "step": 14486 }, { "epoch": 1.84289530594072, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.6922922134399414, "learning_rate": 1e-05, "loss": 0.5806, "mean_token_accuracy": 0.8339691162109375, "num_tokens": 549874981.0, "step": 14487 }, { "epoch": 1.8430225162193106, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.764786958694458, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.871712863445282, "num_tokens": 549917904.0, "step": 14488 }, { "epoch": 1.8431497264979009, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002765655517578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.6644742488861084, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8702428340911865, "num_tokens": 549955421.0, "step": 14489 }, { "epoch": 1.8432769367764914, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7524185180664062, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.876383900642395, "num_tokens": 549986615.0, "step": 14490 }, { "epoch": 1.843404147055082, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.643252372741699, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8640279769897461, "num_tokens": 550022747.0, "step": 14491 }, { "epoch": 1.8435313573336725, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7261974811553955, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8714993596076965, "num_tokens": 550060423.0, "step": 14492 }, { "epoch": 1.843658567612263, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6181318759918213, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8713380098342896, "num_tokens": 550102086.0, "step": 14493 }, { "epoch": 1.8437857778908535, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.709291458129883, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.8867486715316772, "num_tokens": 550138281.0, "step": 14494 }, { "epoch": 1.843912988169444, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.613661527633667, "learning_rate": 1e-05, "loss": 0.4907, "mean_token_accuracy": 0.856701672077179, "num_tokens": 550180219.0, "step": 14495 }, { "epoch": 1.8440401984480346, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.714916944503784, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8724863529205322, "num_tokens": 550218095.0, "step": 14496 }, { "epoch": 1.844167408726625, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.606184244155884, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8664823770523071, "num_tokens": 550256080.0, "step": 14497 }, { "epoch": 1.8442946190052156, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7174413204193115, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8571969270706177, "num_tokens": 550297471.0, "step": 14498 }, { "epoch": 1.8444218292838062, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6001272201538086, "learning_rate": 1e-05, "loss": 0.3869, "mean_token_accuracy": 0.8898510932922363, "num_tokens": 550330495.0, "step": 14499 }, { "epoch": 1.8445490395623967, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7326343059539795, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8711875677108765, "num_tokens": 550362810.0, "step": 14500 }, { "epoch": 1.8446762498409872, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.578437328338623, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8704262375831604, "num_tokens": 550409408.0, "step": 14501 }, { "epoch": 1.8448034601195777, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.754647731781006, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8650144338607788, "num_tokens": 550441366.0, "step": 14502 }, { "epoch": 1.8449306703981683, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.58294939994812, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8790066838264465, "num_tokens": 550478247.0, "step": 14503 }, { "epoch": 1.8450578806767588, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.746183395385742, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8789263963699341, "num_tokens": 550511861.0, "step": 14504 }, { "epoch": 1.8451850909553493, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.565798044204712, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8659729361534119, "num_tokens": 550554959.0, "step": 14505 }, { "epoch": 1.8453123012339399, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7530364990234375, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8754400610923767, "num_tokens": 550589960.0, "step": 14506 }, { "epoch": 1.8454395115125302, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5666751861572266, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8717321157455444, "num_tokens": 550630481.0, "step": 14507 }, { "epoch": 1.8455667217911207, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7648727893829346, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8620530366897583, "num_tokens": 550671819.0, "step": 14508 }, { "epoch": 1.8456939320697112, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.564148426055908, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8799266815185547, "num_tokens": 550712230.0, "step": 14509 }, { "epoch": 1.8458211423483017, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7724831104278564, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8703832626342773, "num_tokens": 550751052.0, "step": 14510 }, { "epoch": 1.8459483526268923, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.583455801010132, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8775684237480164, "num_tokens": 550787374.0, "step": 14511 }, { "epoch": 1.8460755629054828, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.752589464187622, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8787485361099243, "num_tokens": 550829398.0, "step": 14512 }, { "epoch": 1.846202773184073, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.573242664337158, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8628782629966736, "num_tokens": 550868818.0, "step": 14513 }, { "epoch": 1.8463299834626636, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7634027004241943, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8775750994682312, "num_tokens": 550904136.0, "step": 14514 }, { "epoch": 1.8464571937412542, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.58500075340271, "learning_rate": 1e-05, "loss": 0.551, "mean_token_accuracy": 0.8394414186477661, "num_tokens": 550943505.0, "step": 14515 }, { "epoch": 1.8465844040198447, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.752229690551758, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8701574802398682, "num_tokens": 550983633.0, "step": 14516 }, { "epoch": 1.8467116142984352, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.581716775894165, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8808401823043823, "num_tokens": 551017683.0, "step": 14517 }, { "epoch": 1.8468388245770258, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.757167339324951, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8696920871734619, "num_tokens": 551051731.0, "step": 14518 }, { "epoch": 1.8469660348556163, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.572517156600952, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8771337270736694, "num_tokens": 551090818.0, "step": 14519 }, { "epoch": 1.8470932451342068, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.765218496322632, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8790847659111023, "num_tokens": 551122032.0, "step": 14520 }, { "epoch": 1.8472204554127973, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5814411640167236, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8641869425773621, "num_tokens": 551157708.0, "step": 14521 }, { "epoch": 1.8473476656913879, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.763031482696533, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8718916177749634, "num_tokens": 551194222.0, "step": 14522 }, { "epoch": 1.8474748759699784, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5739192962646484, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8702752590179443, "num_tokens": 551234387.0, "step": 14523 }, { "epoch": 1.847602086248569, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7598369121551514, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8762837648391724, "num_tokens": 551273330.0, "step": 14524 }, { "epoch": 1.8477292965271594, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5799665451049805, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8730499744415283, "num_tokens": 551309180.0, "step": 14525 }, { "epoch": 1.84785650680575, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.755410671234131, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.8840560913085938, "num_tokens": 551344455.0, "step": 14526 }, { "epoch": 1.8479837170843405, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5834431648254395, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8628367781639099, "num_tokens": 551380795.0, "step": 14527 }, { "epoch": 1.848110927362931, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7577385902404785, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.8601018190383911, "num_tokens": 551421314.0, "step": 14528 }, { "epoch": 1.8482381376415216, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5733256340026855, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8748696446418762, "num_tokens": 551459021.0, "step": 14529 }, { "epoch": 1.848365347920112, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.76076078414917, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8719760179519653, "num_tokens": 551498055.0, "step": 14530 }, { "epoch": 1.8484925581987026, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5859665870666504, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8669109344482422, "num_tokens": 551536746.0, "step": 14531 }, { "epoch": 1.848619768477293, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.75144362449646, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8612505197525024, "num_tokens": 551576230.0, "step": 14532 }, { "epoch": 1.8487469787558835, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5950493812561035, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8718392848968506, "num_tokens": 551613476.0, "step": 14533 }, { "epoch": 1.848874189034474, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7294764518737793, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8727911114692688, "num_tokens": 551655177.0, "step": 14534 }, { "epoch": 1.8490013993130645, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.588179349899292, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8674722909927368, "num_tokens": 551691236.0, "step": 14535 }, { "epoch": 1.849128609591655, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.726386308670044, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8616628646850586, "num_tokens": 551732152.0, "step": 14536 }, { "epoch": 1.8492558198702456, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.594790458679199, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8771366477012634, "num_tokens": 551765221.0, "step": 14537 }, { "epoch": 1.8493830301488359, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7173027992248535, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8727237582206726, "num_tokens": 551805321.0, "step": 14538 }, { "epoch": 1.8495102404274264, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5788064002990723, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8644648790359497, "num_tokens": 551846752.0, "step": 14539 }, { "epoch": 1.849637450706017, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7229535579681396, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8690558671951294, "num_tokens": 551887877.0, "step": 14540 }, { "epoch": 1.8497646609846075, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5983734130859375, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8647093176841736, "num_tokens": 551919669.0, "step": 14541 }, { "epoch": 1.849891871263198, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.718409776687622, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8733435869216919, "num_tokens": 551955459.0, "step": 14542 }, { "epoch": 1.8500190815417885, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6172118186950684, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8691326379776001, "num_tokens": 551994889.0, "step": 14543 }, { "epoch": 1.850146291820379, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6999988555908203, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8737310171127319, "num_tokens": 552030176.0, "step": 14544 }, { "epoch": 1.8502735020989696, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5939934253692627, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8844941854476929, "num_tokens": 552063317.0, "step": 14545 }, { "epoch": 1.85040071237756, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6902716159820557, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.8901876211166382, "num_tokens": 552100959.0, "step": 14546 }, { "epoch": 1.8505279226561506, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.599332809448242, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.87763512134552, "num_tokens": 552139181.0, "step": 14547 }, { "epoch": 1.8506551329347412, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.702185869216919, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8667645454406738, "num_tokens": 552172293.0, "step": 14548 }, { "epoch": 1.8507823432133317, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.58251690864563, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8685187101364136, "num_tokens": 552214655.0, "step": 14549 }, { "epoch": 1.8509095534919222, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7038917541503906, "learning_rate": 1e-05, "loss": 0.3855, "mean_token_accuracy": 0.8911036849021912, "num_tokens": 552254188.0, "step": 14550 }, { "epoch": 1.8510367637705127, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6067042350769043, "learning_rate": 1e-05, "loss": 0.499, "mean_token_accuracy": 0.8530834913253784, "num_tokens": 552286712.0, "step": 14551 }, { "epoch": 1.8511639740491033, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6993322372436523, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8711323738098145, "num_tokens": 552326846.0, "step": 14552 }, { "epoch": 1.8512911843276938, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5995872020721436, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8827465176582336, "num_tokens": 552356753.0, "step": 14553 }, { "epoch": 1.8514183946062843, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.705573320388794, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.862513542175293, "num_tokens": 552391312.0, "step": 14554 }, { "epoch": 1.8515456048848749, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6100709438323975, "learning_rate": 1e-05, "loss": 0.5122, "mean_token_accuracy": 0.847437858581543, "num_tokens": 552428782.0, "step": 14555 }, { "epoch": 1.8516728151634652, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6870574951171875, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8722267746925354, "num_tokens": 552471131.0, "step": 14556 }, { "epoch": 1.8518000254420557, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.593698740005493, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8755941390991211, "num_tokens": 552512242.0, "step": 14557 }, { "epoch": 1.8519272357206462, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.707768440246582, "learning_rate": 1e-05, "loss": 0.5099, "mean_token_accuracy": 0.8513895273208618, "num_tokens": 552545655.0, "step": 14558 }, { "epoch": 1.8520544459992367, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5842034816741943, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8701884746551514, "num_tokens": 552586340.0, "step": 14559 }, { "epoch": 1.8521816562778273, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7048122882843018, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8714946508407593, "num_tokens": 552624461.0, "step": 14560 }, { "epoch": 1.8523088665564178, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5718343257904053, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8824076652526855, "num_tokens": 552661566.0, "step": 14561 }, { "epoch": 1.852436076835008, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.717651128768921, "learning_rate": 1e-05, "loss": 0.4017, "mean_token_accuracy": 0.882543683052063, "num_tokens": 552698184.0, "step": 14562 }, { "epoch": 1.8525632871135986, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.559424638748169, "learning_rate": 1e-05, "loss": 0.3855, "mean_token_accuracy": 0.8899323344230652, "num_tokens": 552741407.0, "step": 14563 }, { "epoch": 1.8526904973921892, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7372517585754395, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8645153045654297, "num_tokens": 552780038.0, "step": 14564 }, { "epoch": 1.8528177076707797, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.559135675430298, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8778256177902222, "num_tokens": 552818479.0, "step": 14565 }, { "epoch": 1.8529449179493702, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7535903453826904, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8788946866989136, "num_tokens": 552853762.0, "step": 14566 }, { "epoch": 1.8530721282279607, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.563384532928467, "learning_rate": 1e-05, "loss": 0.5063, "mean_token_accuracy": 0.8459465503692627, "num_tokens": 552892693.0, "step": 14567 }, { "epoch": 1.8531993385065513, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7629785537719727, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8782128095626831, "num_tokens": 552930127.0, "step": 14568 }, { "epoch": 1.8533265487851418, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5528435707092285, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8721907734870911, "num_tokens": 552969809.0, "step": 14569 }, { "epoch": 1.8534537590637323, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7824642658233643, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8786159157752991, "num_tokens": 553004747.0, "step": 14570 }, { "epoch": 1.8535809693423229, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5438263416290283, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8594530820846558, "num_tokens": 553049727.0, "step": 14571 }, { "epoch": 1.8537081796209134, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.793574094772339, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8706650733947754, "num_tokens": 553093263.0, "step": 14572 }, { "epoch": 1.853835389899504, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.566608190536499, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8707879781723022, "num_tokens": 553126198.0, "step": 14573 }, { "epoch": 1.8539626001780944, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8150129318237305, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8767871856689453, "num_tokens": 553156798.0, "step": 14574 }, { "epoch": 1.854089810456685, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5498740673065186, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.873932957649231, "num_tokens": 553193612.0, "step": 14575 }, { "epoch": 1.8542170207352755, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.830575942993164, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8777801394462585, "num_tokens": 553227376.0, "step": 14576 }, { "epoch": 1.854344231013866, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5630319118499756, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8685815334320068, "num_tokens": 553263045.0, "step": 14577 }, { "epoch": 1.8544714412924566, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8306808471679688, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8644623756408691, "num_tokens": 553301834.0, "step": 14578 }, { "epoch": 1.854598651571047, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5668017864227295, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8604422211647034, "num_tokens": 553345980.0, "step": 14579 }, { "epoch": 1.8547258618496376, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8331987857818604, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8684371113777161, "num_tokens": 553380645.0, "step": 14580 }, { "epoch": 1.854853072128228, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5801126956939697, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8638336062431335, "num_tokens": 553417952.0, "step": 14581 }, { "epoch": 1.8549802824068184, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8226094245910645, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8774218559265137, "num_tokens": 553454575.0, "step": 14582 }, { "epoch": 1.855107492685409, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.577322244644165, "learning_rate": 1e-05, "loss": 0.3736, "mean_token_accuracy": 0.8928217887878418, "num_tokens": 553489846.0, "step": 14583 }, { "epoch": 1.8552347029639995, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8228702545166016, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.879530668258667, "num_tokens": 553524444.0, "step": 14584 }, { "epoch": 1.85536191324259, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.58620548248291, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8771886229515076, "num_tokens": 553562295.0, "step": 14585 }, { "epoch": 1.8554891235211806, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8181633949279785, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8555120229721069, "num_tokens": 553600240.0, "step": 14586 }, { "epoch": 1.8556163337997709, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.575741767883301, "learning_rate": 1e-05, "loss": 0.3881, "mean_token_accuracy": 0.8868622779846191, "num_tokens": 553641215.0, "step": 14587 }, { "epoch": 1.8557435440783614, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.812674045562744, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8865972757339478, "num_tokens": 553675599.0, "step": 14588 }, { "epoch": 1.855870754356952, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.590264320373535, "learning_rate": 1e-05, "loss": 0.3928, "mean_token_accuracy": 0.8840038180351257, "num_tokens": 553706681.0, "step": 14589 }, { "epoch": 1.8559979646355425, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.806201457977295, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8739898204803467, "num_tokens": 553740919.0, "step": 14590 }, { "epoch": 1.856125174914133, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.616816282272339, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8763362765312195, "num_tokens": 553775509.0, "step": 14591 }, { "epoch": 1.8562523851927235, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7808475494384766, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8638520240783691, "num_tokens": 553809168.0, "step": 14592 }, { "epoch": 1.856379595471314, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5955333709716797, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.883307933807373, "num_tokens": 553848186.0, "step": 14593 }, { "epoch": 1.8565068057499046, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7614023685455322, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8761042356491089, "num_tokens": 553888928.0, "step": 14594 }, { "epoch": 1.856634016028495, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.602820873260498, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8606679439544678, "num_tokens": 553929095.0, "step": 14595 }, { "epoch": 1.8567612263070856, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7488067150115967, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8794532418251038, "num_tokens": 553969610.0, "step": 14596 }, { "epoch": 1.8568884365856761, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6049857139587402, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8734573125839233, "num_tokens": 554006707.0, "step": 14597 }, { "epoch": 1.8570156468642667, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.728025197982788, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.862582802772522, "num_tokens": 554049006.0, "step": 14598 }, { "epoch": 1.8571428571428572, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.598686933517456, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8748891353607178, "num_tokens": 554086682.0, "step": 14599 }, { "epoch": 1.8572700674214477, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7301788330078125, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.867937445640564, "num_tokens": 554122478.0, "step": 14600 }, { "epoch": 1.8573972777000383, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.581960678100586, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8790095448493958, "num_tokens": 554162640.0, "step": 14601 }, { "epoch": 1.8575244879786288, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7248311042785645, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8661905527114868, "num_tokens": 554202161.0, "step": 14602 }, { "epoch": 1.8576516982572193, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.599806785583496, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8770401477813721, "num_tokens": 554235401.0, "step": 14603 }, { "epoch": 1.8577789085358098, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7224700450897217, "learning_rate": 1e-05, "loss": 0.5121, "mean_token_accuracy": 0.8527108430862427, "num_tokens": 554276887.0, "step": 14604 }, { "epoch": 1.8579061188144002, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6123201847076416, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8542001247406006, "num_tokens": 554310513.0, "step": 14605 }, { "epoch": 1.8580333290929907, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6953377723693848, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8745319247245789, "num_tokens": 554354225.0, "step": 14606 }, { "epoch": 1.8581605393715812, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.592573404312134, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8653609752655029, "num_tokens": 554395198.0, "step": 14607 }, { "epoch": 1.8582877496501717, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.706986904144287, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8617935180664062, "num_tokens": 554430234.0, "step": 14608 }, { "epoch": 1.8584149599287623, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.586019992828369, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8710622787475586, "num_tokens": 554469381.0, "step": 14609 }, { "epoch": 1.8585421702073528, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7292702198028564, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8679530024528503, "num_tokens": 554500409.0, "step": 14610 }, { "epoch": 1.858669380485943, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5721380710601807, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8802189826965332, "num_tokens": 554537785.0, "step": 14611 }, { "epoch": 1.8587965907645336, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.730687141418457, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.862403929233551, "num_tokens": 554578420.0, "step": 14612 }, { "epoch": 1.8589238010431242, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.57781982421875, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8747366070747375, "num_tokens": 554615264.0, "step": 14613 }, { "epoch": 1.8590510113217147, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7477989196777344, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8711644411087036, "num_tokens": 554647327.0, "step": 14614 }, { "epoch": 1.8591782216003052, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5688788890838623, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8727865219116211, "num_tokens": 554682552.0, "step": 14615 }, { "epoch": 1.8593054318788957, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.742488384246826, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8642176389694214, "num_tokens": 554721348.0, "step": 14616 }, { "epoch": 1.8594326421574863, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5561258792877197, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8807507753372192, "num_tokens": 554760618.0, "step": 14617 }, { "epoch": 1.8595598524360768, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.757567882537842, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8655357360839844, "num_tokens": 554801831.0, "step": 14618 }, { "epoch": 1.8596870627146673, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.566161632537842, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8763246536254883, "num_tokens": 554838307.0, "step": 14619 }, { "epoch": 1.8598142729932579, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7572195529937744, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8678692579269409, "num_tokens": 554883833.0, "step": 14620 }, { "epoch": 1.8599414832718484, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.558274507522583, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8703241348266602, "num_tokens": 554920944.0, "step": 14621 }, { "epoch": 1.860068693550439, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7725610733032227, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8650133013725281, "num_tokens": 554958656.0, "step": 14622 }, { "epoch": 1.8601959038290294, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.561522960662842, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8644071221351624, "num_tokens": 555001864.0, "step": 14623 }, { "epoch": 1.86032311410762, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7935359477996826, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8765923976898193, "num_tokens": 555034268.0, "step": 14624 }, { "epoch": 1.8604503243862105, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5536892414093018, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8673639297485352, "num_tokens": 555075275.0, "step": 14625 }, { "epoch": 1.860577534664801, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.824456214904785, "learning_rate": 1e-05, "loss": 0.5499, "mean_token_accuracy": 0.8408977389335632, "num_tokens": 555110961.0, "step": 14626 }, { "epoch": 1.8607047449433916, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.556133508682251, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8574166297912598, "num_tokens": 555150369.0, "step": 14627 }, { "epoch": 1.860831955221982, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8330507278442383, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8623966574668884, "num_tokens": 555182555.0, "step": 14628 }, { "epoch": 1.8609591655005726, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.62274169921875, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8853191137313843, "num_tokens": 555223267.0, "step": 14629 }, { "epoch": 1.861086375779163, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.790309190750122, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8686665892601013, "num_tokens": 555263461.0, "step": 14630 }, { "epoch": 1.8612135860577534, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.568254232406616, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8706446886062622, "num_tokens": 555306201.0, "step": 14631 }, { "epoch": 1.861340796336344, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7958428859710693, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8730266690254211, "num_tokens": 555344418.0, "step": 14632 }, { "epoch": 1.8614680066149345, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5708577632904053, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8688451051712036, "num_tokens": 555386096.0, "step": 14633 }, { "epoch": 1.861595216893525, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7929606437683105, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8753553628921509, "num_tokens": 555423368.0, "step": 14634 }, { "epoch": 1.8617224271721156, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5856664180755615, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8669273853302002, "num_tokens": 555458624.0, "step": 14635 }, { "epoch": 1.8618496374507059, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7945494651794434, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.87784343957901, "num_tokens": 555491433.0, "step": 14636 }, { "epoch": 1.8619768477292964, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5836308002471924, "learning_rate": 1e-05, "loss": 0.3868, "mean_token_accuracy": 0.8870482444763184, "num_tokens": 555527198.0, "step": 14637 }, { "epoch": 1.862104058007887, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7853739261627197, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.878571629524231, "num_tokens": 555563232.0, "step": 14638 }, { "epoch": 1.8622312682864774, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.597567558288574, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8603149652481079, "num_tokens": 555599805.0, "step": 14639 }, { "epoch": 1.862358478565068, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7972490787506104, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8664431571960449, "num_tokens": 555628835.0, "step": 14640 }, { "epoch": 1.8624856888436585, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5777575969696045, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8786346912384033, "num_tokens": 555668120.0, "step": 14641 }, { "epoch": 1.862612899122249, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7957746982574463, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.856308102607727, "num_tokens": 555704662.0, "step": 14642 }, { "epoch": 1.8627401094008396, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.7965826988220215, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8822858929634094, "num_tokens": 555738867.0, "step": 14643 }, { "epoch": 1.86286731967943, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6843607425689697, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8674240112304688, "num_tokens": 555773500.0, "step": 14644 }, { "epoch": 1.8629945299580206, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6412229537963867, "learning_rate": 1e-05, "loss": 0.3964, "mean_token_accuracy": 0.8859480023384094, "num_tokens": 555812139.0, "step": 14645 }, { "epoch": 1.8631217402366111, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6547482013702393, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8604112863540649, "num_tokens": 555850823.0, "step": 14646 }, { "epoch": 1.8632489505152017, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6406266689300537, "learning_rate": 1e-05, "loss": 0.3808, "mean_token_accuracy": 0.8898424506187439, "num_tokens": 555888554.0, "step": 14647 }, { "epoch": 1.8633761607937922, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6802659034729004, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8705594539642334, "num_tokens": 555932629.0, "step": 14648 }, { "epoch": 1.8635033710723827, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6300277709960938, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8633785843849182, "num_tokens": 555970722.0, "step": 14649 }, { "epoch": 1.8636305813509733, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6627025604248047, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.879206657409668, "num_tokens": 556005784.0, "step": 14650 }, { "epoch": 1.8637577916295638, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6074254512786865, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8798006176948547, "num_tokens": 556046053.0, "step": 14651 }, { "epoch": 1.8638850019081543, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6763269901275635, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8658504486083984, "num_tokens": 556084265.0, "step": 14652 }, { "epoch": 1.8640122121867448, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5950517654418945, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8767061829566956, "num_tokens": 556125062.0, "step": 14653 }, { "epoch": 1.8641394224653351, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.684516191482544, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8573591113090515, "num_tokens": 556162590.0, "step": 14654 }, { "epoch": 1.8642666327439257, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6019206047058105, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8683830499649048, "num_tokens": 556200560.0, "step": 14655 }, { "epoch": 1.8643938430225162, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.688650369644165, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8788928389549255, "num_tokens": 556237199.0, "step": 14656 }, { "epoch": 1.8645210533011067, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5914366245269775, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8700699806213379, "num_tokens": 556278620.0, "step": 14657 }, { "epoch": 1.8646482635796973, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.693635940551758, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.876607358455658, "num_tokens": 556312844.0, "step": 14658 }, { "epoch": 1.8647754738582878, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5761964321136475, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8787131309509277, "num_tokens": 556351218.0, "step": 14659 }, { "epoch": 1.864902684136878, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.70536208152771, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8817398548126221, "num_tokens": 556388176.0, "step": 14660 }, { "epoch": 1.8650298944154686, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5848140716552734, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8651990294456482, "num_tokens": 556424996.0, "step": 14661 }, { "epoch": 1.8651571046940592, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7204430103302, "learning_rate": 1e-05, "loss": 0.5116, "mean_token_accuracy": 0.8534518480300903, "num_tokens": 556465961.0, "step": 14662 }, { "epoch": 1.8652843149726497, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5673141479492188, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8733941316604614, "num_tokens": 556504777.0, "step": 14663 }, { "epoch": 1.8654115252512402, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.72572660446167, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8780394792556763, "num_tokens": 556539737.0, "step": 14664 }, { "epoch": 1.8655387355298307, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.560610294342041, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8849565982818604, "num_tokens": 556578028.0, "step": 14665 }, { "epoch": 1.8656659458084213, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7629315853118896, "learning_rate": 1e-05, "loss": 0.4817, "mean_token_accuracy": 0.8610761165618896, "num_tokens": 556611362.0, "step": 14666 }, { "epoch": 1.8657931560870118, "ewc_loss": 0.06689453125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.551286458969116, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8702175617218018, "num_tokens": 556652329.0, "step": 14667 }, { "epoch": 1.8659203663656023, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7652788162231445, "learning_rate": 1e-05, "loss": 0.3724, "mean_token_accuracy": 0.8946621417999268, "num_tokens": 556689531.0, "step": 14668 }, { "epoch": 1.8660475766441929, "ewc_loss": 0.06640625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.886222839355469e-05, "grad_norm": 2.562709331512451, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8792754411697388, "num_tokens": 556722309.0, "step": 14669 }, { "epoch": 1.8661747869227834, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7805395126342773, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8822457790374756, "num_tokens": 556755682.0, "step": 14670 }, { "epoch": 1.866301997201374, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5564184188842773, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8807411193847656, "num_tokens": 556792591.0, "step": 14671 }, { "epoch": 1.8664292074799644, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7735326290130615, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8770058155059814, "num_tokens": 556835168.0, "step": 14672 }, { "epoch": 1.866556417758555, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5547425746917725, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8743659257888794, "num_tokens": 556876106.0, "step": 14673 }, { "epoch": 1.8666836280371455, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.795307159423828, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8677959442138672, "num_tokens": 556917154.0, "step": 14674 }, { "epoch": 1.866810838315736, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.564241409301758, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8674285411834717, "num_tokens": 556953706.0, "step": 14675 }, { "epoch": 1.8669380485943265, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.8006539344787598, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8740605711936951, "num_tokens": 556992035.0, "step": 14676 }, { "epoch": 1.867065258872917, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5518620014190674, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8748446106910706, "num_tokens": 557034785.0, "step": 14677 }, { "epoch": 1.8671924691515076, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.812514305114746, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.863880455493927, "num_tokens": 557074910.0, "step": 14678 }, { "epoch": 1.867319679430098, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.577040195465088, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8740806579589844, "num_tokens": 557105327.0, "step": 14679 }, { "epoch": 1.8674468897086884, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7995057106018066, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8693011999130249, "num_tokens": 557149261.0, "step": 14680 }, { "epoch": 1.867574099987279, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5677058696746826, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8685622215270996, "num_tokens": 557190550.0, "step": 14681 }, { "epoch": 1.8677013102658695, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.81782865524292, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.856573224067688, "num_tokens": 557222828.0, "step": 14682 }, { "epoch": 1.86782852054446, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.569197654724121, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8664528131484985, "num_tokens": 557264228.0, "step": 14683 }, { "epoch": 1.8679557308230506, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8067357540130615, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8906331658363342, "num_tokens": 557301310.0, "step": 14684 }, { "epoch": 1.8680829411016409, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5754833221435547, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.876524806022644, "num_tokens": 557338596.0, "step": 14685 }, { "epoch": 1.8682101513802314, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8183629512786865, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8712563514709473, "num_tokens": 557374939.0, "step": 14686 }, { "epoch": 1.868337361658822, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.582815408706665, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8714816570281982, "num_tokens": 557411711.0, "step": 14687 }, { "epoch": 1.8684645719374124, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.805642604827881, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8743671178817749, "num_tokens": 557448892.0, "step": 14688 }, { "epoch": 1.868591782216003, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.57905912399292, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8662219643592834, "num_tokens": 557490911.0, "step": 14689 }, { "epoch": 1.8687189924945935, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8071956634521484, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8740973472595215, "num_tokens": 557529721.0, "step": 14690 }, { "epoch": 1.868846202773184, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5866904258728027, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8760839104652405, "num_tokens": 557567591.0, "step": 14691 }, { "epoch": 1.8689734130517746, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.790074348449707, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8682023286819458, "num_tokens": 557609105.0, "step": 14692 }, { "epoch": 1.869100623330365, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5916144847869873, "learning_rate": 1e-05, "loss": 0.5419, "mean_token_accuracy": 0.8408668041229248, "num_tokens": 557658798.0, "step": 14693 }, { "epoch": 1.8692278336089556, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.780515670776367, "learning_rate": 1e-05, "loss": 0.3955, "mean_token_accuracy": 0.8868184089660645, "num_tokens": 557695572.0, "step": 14694 }, { "epoch": 1.8693550438875461, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6083199977874756, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8748481273651123, "num_tokens": 557728796.0, "step": 14695 }, { "epoch": 1.8694822541661367, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.756805658340454, "learning_rate": 1e-05, "loss": 0.3994, "mean_token_accuracy": 0.8840641975402832, "num_tokens": 557761786.0, "step": 14696 }, { "epoch": 1.8696094644447272, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5905535221099854, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8637546896934509, "num_tokens": 557809454.0, "step": 14697 }, { "epoch": 1.8697366747233177, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7389938831329346, "learning_rate": 1e-05, "loss": 0.3708, "mean_token_accuracy": 0.8928096294403076, "num_tokens": 557846959.0, "step": 14698 }, { "epoch": 1.8698638850019083, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6050820350646973, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8831352591514587, "num_tokens": 557879873.0, "step": 14699 }, { "epoch": 1.8699910952804988, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.73773193359375, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8665497303009033, "num_tokens": 557918020.0, "step": 14700 }, { "epoch": 1.8701183055590893, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5901877880096436, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8680983185768127, "num_tokens": 557960633.0, "step": 14701 }, { "epoch": 1.8702455158376798, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7290453910827637, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8710987567901611, "num_tokens": 557998430.0, "step": 14702 }, { "epoch": 1.8703727261162701, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.58872389793396, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8626813888549805, "num_tokens": 558038283.0, "step": 14703 }, { "epoch": 1.8704999363948607, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.744121551513672, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8714941740036011, "num_tokens": 558079636.0, "step": 14704 }, { "epoch": 1.8706271466734512, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5743324756622314, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8809657096862793, "num_tokens": 558117670.0, "step": 14705 }, { "epoch": 1.8707543569520417, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7497355937957764, "learning_rate": 1e-05, "loss": 0.4906, "mean_token_accuracy": 0.8573044538497925, "num_tokens": 558155732.0, "step": 14706 }, { "epoch": 1.8708815672306323, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.574226140975952, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8687436580657959, "num_tokens": 558195389.0, "step": 14707 }, { "epoch": 1.8710087775092228, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7336301803588867, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8836098909378052, "num_tokens": 558239535.0, "step": 14708 }, { "epoch": 1.871135987787813, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.582756757736206, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8707771301269531, "num_tokens": 558277441.0, "step": 14709 }, { "epoch": 1.8712631980664036, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7384650707244873, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8682913780212402, "num_tokens": 558319331.0, "step": 14710 }, { "epoch": 1.8713904083449941, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.570279121398926, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8747573494911194, "num_tokens": 558365494.0, "step": 14711 }, { "epoch": 1.8715176186235847, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.8058929443359375, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8658893704414368, "num_tokens": 558407608.0, "step": 14712 }, { "epoch": 1.8716448289021752, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5950734615325928, "learning_rate": 1e-05, "loss": 0.5101, "mean_token_accuracy": 0.8552883863449097, "num_tokens": 558444848.0, "step": 14713 }, { "epoch": 1.8717720391807657, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7614681720733643, "learning_rate": 1e-05, "loss": 0.5103, "mean_token_accuracy": 0.8507773876190186, "num_tokens": 558486630.0, "step": 14714 }, { "epoch": 1.8718992494593563, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5608344078063965, "learning_rate": 1e-05, "loss": 0.3899, "mean_token_accuracy": 0.8825013637542725, "num_tokens": 558523609.0, "step": 14715 }, { "epoch": 1.8720264597379468, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7712066173553467, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.878538966178894, "num_tokens": 558561065.0, "step": 14716 }, { "epoch": 1.8721536700165373, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5600779056549072, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8793025016784668, "num_tokens": 558602007.0, "step": 14717 }, { "epoch": 1.8722808802951278, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.8010122776031494, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8714108467102051, "num_tokens": 558637712.0, "step": 14718 }, { "epoch": 1.8724080905737184, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5659122467041016, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8761485815048218, "num_tokens": 558671939.0, "step": 14719 }, { "epoch": 1.872535300852309, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8010001182556152, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8680973052978516, "num_tokens": 558708413.0, "step": 14720 }, { "epoch": 1.8726625111308994, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5781991481781006, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8453472852706909, "num_tokens": 558748993.0, "step": 14721 }, { "epoch": 1.87278972140949, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7962141036987305, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8667111396789551, "num_tokens": 558788058.0, "step": 14722 }, { "epoch": 1.8729169316880805, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5778725147247314, "learning_rate": 1e-05, "loss": 0.5144, "mean_token_accuracy": 0.8486440181732178, "num_tokens": 558828046.0, "step": 14723 }, { "epoch": 1.873044141966671, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7885842323303223, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8736675977706909, "num_tokens": 558862445.0, "step": 14724 }, { "epoch": 1.8731713522452615, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.561633348464966, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.881658136844635, "num_tokens": 558907645.0, "step": 14725 }, { "epoch": 1.873298562523852, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.798001289367676, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8744338750839233, "num_tokens": 558943253.0, "step": 14726 }, { "epoch": 1.8734257728024426, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5730338096618652, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8755397200584412, "num_tokens": 558980341.0, "step": 14727 }, { "epoch": 1.873552983081033, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8053786754608154, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8621498942375183, "num_tokens": 559018659.0, "step": 14728 }, { "epoch": 1.8736801933596234, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5660367012023926, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8740836381912231, "num_tokens": 559060254.0, "step": 14729 }, { "epoch": 1.873807403638214, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.812830686569214, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.874122142791748, "num_tokens": 559099166.0, "step": 14730 }, { "epoch": 1.8739346139168045, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5758159160614014, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8801791667938232, "num_tokens": 559135189.0, "step": 14731 }, { "epoch": 1.874061824195395, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7994134426116943, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.874299943447113, "num_tokens": 559178491.0, "step": 14732 }, { "epoch": 1.8741890344739855, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5926930904388428, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8639947175979614, "num_tokens": 559213977.0, "step": 14733 }, { "epoch": 1.8743162447525759, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7803831100463867, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8714767098426819, "num_tokens": 559254309.0, "step": 14734 }, { "epoch": 1.8744434550311664, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5840179920196533, "learning_rate": 1e-05, "loss": 0.3988, "mean_token_accuracy": 0.8842476606369019, "num_tokens": 559289756.0, "step": 14735 }, { "epoch": 1.874570665309757, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7750864028930664, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8725016117095947, "num_tokens": 559331143.0, "step": 14736 }, { "epoch": 1.8746978755883474, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.603423833847046, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8692063093185425, "num_tokens": 559368256.0, "step": 14737 }, { "epoch": 1.874825085866938, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7601284980773926, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8524079322814941, "num_tokens": 559408853.0, "step": 14738 }, { "epoch": 1.8749522961455285, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.604607343673706, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8628873229026794, "num_tokens": 559445606.0, "step": 14739 }, { "epoch": 1.875079506424119, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7503974437713623, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8735673427581787, "num_tokens": 559482482.0, "step": 14740 }, { "epoch": 1.8752067167027096, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5831246376037598, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8719041347503662, "num_tokens": 559522757.0, "step": 14741 }, { "epoch": 1.8753339269813, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.749358654022217, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8792753219604492, "num_tokens": 559556684.0, "step": 14742 }, { "epoch": 1.8754611372598906, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5899951457977295, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8706157803535461, "num_tokens": 559597548.0, "step": 14743 }, { "epoch": 1.8755883475384811, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7419707775115967, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8659493327140808, "num_tokens": 559638440.0, "step": 14744 }, { "epoch": 1.8757155578170717, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5858325958251953, "learning_rate": 1e-05, "loss": 0.5102, "mean_token_accuracy": 0.8497799038887024, "num_tokens": 559680819.0, "step": 14745 }, { "epoch": 1.8758427680956622, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7399959564208984, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8692693710327148, "num_tokens": 559721861.0, "step": 14746 }, { "epoch": 1.8759699783742527, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5898423194885254, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8703619837760925, "num_tokens": 559757370.0, "step": 14747 }, { "epoch": 1.8760971886528433, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7511041164398193, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.864077091217041, "num_tokens": 559795570.0, "step": 14748 }, { "epoch": 1.8762243989314338, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.570622444152832, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8753097653388977, "num_tokens": 559835115.0, "step": 14749 }, { "epoch": 1.8763516092100243, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.752562999725342, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8658175468444824, "num_tokens": 559876119.0, "step": 14750 }, { "epoch": 1.8764788194886148, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.591442584991455, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8755357265472412, "num_tokens": 559908349.0, "step": 14751 }, { "epoch": 1.8766060297672051, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.748410224914551, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8571693897247314, "num_tokens": 559947336.0, "step": 14752 }, { "epoch": 1.8767332400457957, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5851166248321533, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8624154329299927, "num_tokens": 559989571.0, "step": 14753 }, { "epoch": 1.8768604503243862, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7672057151794434, "learning_rate": 1e-05, "loss": 0.4873, "mean_token_accuracy": 0.8517783880233765, "num_tokens": 560022840.0, "step": 14754 }, { "epoch": 1.8769876606029767, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.000278472900390625, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.572293281555176, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8624497652053833, "num_tokens": 560061939.0, "step": 14755 }, { "epoch": 1.8771148708815673, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7707624435424805, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8679933547973633, "num_tokens": 560097480.0, "step": 14756 }, { "epoch": 1.8772420811601578, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.566845178604126, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8688470721244812, "num_tokens": 560137957.0, "step": 14757 }, { "epoch": 1.877369291438748, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7770066261291504, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8663510084152222, "num_tokens": 560173717.0, "step": 14758 }, { "epoch": 1.8774965017173386, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.573637008666992, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8690440058708191, "num_tokens": 560209833.0, "step": 14759 }, { "epoch": 1.8776237119959291, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.78537917137146, "learning_rate": 1e-05, "loss": 0.5302, "mean_token_accuracy": 0.845634937286377, "num_tokens": 560250406.0, "step": 14760 }, { "epoch": 1.8777509222745197, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5732948780059814, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8699570894241333, "num_tokens": 560288828.0, "step": 14761 }, { "epoch": 1.8778781325531102, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7800216674804688, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8795070648193359, "num_tokens": 560327431.0, "step": 14762 }, { "epoch": 1.8780053428317007, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.565129041671753, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8887163400650024, "num_tokens": 560367767.0, "step": 14763 }, { "epoch": 1.8781325531102913, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7981691360473633, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.882404088973999, "num_tokens": 560401179.0, "step": 14764 }, { "epoch": 1.8782597633888818, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.600961208343506, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.865506649017334, "num_tokens": 560431198.0, "step": 14765 }, { "epoch": 1.8783869736674723, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7766523361206055, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8704169988632202, "num_tokens": 560470244.0, "step": 14766 }, { "epoch": 1.8785141839460628, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5760486125946045, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8784183859825134, "num_tokens": 560511696.0, "step": 14767 }, { "epoch": 1.8786413942246534, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.769078254699707, "learning_rate": 1e-05, "loss": 0.3766, "mean_token_accuracy": 0.8940262198448181, "num_tokens": 560547555.0, "step": 14768 }, { "epoch": 1.878768604503244, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5828635692596436, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8692588806152344, "num_tokens": 560584349.0, "step": 14769 }, { "epoch": 1.8788958147818344, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.761749744415283, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8793042898178101, "num_tokens": 560628403.0, "step": 14770 }, { "epoch": 1.879023025060425, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5991294384002686, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8748891949653625, "num_tokens": 560663454.0, "step": 14771 }, { "epoch": 1.8791502353390155, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7531487941741943, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8690676689147949, "num_tokens": 560700506.0, "step": 14772 }, { "epoch": 1.879277445617606, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.585911750793457, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8765579462051392, "num_tokens": 560739690.0, "step": 14773 }, { "epoch": 1.8794046558961965, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7373125553131104, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8735868334770203, "num_tokens": 560786364.0, "step": 14774 }, { "epoch": 1.879531866174787, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5858471393585205, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8659732341766357, "num_tokens": 560828850.0, "step": 14775 }, { "epoch": 1.8796590764533776, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.756380319595337, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8790183067321777, "num_tokens": 560863401.0, "step": 14776 }, { "epoch": 1.879786286731968, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.651618719100952, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8705606460571289, "num_tokens": 560894347.0, "step": 14777 }, { "epoch": 1.8799134970105584, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.84512996673584, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8568998575210571, "num_tokens": 560935302.0, "step": 14778 }, { "epoch": 1.880040707289149, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.565232753753662, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8804680705070496, "num_tokens": 560971960.0, "step": 14779 }, { "epoch": 1.8801679175677395, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7702677249908447, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8645265102386475, "num_tokens": 561012267.0, "step": 14780 }, { "epoch": 1.88029512784633, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5627987384796143, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8706753849983215, "num_tokens": 561048861.0, "step": 14781 }, { "epoch": 1.8804223381249205, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7822892665863037, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.874146580696106, "num_tokens": 561087832.0, "step": 14782 }, { "epoch": 1.8805495484035109, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5775680541992188, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8801729679107666, "num_tokens": 561120657.0, "step": 14783 }, { "epoch": 1.8806767586821014, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.77107572555542, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8820002675056458, "num_tokens": 561161762.0, "step": 14784 }, { "epoch": 1.880803968960692, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.580688238143921, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8733018636703491, "num_tokens": 561199459.0, "step": 14785 }, { "epoch": 1.8809311792392824, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.787498950958252, "learning_rate": 1e-05, "loss": 0.4985, "mean_token_accuracy": 0.856348991394043, "num_tokens": 561237958.0, "step": 14786 }, { "epoch": 1.881058389517873, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.595933675765991, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8655529618263245, "num_tokens": 561272086.0, "step": 14787 }, { "epoch": 1.8811855997964635, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7631397247314453, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8791007995605469, "num_tokens": 561309821.0, "step": 14788 }, { "epoch": 1.881312810075054, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5819709300994873, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8693845868110657, "num_tokens": 561351565.0, "step": 14789 }, { "epoch": 1.8814400203536445, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7812628746032715, "learning_rate": 1e-05, "loss": 0.5104, "mean_token_accuracy": 0.8500431776046753, "num_tokens": 561387399.0, "step": 14790 }, { "epoch": 1.881567230632235, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.580805540084839, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8676150441169739, "num_tokens": 561427917.0, "step": 14791 }, { "epoch": 1.8816944409108256, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7765824794769287, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8662224411964417, "num_tokens": 561465398.0, "step": 14792 }, { "epoch": 1.8818216511894161, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5684709548950195, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8739122152328491, "num_tokens": 561510264.0, "step": 14793 }, { "epoch": 1.8819488614680067, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.770890235900879, "learning_rate": 1e-05, "loss": 0.4689, "mean_token_accuracy": 0.8655515909194946, "num_tokens": 561550942.0, "step": 14794 }, { "epoch": 1.8820760717465972, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5968282222747803, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8764946460723877, "num_tokens": 561580571.0, "step": 14795 }, { "epoch": 1.8822032820251877, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.758410692214966, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.87395840883255, "num_tokens": 561618993.0, "step": 14796 }, { "epoch": 1.8823304923037782, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5926756858825684, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.853935956954956, "num_tokens": 561659389.0, "step": 14797 }, { "epoch": 1.8824577025823688, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.746690511703491, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.872106671333313, "num_tokens": 561707703.0, "step": 14798 }, { "epoch": 1.8825849128609593, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5828540325164795, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8693803548812866, "num_tokens": 561749367.0, "step": 14799 }, { "epoch": 1.8827121231395498, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7650105953216553, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8618234992027283, "num_tokens": 561785234.0, "step": 14800 }, { "epoch": 1.8828393334181401, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.56795072555542, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8790037631988525, "num_tokens": 561825801.0, "step": 14801 }, { "epoch": 1.8829665436967307, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.766458034515381, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8627239465713501, "num_tokens": 561865030.0, "step": 14802 }, { "epoch": 1.8830937539753212, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5752575397491455, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8811347484588623, "num_tokens": 561899767.0, "step": 14803 }, { "epoch": 1.8832209642539117, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.780540943145752, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8643029928207397, "num_tokens": 561936794.0, "step": 14804 }, { "epoch": 1.8833481745325023, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.572093963623047, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8693214058876038, "num_tokens": 561971770.0, "step": 14805 }, { "epoch": 1.8834753848110928, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7645421028137207, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8665939569473267, "num_tokens": 562016400.0, "step": 14806 }, { "epoch": 1.883602595089683, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5762178897857666, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8782564401626587, "num_tokens": 562050732.0, "step": 14807 }, { "epoch": 1.8837298053682736, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7592427730560303, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8721901774406433, "num_tokens": 562093619.0, "step": 14808 }, { "epoch": 1.8838570156468641, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5797059535980225, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.871849775314331, "num_tokens": 562131434.0, "step": 14809 }, { "epoch": 1.8839842259254547, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7721896171569824, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8604414463043213, "num_tokens": 562169914.0, "step": 14810 }, { "epoch": 1.8841114362040452, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.583253860473633, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8747942447662354, "num_tokens": 562207190.0, "step": 14811 }, { "epoch": 1.8842386464826357, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7649688720703125, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8642504215240479, "num_tokens": 562246236.0, "step": 14812 }, { "epoch": 1.8843658567612263, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5773699283599854, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8815223574638367, "num_tokens": 562286875.0, "step": 14813 }, { "epoch": 1.8844930670398168, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7708516120910645, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8773437738418579, "num_tokens": 562325807.0, "step": 14814 }, { "epoch": 1.8846202773184073, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5822441577911377, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8733721971511841, "num_tokens": 562363663.0, "step": 14815 }, { "epoch": 1.8847474875969978, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.770897150039673, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8717079758644104, "num_tokens": 562399700.0, "step": 14816 }, { "epoch": 1.8848746978755884, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5935206413269043, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8623033761978149, "num_tokens": 562434395.0, "step": 14817 }, { "epoch": 1.885001908154179, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.771062135696411, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8547773361206055, "num_tokens": 562472140.0, "step": 14818 }, { "epoch": 1.8851291184327694, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5808608531951904, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.869174599647522, "num_tokens": 562510556.0, "step": 14819 }, { "epoch": 1.88525632871136, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7653279304504395, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8671494722366333, "num_tokens": 562546679.0, "step": 14820 }, { "epoch": 1.8853835389899505, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.596220016479492, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8670370578765869, "num_tokens": 562580707.0, "step": 14821 }, { "epoch": 1.885510749268541, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7454841136932373, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.869576096534729, "num_tokens": 562621753.0, "step": 14822 }, { "epoch": 1.8856379595471315, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5980448722839355, "learning_rate": 1e-05, "loss": 0.3804, "mean_token_accuracy": 0.89068603515625, "num_tokens": 562654064.0, "step": 14823 }, { "epoch": 1.885765169825722, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.749767303466797, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8608328700065613, "num_tokens": 562692575.0, "step": 14824 }, { "epoch": 1.8858923801043126, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5848655700683594, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8817273378372192, "num_tokens": 562732655.0, "step": 14825 }, { "epoch": 1.886019590382903, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.754641056060791, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.868308424949646, "num_tokens": 562768618.0, "step": 14826 }, { "epoch": 1.8861468006614934, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5813047885894775, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.857108473777771, "num_tokens": 562805357.0, "step": 14827 }, { "epoch": 1.886274010940084, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7514240741729736, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8649199604988098, "num_tokens": 562848798.0, "step": 14828 }, { "epoch": 1.8864012212186745, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.573514699935913, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8783366084098816, "num_tokens": 562890108.0, "step": 14829 }, { "epoch": 1.886528431497265, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7745792865753174, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8761599659919739, "num_tokens": 562921089.0, "step": 14830 }, { "epoch": 1.8866556417758555, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5893845558166504, "learning_rate": 1e-05, "loss": 0.5395, "mean_token_accuracy": 0.8371396064758301, "num_tokens": 562961705.0, "step": 14831 }, { "epoch": 1.8867828520544458, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7495532035827637, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8867887854576111, "num_tokens": 563000011.0, "step": 14832 }, { "epoch": 1.8869100623330364, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5680043697357178, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8745669722557068, "num_tokens": 563042568.0, "step": 14833 }, { "epoch": 1.887037272611627, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.768389940261841, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8777503967285156, "num_tokens": 563075892.0, "step": 14834 }, { "epoch": 1.8871644828902174, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5643458366394043, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.86930251121521, "num_tokens": 563117276.0, "step": 14835 }, { "epoch": 1.887291693168808, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7892355918884277, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8721418380737305, "num_tokens": 563152557.0, "step": 14836 }, { "epoch": 1.8874189034473985, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5757038593292236, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8732632398605347, "num_tokens": 563188620.0, "step": 14837 }, { "epoch": 1.887546113725989, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7844438552856445, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.865971028804779, "num_tokens": 563227582.0, "step": 14838 }, { "epoch": 1.8876733240045795, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5735368728637695, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8836287260055542, "num_tokens": 563260057.0, "step": 14839 }, { "epoch": 1.88780053428317, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.780322313308716, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8694347739219666, "num_tokens": 563299612.0, "step": 14840 }, { "epoch": 1.8879277445617606, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.606171131134033, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8729877471923828, "num_tokens": 563331659.0, "step": 14841 }, { "epoch": 1.8880549548403511, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7648580074310303, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8631375432014465, "num_tokens": 563371411.0, "step": 14842 }, { "epoch": 1.8881821651189417, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6004717350006104, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8700982332229614, "num_tokens": 563401498.0, "step": 14843 }, { "epoch": 1.8883093753975322, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7310523986816406, "learning_rate": 1e-05, "loss": 0.3927, "mean_token_accuracy": 0.8863459825515747, "num_tokens": 563441383.0, "step": 14844 }, { "epoch": 1.8884365856761227, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6096620559692383, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8605897426605225, "num_tokens": 563478547.0, "step": 14845 }, { "epoch": 1.8885637959547132, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.708991050720215, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.870002031326294, "num_tokens": 563521366.0, "step": 14846 }, { "epoch": 1.8886910062333038, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.611816167831421, "learning_rate": 1e-05, "loss": 0.489, "mean_token_accuracy": 0.8538599014282227, "num_tokens": 563560062.0, "step": 14847 }, { "epoch": 1.8888182165118943, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7134451866149902, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8636700510978699, "num_tokens": 563598448.0, "step": 14848 }, { "epoch": 1.8889454267904848, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5936453342437744, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8785330057144165, "num_tokens": 563639274.0, "step": 14849 }, { "epoch": 1.8890726370690751, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.712918758392334, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8688110113143921, "num_tokens": 563679309.0, "step": 14850 }, { "epoch": 1.8891998473476657, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.594144105911255, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8696774244308472, "num_tokens": 563717594.0, "step": 14851 }, { "epoch": 1.8893270576262562, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7256252765655518, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8807759881019592, "num_tokens": 563749442.0, "step": 14852 }, { "epoch": 1.8894542679048467, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5926172733306885, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.863919198513031, "num_tokens": 563788519.0, "step": 14853 }, { "epoch": 1.8895814781834372, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7256762981414795, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8615559339523315, "num_tokens": 563827233.0, "step": 14854 }, { "epoch": 1.8897086884620278, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5899837017059326, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8697437644004822, "num_tokens": 563863534.0, "step": 14855 }, { "epoch": 1.889835898740618, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.730381965637207, "learning_rate": 1e-05, "loss": 0.4986, "mean_token_accuracy": 0.8532716631889343, "num_tokens": 563903237.0, "step": 14856 }, { "epoch": 1.8899631090192086, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.567009210586548, "learning_rate": 1e-05, "loss": 0.3648, "mean_token_accuracy": 0.893579363822937, "num_tokens": 563944334.0, "step": 14857 }, { "epoch": 1.8900903192977991, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7476539611816406, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8668417930603027, "num_tokens": 563979801.0, "step": 14858 }, { "epoch": 1.8902175295763897, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.565948486328125, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8641015291213989, "num_tokens": 564022348.0, "step": 14859 }, { "epoch": 1.8903447398549802, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7539219856262207, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.8830548524856567, "num_tokens": 564063285.0, "step": 14860 }, { "epoch": 1.8904719501335707, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.563382387161255, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8708254098892212, "num_tokens": 564105323.0, "step": 14861 }, { "epoch": 1.8905991604121613, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7784955501556396, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.86181640625, "num_tokens": 564140238.0, "step": 14862 }, { "epoch": 1.8907263706907518, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5743870735168457, "learning_rate": 1e-05, "loss": 0.5244, "mean_token_accuracy": 0.8425223231315613, "num_tokens": 564178441.0, "step": 14863 }, { "epoch": 1.8908535809693423, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.768326759338379, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8715379238128662, "num_tokens": 564217586.0, "step": 14864 }, { "epoch": 1.8909807912479328, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.566967248916626, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8606059551239014, "num_tokens": 564259198.0, "step": 14865 }, { "epoch": 1.8911080015265234, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.766113758087158, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.8843425512313843, "num_tokens": 564298622.0, "step": 14866 }, { "epoch": 1.891235211805114, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.575291872024536, "learning_rate": 1e-05, "loss": 0.4021, "mean_token_accuracy": 0.8763673305511475, "num_tokens": 564331573.0, "step": 14867 }, { "epoch": 1.8913624220837044, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.77286958694458, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8640774488449097, "num_tokens": 564371461.0, "step": 14868 }, { "epoch": 1.891489632362295, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5583603382110596, "learning_rate": 1e-05, "loss": 0.3804, "mean_token_accuracy": 0.8887460231781006, "num_tokens": 564410160.0, "step": 14869 }, { "epoch": 1.8916168426408855, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.781785011291504, "learning_rate": 1e-05, "loss": 0.4797, "mean_token_accuracy": 0.8590624928474426, "num_tokens": 564455954.0, "step": 14870 }, { "epoch": 1.891744052919476, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5745930671691895, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8715901374816895, "num_tokens": 564490196.0, "step": 14871 }, { "epoch": 1.8918712631980665, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7721636295318604, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8760154843330383, "num_tokens": 564530362.0, "step": 14872 }, { "epoch": 1.891998473476657, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.584709644317627, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8671432733535767, "num_tokens": 564563528.0, "step": 14873 }, { "epoch": 1.8921256837552476, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7819883823394775, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8724099397659302, "num_tokens": 564596865.0, "step": 14874 }, { "epoch": 1.892252894033838, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5833890438079834, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8643067479133606, "num_tokens": 564635740.0, "step": 14875 }, { "epoch": 1.8923801043124284, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.772782325744629, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8602889776229858, "num_tokens": 564679007.0, "step": 14876 }, { "epoch": 1.892507314591019, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5724966526031494, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.875251054763794, "num_tokens": 564719211.0, "step": 14877 }, { "epoch": 1.8926345248696095, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7749428749084473, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8852481842041016, "num_tokens": 564756482.0, "step": 14878 }, { "epoch": 1.8927617351482, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5761396884918213, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8718870282173157, "num_tokens": 564796304.0, "step": 14879 }, { "epoch": 1.8928889454267905, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7748594284057617, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8677842020988464, "num_tokens": 564835444.0, "step": 14880 }, { "epoch": 1.8930161557053808, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5893681049346924, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8562989234924316, "num_tokens": 564873637.0, "step": 14881 }, { "epoch": 1.8931433659839714, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7567005157470703, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.873013973236084, "num_tokens": 564914742.0, "step": 14882 }, { "epoch": 1.893270576262562, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5831007957458496, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8720675706863403, "num_tokens": 564953519.0, "step": 14883 }, { "epoch": 1.8933977865411524, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7801692485809326, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8638122081756592, "num_tokens": 564984994.0, "step": 14884 }, { "epoch": 1.893524996819743, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5769827365875244, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8596833944320679, "num_tokens": 565028584.0, "step": 14885 }, { "epoch": 1.8936522070983335, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7705912590026855, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8695257902145386, "num_tokens": 565068726.0, "step": 14886 }, { "epoch": 1.893779417376924, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5799977779388428, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8621014952659607, "num_tokens": 565112156.0, "step": 14887 }, { "epoch": 1.8939066276555145, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.780245542526245, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.874375581741333, "num_tokens": 565146158.0, "step": 14888 }, { "epoch": 1.894033837934105, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5819802284240723, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.882970929145813, "num_tokens": 565176571.0, "step": 14889 }, { "epoch": 1.8941610482126956, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.774505138397217, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8680561780929565, "num_tokens": 565216162.0, "step": 14890 }, { "epoch": 1.8942882584912861, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5752949714660645, "learning_rate": 1e-05, "loss": 0.3946, "mean_token_accuracy": 0.8841248750686646, "num_tokens": 565252010.0, "step": 14891 }, { "epoch": 1.8944154687698767, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7740318775177, "learning_rate": 1e-05, "loss": 0.4895, "mean_token_accuracy": 0.8540700674057007, "num_tokens": 565297159.0, "step": 14892 }, { "epoch": 1.8945426790484672, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.580857753753662, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8740226030349731, "num_tokens": 565337074.0, "step": 14893 }, { "epoch": 1.8946698893270577, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.766418695449829, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8635938763618469, "num_tokens": 565379551.0, "step": 14894 }, { "epoch": 1.8947970996056482, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.58170485496521, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8745355010032654, "num_tokens": 565418611.0, "step": 14895 }, { "epoch": 1.8949243098842388, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.779752016067505, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8730703592300415, "num_tokens": 565452804.0, "step": 14896 }, { "epoch": 1.8950515201628293, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.579362630844116, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8662084341049194, "num_tokens": 565490628.0, "step": 14897 }, { "epoch": 1.8951787304414198, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7915210723876953, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8636929988861084, "num_tokens": 565523990.0, "step": 14898 }, { "epoch": 1.8953059407200101, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.572404384613037, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8780326843261719, "num_tokens": 565562684.0, "step": 14899 }, { "epoch": 1.8954331509986007, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7787888050079346, "learning_rate": 1e-05, "loss": 0.3756, "mean_token_accuracy": 0.8920828104019165, "num_tokens": 565596479.0, "step": 14900 }, { "epoch": 1.8955603612771912, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.581209182739258, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8677623271942139, "num_tokens": 565639112.0, "step": 14901 }, { "epoch": 1.8956875715557817, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7866971492767334, "learning_rate": 1e-05, "loss": 0.4156, "mean_token_accuracy": 0.8789371848106384, "num_tokens": 565671379.0, "step": 14902 }, { "epoch": 1.8958147818343722, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5784287452697754, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.877179741859436, "num_tokens": 565708115.0, "step": 14903 }, { "epoch": 1.8959419921129628, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7948904037475586, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8644158840179443, "num_tokens": 565744767.0, "step": 14904 }, { "epoch": 1.896069202391553, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.575653076171875, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.870276927947998, "num_tokens": 565783709.0, "step": 14905 }, { "epoch": 1.8961964126701436, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.79538631439209, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.874904453754425, "num_tokens": 565818503.0, "step": 14906 }, { "epoch": 1.8963236229487341, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5853941440582275, "learning_rate": 1e-05, "loss": 0.3822, "mean_token_accuracy": 0.8898969888687134, "num_tokens": 565850188.0, "step": 14907 }, { "epoch": 1.8964508332273247, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.784274101257324, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8633129000663757, "num_tokens": 565889177.0, "step": 14908 }, { "epoch": 1.8965780435059152, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.604468822479248, "learning_rate": 1e-05, "loss": 0.5126, "mean_token_accuracy": 0.8500732779502869, "num_tokens": 565929316.0, "step": 14909 }, { "epoch": 1.8967052537845057, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.777073383331299, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8882648348808289, "num_tokens": 565961874.0, "step": 14910 }, { "epoch": 1.8968324640630962, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.610395669937134, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8552977442741394, "num_tokens": 565995143.0, "step": 14911 }, { "epoch": 1.8969596743416868, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7592947483062744, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8707075119018555, "num_tokens": 566032297.0, "step": 14912 }, { "epoch": 1.8970868846202773, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.590334177017212, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8747454881668091, "num_tokens": 566073138.0, "step": 14913 }, { "epoch": 1.8972140948988678, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.744368076324463, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8743466138839722, "num_tokens": 566115471.0, "step": 14914 }, { "epoch": 1.8973413051774584, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.596593141555786, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8672984838485718, "num_tokens": 566152644.0, "step": 14915 }, { "epoch": 1.8974685154560489, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.727208375930786, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8686774373054504, "num_tokens": 566197411.0, "step": 14916 }, { "epoch": 1.8975957257346394, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.580226421356201, "learning_rate": 1e-05, "loss": 0.3633, "mean_token_accuracy": 0.8940569162368774, "num_tokens": 566234738.0, "step": 14917 }, { "epoch": 1.89772293601323, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.74355149269104, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8536840677261353, "num_tokens": 566274445.0, "step": 14918 }, { "epoch": 1.8978501462918205, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5777008533477783, "learning_rate": 1e-05, "loss": 0.3874, "mean_token_accuracy": 0.8895906209945679, "num_tokens": 566312710.0, "step": 14919 }, { "epoch": 1.897977356570411, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.749274730682373, "learning_rate": 1e-05, "loss": 0.4625, "mean_token_accuracy": 0.8683120012283325, "num_tokens": 566349229.0, "step": 14920 }, { "epoch": 1.8981045668490015, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5906810760498047, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8643768429756165, "num_tokens": 566387208.0, "step": 14921 }, { "epoch": 1.898231777127592, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.745958089828491, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8707523941993713, "num_tokens": 566423616.0, "step": 14922 }, { "epoch": 1.8983589874061826, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.582230806350708, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8743621110916138, "num_tokens": 566457984.0, "step": 14923 }, { "epoch": 1.898486197684773, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.752220630645752, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8658350110054016, "num_tokens": 566495664.0, "step": 14924 }, { "epoch": 1.8986134079633634, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.577502727508545, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.863998532295227, "num_tokens": 566533750.0, "step": 14925 }, { "epoch": 1.898740618241954, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.747121810913086, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8617299795150757, "num_tokens": 566576247.0, "step": 14926 }, { "epoch": 1.8988678285205445, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.580242872238159, "learning_rate": 1e-05, "loss": 0.4949, "mean_token_accuracy": 0.8547629117965698, "num_tokens": 566618185.0, "step": 14927 }, { "epoch": 1.898995038799135, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.751668930053711, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8578135967254639, "num_tokens": 566659219.0, "step": 14928 }, { "epoch": 1.8991222490777255, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.57938814163208, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8780256509780884, "num_tokens": 566695101.0, "step": 14929 }, { "epoch": 1.8992494593563158, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7520508766174316, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8683127164840698, "num_tokens": 566735286.0, "step": 14930 }, { "epoch": 1.8993766696349064, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5761594772338867, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8717144727706909, "num_tokens": 566773534.0, "step": 14931 }, { "epoch": 1.899503879913497, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7489209175109863, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8762679696083069, "num_tokens": 566819141.0, "step": 14932 }, { "epoch": 1.8996310901920874, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5768258571624756, "learning_rate": 1e-05, "loss": 0.4004, "mean_token_accuracy": 0.886681854724884, "num_tokens": 566855095.0, "step": 14933 }, { "epoch": 1.899758300470678, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7692301273345947, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8634663224220276, "num_tokens": 566888989.0, "step": 14934 }, { "epoch": 1.8998855107492685, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.578317642211914, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8827930092811584, "num_tokens": 566923049.0, "step": 14935 }, { "epoch": 1.900012721027859, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7703278064727783, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.8508756160736084, "num_tokens": 566962049.0, "step": 14936 }, { "epoch": 1.9001399313064495, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.562913417816162, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8856914043426514, "num_tokens": 567003538.0, "step": 14937 }, { "epoch": 1.90026714158504, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7792208194732666, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8812479972839355, "num_tokens": 567035821.0, "step": 14938 }, { "epoch": 1.9003943518636306, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.579909324645996, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8792847394943237, "num_tokens": 567068095.0, "step": 14939 }, { "epoch": 1.9005215621422211, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.780942678451538, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8748083114624023, "num_tokens": 567104741.0, "step": 14940 }, { "epoch": 1.9006487724208116, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5767476558685303, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8683730363845825, "num_tokens": 567142500.0, "step": 14941 }, { "epoch": 1.9007759826994022, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7785115242004395, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8701344728469849, "num_tokens": 567182001.0, "step": 14942 }, { "epoch": 1.9009031929779927, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.610018014907837, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8673309087753296, "num_tokens": 567219809.0, "step": 14943 }, { "epoch": 1.9010304032565832, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7732412815093994, "learning_rate": 1e-05, "loss": 0.5316, "mean_token_accuracy": 0.8433922529220581, "num_tokens": 567260309.0, "step": 14944 }, { "epoch": 1.9011576135351738, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.594271421432495, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8581632375717163, "num_tokens": 567295840.0, "step": 14945 }, { "epoch": 1.9012848238137643, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.758556842803955, "learning_rate": 1e-05, "loss": 0.5091, "mean_token_accuracy": 0.8495489358901978, "num_tokens": 567335546.0, "step": 14946 }, { "epoch": 1.9014120340923548, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5774593353271484, "learning_rate": 1e-05, "loss": 0.3964, "mean_token_accuracy": 0.8875190019607544, "num_tokens": 567372996.0, "step": 14947 }, { "epoch": 1.9015392443709451, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.75856876373291, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8782578706741333, "num_tokens": 567411758.0, "step": 14948 }, { "epoch": 1.9016664546495357, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.591770887374878, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8682302236557007, "num_tokens": 567446859.0, "step": 14949 }, { "epoch": 1.9017936649281262, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7621607780456543, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8806395530700684, "num_tokens": 567480890.0, "step": 14950 }, { "epoch": 1.9019208752067167, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.567777156829834, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8765340447425842, "num_tokens": 567523381.0, "step": 14951 }, { "epoch": 1.9020480854853072, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.774989604949951, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8695579767227173, "num_tokens": 567564926.0, "step": 14952 }, { "epoch": 1.9021752957638978, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.570660352706909, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8687365651130676, "num_tokens": 567603351.0, "step": 14953 }, { "epoch": 1.902302506042488, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7835307121276855, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8857107162475586, "num_tokens": 567636710.0, "step": 14954 }, { "epoch": 1.9024297163210786, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.560443639755249, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8834453821182251, "num_tokens": 567676325.0, "step": 14955 }, { "epoch": 1.9025569265996691, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7763593196868896, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8907423615455627, "num_tokens": 567720663.0, "step": 14956 }, { "epoch": 1.9026841368782597, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5863659381866455, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8625986576080322, "num_tokens": 567761589.0, "step": 14957 }, { "epoch": 1.9028113471568502, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7728965282440186, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8817911744117737, "num_tokens": 567801649.0, "step": 14958 }, { "epoch": 1.9029385574354407, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5771055221557617, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8803184032440186, "num_tokens": 567839953.0, "step": 14959 }, { "epoch": 1.9030657677140312, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7678542137145996, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8787442445755005, "num_tokens": 567877634.0, "step": 14960 }, { "epoch": 1.9031929779926218, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6059107780456543, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8584762811660767, "num_tokens": 567911504.0, "step": 14961 }, { "epoch": 1.9033201882712123, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7911534309387207, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8757541179656982, "num_tokens": 567954904.0, "step": 14962 }, { "epoch": 1.9034473985498028, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.584139823913574, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8649312853813171, "num_tokens": 567992384.0, "step": 14963 }, { "epoch": 1.9035746088283934, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.779280662536621, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8678540587425232, "num_tokens": 568027692.0, "step": 14964 }, { "epoch": 1.9037018191069839, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5799663066864014, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8646983504295349, "num_tokens": 568063385.0, "step": 14965 }, { "epoch": 1.9038290293855744, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7765450477600098, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8808107972145081, "num_tokens": 568098315.0, "step": 14966 }, { "epoch": 1.903956239664165, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5732405185699463, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8616530895233154, "num_tokens": 568144032.0, "step": 14967 }, { "epoch": 1.9040834499427555, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7939436435699463, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8687076568603516, "num_tokens": 568174988.0, "step": 14968 }, { "epoch": 1.904210660221346, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5692861080169678, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8587660789489746, "num_tokens": 568219794.0, "step": 14969 }, { "epoch": 1.9043378704999365, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.785017967224121, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.877555251121521, "num_tokens": 568264995.0, "step": 14970 }, { "epoch": 1.904465080778527, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.586930990219116, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8738595247268677, "num_tokens": 568300876.0, "step": 14971 }, { "epoch": 1.9045922910571176, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7812726497650146, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.867046594619751, "num_tokens": 568337341.0, "step": 14972 }, { "epoch": 1.9047195013357079, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.589735984802246, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8666819334030151, "num_tokens": 568373524.0, "step": 14973 }, { "epoch": 1.9048467116142984, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.76187801361084, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8843264579772949, "num_tokens": 568412575.0, "step": 14974 }, { "epoch": 1.904973921892889, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.593425750732422, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8741729259490967, "num_tokens": 568446794.0, "step": 14975 }, { "epoch": 1.9051011321714795, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7738468647003174, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8675926327705383, "num_tokens": 568490645.0, "step": 14976 }, { "epoch": 1.90522834245007, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5833325386047363, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8715801239013672, "num_tokens": 568529281.0, "step": 14977 }, { "epoch": 1.9053555527286605, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7670247554779053, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8800243139266968, "num_tokens": 568568722.0, "step": 14978 }, { "epoch": 1.9054827630072508, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5918688774108887, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8750003576278687, "num_tokens": 568604736.0, "step": 14979 }, { "epoch": 1.9056099732858414, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7541513442993164, "learning_rate": 1e-05, "loss": 0.4033, "mean_token_accuracy": 0.8850116729736328, "num_tokens": 568643720.0, "step": 14980 }, { "epoch": 1.905737183564432, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.584540843963623, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8824728727340698, "num_tokens": 568681805.0, "step": 14981 }, { "epoch": 1.9058643938430224, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7535085678100586, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8659554719924927, "num_tokens": 568723107.0, "step": 14982 }, { "epoch": 1.905991604121613, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.603062629699707, "learning_rate": 1e-05, "loss": 0.5046, "mean_token_accuracy": 0.8527106046676636, "num_tokens": 568760229.0, "step": 14983 }, { "epoch": 1.9061188144002035, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7567386627197266, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8592414855957031, "num_tokens": 568794539.0, "step": 14984 }, { "epoch": 1.906246024678794, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5896952152252197, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8653707504272461, "num_tokens": 568834808.0, "step": 14985 }, { "epoch": 1.9063732349573845, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7397139072418213, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8709821701049805, "num_tokens": 568875004.0, "step": 14986 }, { "epoch": 1.906500445235975, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5768744945526123, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8740496039390564, "num_tokens": 568917372.0, "step": 14987 }, { "epoch": 1.9066276555145656, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.751722574234009, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8645588755607605, "num_tokens": 568953567.0, "step": 14988 }, { "epoch": 1.9067548657931561, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5964598655700684, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8617468476295471, "num_tokens": 568988634.0, "step": 14989 }, { "epoch": 1.9068820760717466, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.737431049346924, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8762269616127014, "num_tokens": 569023929.0, "step": 14990 }, { "epoch": 1.9070092863503372, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5896668434143066, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8793960213661194, "num_tokens": 569054483.0, "step": 14991 }, { "epoch": 1.9071364966289277, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7345101833343506, "learning_rate": 1e-05, "loss": 0.4765, "mean_token_accuracy": 0.8610045313835144, "num_tokens": 569096919.0, "step": 14992 }, { "epoch": 1.9072637069075182, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5914268493652344, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8664519190788269, "num_tokens": 569134161.0, "step": 14993 }, { "epoch": 1.9073909171861088, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.735344409942627, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.871300995349884, "num_tokens": 569173391.0, "step": 14994 }, { "epoch": 1.9075181274646993, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.605459690093994, "learning_rate": 1e-05, "loss": 0.5059, "mean_token_accuracy": 0.8527078628540039, "num_tokens": 569206173.0, "step": 14995 }, { "epoch": 1.9076453377432898, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7329678535461426, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8635560274124146, "num_tokens": 569241692.0, "step": 14996 }, { "epoch": 1.9077725480218801, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.596222162246704, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8802382349967957, "num_tokens": 569276361.0, "step": 14997 }, { "epoch": 1.9078997583004706, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.725881576538086, "learning_rate": 1e-05, "loss": 0.5199, "mean_token_accuracy": 0.8462965488433838, "num_tokens": 569314478.0, "step": 14998 }, { "epoch": 1.9080269685790612, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.582601308822632, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8743021488189697, "num_tokens": 569354867.0, "step": 14999 }, { "epoch": 1.9081541788576517, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7318766117095947, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8786495327949524, "num_tokens": 569394031.0, "step": 15000 }, { "epoch": 1.9082813891362422, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.584219217300415, "learning_rate": 1e-05, "loss": 0.3619, "mean_token_accuracy": 0.8972117900848389, "num_tokens": 569430899.0, "step": 15001 }, { "epoch": 1.9084085994148328, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7389187812805176, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8515926599502563, "num_tokens": 569469871.0, "step": 15002 }, { "epoch": 1.908535809693423, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5812265872955322, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8721537590026855, "num_tokens": 569508385.0, "step": 15003 }, { "epoch": 1.9086630199720136, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.731919527053833, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8722636699676514, "num_tokens": 569549705.0, "step": 15004 }, { "epoch": 1.9087902302506041, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5914270877838135, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8597329258918762, "num_tokens": 569584092.0, "step": 15005 }, { "epoch": 1.9089174405291947, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.736976385116577, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8712337017059326, "num_tokens": 569619683.0, "step": 15006 }, { "epoch": 1.9090446508077852, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.589007616043091, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8793036937713623, "num_tokens": 569656347.0, "step": 15007 }, { "epoch": 1.9091718610863757, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7349958419799805, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.87668776512146, "num_tokens": 569689633.0, "step": 15008 }, { "epoch": 1.9092990713649662, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.574337959289551, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8586621284484863, "num_tokens": 569730685.0, "step": 15009 }, { "epoch": 1.9094262816435568, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.740079402923584, "learning_rate": 1e-05, "loss": 0.4863, "mean_token_accuracy": 0.8592844009399414, "num_tokens": 569773343.0, "step": 15010 }, { "epoch": 1.9095534919221473, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.578085422515869, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8698948621749878, "num_tokens": 569814417.0, "step": 15011 }, { "epoch": 1.9096807022007378, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7393765449523926, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8771102428436279, "num_tokens": 569854436.0, "step": 15012 }, { "epoch": 1.9098079124793284, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.581092596054077, "learning_rate": 1e-05, "loss": 0.4913, "mean_token_accuracy": 0.8552875518798828, "num_tokens": 569897530.0, "step": 15013 }, { "epoch": 1.9099351227579189, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.758577346801758, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8757613897323608, "num_tokens": 569931927.0, "step": 15014 }, { "epoch": 1.9100623330365094, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5747976303100586, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8736766576766968, "num_tokens": 569968283.0, "step": 15015 }, { "epoch": 1.9101895433151, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.745363235473633, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8676849007606506, "num_tokens": 570009073.0, "step": 15016 }, { "epoch": 1.9103167535936905, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.581486940383911, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8700763583183289, "num_tokens": 570046421.0, "step": 15017 }, { "epoch": 1.910443963872281, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.740535020828247, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8659915924072266, "num_tokens": 570093679.0, "step": 15018 }, { "epoch": 1.9105711741508715, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.572763442993164, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8789881467819214, "num_tokens": 570132575.0, "step": 15019 }, { "epoch": 1.910698384429462, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.760716438293457, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8672121167182922, "num_tokens": 570169782.0, "step": 15020 }, { "epoch": 1.9108255947080524, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5822043418884277, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8751699328422546, "num_tokens": 570205034.0, "step": 15021 }, { "epoch": 1.9109528049866429, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.766549587249756, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8718565106391907, "num_tokens": 570236804.0, "step": 15022 }, { "epoch": 1.9110800152652334, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.585371494293213, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8721791505813599, "num_tokens": 570266893.0, "step": 15023 }, { "epoch": 1.911207225543824, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.765490770339966, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8807076811790466, "num_tokens": 570297779.0, "step": 15024 }, { "epoch": 1.9113344358224145, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5667717456817627, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8759324550628662, "num_tokens": 570334774.0, "step": 15025 }, { "epoch": 1.911461646101005, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7703137397766113, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8655675649642944, "num_tokens": 570373788.0, "step": 15026 }, { "epoch": 1.9115888563795955, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5707743167877197, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.876014769077301, "num_tokens": 570412156.0, "step": 15027 }, { "epoch": 1.9117160666581858, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.78912091255188, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8767232894897461, "num_tokens": 570442702.0, "step": 15028 }, { "epoch": 1.9118432769367764, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5646555423736572, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8585202693939209, "num_tokens": 570485302.0, "step": 15029 }, { "epoch": 1.9119704872153669, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7928245067596436, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8752085566520691, "num_tokens": 570524820.0, "step": 15030 }, { "epoch": 1.9120976974939574, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.556809663772583, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8587542176246643, "num_tokens": 570567973.0, "step": 15031 }, { "epoch": 1.912224907772548, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8034920692443848, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8754451274871826, "num_tokens": 570610369.0, "step": 15032 }, { "epoch": 1.9123521180511385, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5612101554870605, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8661966323852539, "num_tokens": 570653226.0, "step": 15033 }, { "epoch": 1.912479328329729, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8066272735595703, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8676385879516602, "num_tokens": 570691505.0, "step": 15034 }, { "epoch": 1.9126065386083195, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5646443367004395, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8628849983215332, "num_tokens": 570733781.0, "step": 15035 }, { "epoch": 1.91273374888691, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8110899925231934, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8752431273460388, "num_tokens": 570772675.0, "step": 15036 }, { "epoch": 1.9128609591655006, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.566340208053589, "learning_rate": 1e-05, "loss": 0.3946, "mean_token_accuracy": 0.8848361372947693, "num_tokens": 570813967.0, "step": 15037 }, { "epoch": 1.9129881694440911, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.814398765563965, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.8803449869155884, "num_tokens": 570850315.0, "step": 15038 }, { "epoch": 1.9131153797226816, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.588209390640259, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8643521666526794, "num_tokens": 570890191.0, "step": 15039 }, { "epoch": 1.9132425900012722, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.803337812423706, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8745579719543457, "num_tokens": 570926564.0, "step": 15040 }, { "epoch": 1.9133698002798627, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.588942050933838, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8738745450973511, "num_tokens": 570966334.0, "step": 15041 }, { "epoch": 1.9134970105584532, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.795032501220703, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8657521605491638, "num_tokens": 571008008.0, "step": 15042 }, { "epoch": 1.9136242208370438, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.601390838623047, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8604668974876404, "num_tokens": 571046915.0, "step": 15043 }, { "epoch": 1.9137514311156343, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7833645343780518, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8711137771606445, "num_tokens": 571085945.0, "step": 15044 }, { "epoch": 1.9138786413942248, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5962584018707275, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.886078953742981, "num_tokens": 571119839.0, "step": 15045 }, { "epoch": 1.9140058516728151, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7661514282226562, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.882020115852356, "num_tokens": 571156888.0, "step": 15046 }, { "epoch": 1.9141330619514056, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.584892988204956, "learning_rate": 1e-05, "loss": 0.3834, "mean_token_accuracy": 0.8903288841247559, "num_tokens": 571197022.0, "step": 15047 }, { "epoch": 1.9142602722299962, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7702362537384033, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8640025854110718, "num_tokens": 571234780.0, "step": 15048 }, { "epoch": 1.9143874825085867, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.596773147583008, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8746925592422485, "num_tokens": 571272804.0, "step": 15049 }, { "epoch": 1.9145146927871772, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7550852298736572, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8687624931335449, "num_tokens": 571315455.0, "step": 15050 }, { "epoch": 1.9146419030657678, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.595930337905884, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8776806592941284, "num_tokens": 571350522.0, "step": 15051 }, { "epoch": 1.914769113344358, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7549920082092285, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8774949312210083, "num_tokens": 571384452.0, "step": 15052 }, { "epoch": 1.9148963236229486, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5783886909484863, "learning_rate": 1e-05, "loss": 0.3928, "mean_token_accuracy": 0.8885534405708313, "num_tokens": 571423041.0, "step": 15053 }, { "epoch": 1.9150235339015391, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.757258176803589, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.8862369060516357, "num_tokens": 571454744.0, "step": 15054 }, { "epoch": 1.9151507441801296, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5838587284088135, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.881645917892456, "num_tokens": 571492937.0, "step": 15055 }, { "epoch": 1.9152779544587202, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7439887523651123, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8848928213119507, "num_tokens": 571531756.0, "step": 15056 }, { "epoch": 1.9154051647373107, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5937321186065674, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8619309663772583, "num_tokens": 571566095.0, "step": 15057 }, { "epoch": 1.9155323750159012, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.745978832244873, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8752808570861816, "num_tokens": 571599169.0, "step": 15058 }, { "epoch": 1.9156595852944918, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.595261812210083, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8800677061080933, "num_tokens": 571632063.0, "step": 15059 }, { "epoch": 1.9157867955730823, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7282257080078125, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8743976354598999, "num_tokens": 571670867.0, "step": 15060 }, { "epoch": 1.9159140058516728, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5919783115386963, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8709890246391296, "num_tokens": 571710445.0, "step": 15061 }, { "epoch": 1.9160412161302633, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7194671630859375, "learning_rate": 1e-05, "loss": 0.3845, "mean_token_accuracy": 0.8897374868392944, "num_tokens": 571748410.0, "step": 15062 }, { "epoch": 1.9161684264088539, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6011862754821777, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8583985567092896, "num_tokens": 571786386.0, "step": 15063 }, { "epoch": 1.9162956366874444, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7204267978668213, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8656297922134399, "num_tokens": 571824414.0, "step": 15064 }, { "epoch": 1.916422846966035, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.603726625442505, "learning_rate": 1e-05, "loss": 0.5015, "mean_token_accuracy": 0.8510684967041016, "num_tokens": 571863085.0, "step": 15065 }, { "epoch": 1.9165500572446255, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.73136043548584, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8703930974006653, "num_tokens": 571894749.0, "step": 15066 }, { "epoch": 1.916677267523216, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.613036870956421, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8630343675613403, "num_tokens": 571925962.0, "step": 15067 }, { "epoch": 1.9168044778018065, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7133796215057373, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8616636991500854, "num_tokens": 571966021.0, "step": 15068 }, { "epoch": 1.916931688080397, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.593538761138916, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8815274834632874, "num_tokens": 572002684.0, "step": 15069 }, { "epoch": 1.9170588983589874, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.720972776412964, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8735291957855225, "num_tokens": 572038054.0, "step": 15070 }, { "epoch": 1.9171861086375779, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.595987558364868, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8685639500617981, "num_tokens": 572076851.0, "step": 15071 }, { "epoch": 1.9173133189161684, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7129619121551514, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8759250640869141, "num_tokens": 572115642.0, "step": 15072 }, { "epoch": 1.917440529194759, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.58449649810791, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8799400925636292, "num_tokens": 572154803.0, "step": 15073 }, { "epoch": 1.9175677394733495, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.736933469772339, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8671493530273438, "num_tokens": 572189241.0, "step": 15074 }, { "epoch": 1.91769494975194, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5693559646606445, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8860031366348267, "num_tokens": 572230464.0, "step": 15075 }, { "epoch": 1.9178221600305305, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.741309881210327, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8822761178016663, "num_tokens": 572268863.0, "step": 15076 }, { "epoch": 1.9179493703091208, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.573269844055176, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8677353262901306, "num_tokens": 572307484.0, "step": 15077 }, { "epoch": 1.9180765805877114, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.74442195892334, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8702290654182434, "num_tokens": 572347653.0, "step": 15078 }, { "epoch": 1.9182037908663019, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5836706161499023, "learning_rate": 1e-05, "loss": 0.4884, "mean_token_accuracy": 0.858608603477478, "num_tokens": 572386850.0, "step": 15079 }, { "epoch": 1.9183310011448924, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7569332122802734, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8614553213119507, "num_tokens": 572422416.0, "step": 15080 }, { "epoch": 1.918458211423483, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5787370204925537, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8601529002189636, "num_tokens": 572458845.0, "step": 15081 }, { "epoch": 1.9185854217020735, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7513482570648193, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.873889148235321, "num_tokens": 572500553.0, "step": 15082 }, { "epoch": 1.918712631980664, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5777225494384766, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8636122941970825, "num_tokens": 572541315.0, "step": 15083 }, { "epoch": 1.9188398422592545, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7690203189849854, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8791394233703613, "num_tokens": 572574461.0, "step": 15084 }, { "epoch": 1.918967052537845, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5708069801330566, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8638318777084351, "num_tokens": 572616860.0, "step": 15085 }, { "epoch": 1.9190942628164356, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7708394527435303, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8766098022460938, "num_tokens": 572654228.0, "step": 15086 }, { "epoch": 1.919221473095026, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.570857286453247, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8704337477684021, "num_tokens": 572693603.0, "step": 15087 }, { "epoch": 1.9193486833736166, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7817494869232178, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8644886016845703, "num_tokens": 572730292.0, "step": 15088 }, { "epoch": 1.9194758936522072, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5815885066986084, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8667519092559814, "num_tokens": 572764348.0, "step": 15089 }, { "epoch": 1.9196031039307977, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.771106481552124, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8776594400405884, "num_tokens": 572798410.0, "step": 15090 }, { "epoch": 1.9197303142093882, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.585907220840454, "learning_rate": 1e-05, "loss": 0.4951, "mean_token_accuracy": 0.8592255115509033, "num_tokens": 572832067.0, "step": 15091 }, { "epoch": 1.9198575244879788, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7693943977355957, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8762470483779907, "num_tokens": 572866676.0, "step": 15092 }, { "epoch": 1.9199847347665693, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5660409927368164, "learning_rate": 1e-05, "loss": 0.4719, "mean_token_accuracy": 0.8616604804992676, "num_tokens": 572913631.0, "step": 15093 }, { "epoch": 1.9201119450451598, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.77778697013855, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8765244483947754, "num_tokens": 572953154.0, "step": 15094 }, { "epoch": 1.9202391553237501, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.575211524963379, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8648995161056519, "num_tokens": 572992259.0, "step": 15095 }, { "epoch": 1.9203663656023406, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7847177982330322, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8729948997497559, "num_tokens": 573024464.0, "step": 15096 }, { "epoch": 1.9204935758809312, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.577960252761841, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8772737979888916, "num_tokens": 573060620.0, "step": 15097 }, { "epoch": 1.9206207861595217, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7846133708953857, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8677773475646973, "num_tokens": 573098284.0, "step": 15098 }, { "epoch": 1.9207479964381122, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5803515911102295, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8723056316375732, "num_tokens": 573136253.0, "step": 15099 }, { "epoch": 1.9208752067167028, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7864718437194824, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8704757690429688, "num_tokens": 573174843.0, "step": 15100 }, { "epoch": 1.921002416995293, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.579026222229004, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8593051433563232, "num_tokens": 573215759.0, "step": 15101 }, { "epoch": 1.9211296272738836, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7796807289123535, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8743035197257996, "num_tokens": 573255297.0, "step": 15102 }, { "epoch": 1.9212568375524741, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5591914653778076, "learning_rate": 1e-05, "loss": 0.3573, "mean_token_accuracy": 0.8987504839897156, "num_tokens": 573293152.0, "step": 15103 }, { "epoch": 1.9213840478310646, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.790754795074463, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8702321648597717, "num_tokens": 573330456.0, "step": 15104 }, { "epoch": 1.9215112581096552, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5772626399993896, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8627976179122925, "num_tokens": 573371595.0, "step": 15105 }, { "epoch": 1.9216384683882457, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7871944904327393, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8665227890014648, "num_tokens": 573417414.0, "step": 15106 }, { "epoch": 1.9217656786668362, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5896031856536865, "learning_rate": 1e-05, "loss": 0.4657, "mean_token_accuracy": 0.8660928606987, "num_tokens": 573452146.0, "step": 15107 }, { "epoch": 1.9218928889454268, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.801295042037964, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8638821840286255, "num_tokens": 573485469.0, "step": 15108 }, { "epoch": 1.9220200992240173, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.575183391571045, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8760532140731812, "num_tokens": 573524763.0, "step": 15109 }, { "epoch": 1.9221473095026078, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8015191555023193, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8663015961647034, "num_tokens": 573559326.0, "step": 15110 }, { "epoch": 1.9222745197811983, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.567080497741699, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8665317296981812, "num_tokens": 573604700.0, "step": 15111 }, { "epoch": 1.9224017300597889, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.801478385925293, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.8838110566139221, "num_tokens": 573639570.0, "step": 15112 }, { "epoch": 1.9225289403383794, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6026535034179688, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.867189884185791, "num_tokens": 573669849.0, "step": 15113 }, { "epoch": 1.92265615061697, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.803464651107788, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8688023090362549, "num_tokens": 573702402.0, "step": 15114 }, { "epoch": 1.9227833608955605, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.593191623687744, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8671582937240601, "num_tokens": 573737428.0, "step": 15115 }, { "epoch": 1.922910571174151, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.776129722595215, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8771843910217285, "num_tokens": 573775111.0, "step": 15116 }, { "epoch": 1.9230377814527415, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5954415798187256, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8624093532562256, "num_tokens": 573814550.0, "step": 15117 }, { "epoch": 1.923164991731332, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.782186269760132, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8607257604598999, "num_tokens": 573852932.0, "step": 15118 }, { "epoch": 1.9232922020099223, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5965464115142822, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.879298210144043, "num_tokens": 573887769.0, "step": 15119 }, { "epoch": 1.9234194122885129, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7675018310546875, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8655411005020142, "num_tokens": 573925258.0, "step": 15120 }, { "epoch": 1.9235466225671034, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.596813440322876, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8823267221450806, "num_tokens": 573956690.0, "step": 15121 }, { "epoch": 1.923673832845694, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.764181613922119, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8587691187858582, "num_tokens": 573993822.0, "step": 15122 }, { "epoch": 1.9238010431242845, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5964314937591553, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8691614270210266, "num_tokens": 574030087.0, "step": 15123 }, { "epoch": 1.923928253402875, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7459254264831543, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8783504366874695, "num_tokens": 574071964.0, "step": 15124 }, { "epoch": 1.9240554636814655, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6044514179229736, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8659651279449463, "num_tokens": 574105706.0, "step": 15125 }, { "epoch": 1.9241826739600558, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.736768960952759, "learning_rate": 1e-05, "loss": 0.496, "mean_token_accuracy": 0.8551186323165894, "num_tokens": 574145798.0, "step": 15126 }, { "epoch": 1.9243098842386464, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5759119987487793, "learning_rate": 1e-05, "loss": 0.3922, "mean_token_accuracy": 0.886154294013977, "num_tokens": 574192631.0, "step": 15127 }, { "epoch": 1.9244370945172369, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.758629083633423, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8523956537246704, "num_tokens": 574228601.0, "step": 15128 }, { "epoch": 1.9245643047958274, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5810582637786865, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8770910501480103, "num_tokens": 574266416.0, "step": 15129 }, { "epoch": 1.924691515074418, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7812082767486572, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8601382970809937, "num_tokens": 574297414.0, "step": 15130 }, { "epoch": 1.9248187253530085, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.570852756500244, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.888809323310852, "num_tokens": 574333749.0, "step": 15131 }, { "epoch": 1.924945935631599, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7689948081970215, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.869928240776062, "num_tokens": 574377186.0, "step": 15132 }, { "epoch": 1.9250731459101895, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5752687454223633, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8754346966743469, "num_tokens": 574414814.0, "step": 15133 }, { "epoch": 1.92520035618878, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7791495323181152, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8645431399345398, "num_tokens": 574453070.0, "step": 15134 }, { "epoch": 1.9253275664673706, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5817742347717285, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.8787298202514648, "num_tokens": 574487564.0, "step": 15135 }, { "epoch": 1.925454776745961, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7771735191345215, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8689941167831421, "num_tokens": 574528456.0, "step": 15136 }, { "epoch": 1.9255819870245516, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5784354209899902, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8723529577255249, "num_tokens": 574567724.0, "step": 15137 }, { "epoch": 1.9257091973031422, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.79525089263916, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8729625344276428, "num_tokens": 574599702.0, "step": 15138 }, { "epoch": 1.9258364075817327, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5845961570739746, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8770709037780762, "num_tokens": 574633709.0, "step": 15139 }, { "epoch": 1.9259636178603232, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7805614471435547, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8694443702697754, "num_tokens": 574670490.0, "step": 15140 }, { "epoch": 1.9260908281389137, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.593895196914673, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8571351766586304, "num_tokens": 574706563.0, "step": 15141 }, { "epoch": 1.9262180384175043, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7863950729370117, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8748928904533386, "num_tokens": 574744743.0, "step": 15142 }, { "epoch": 1.9263452486960948, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5806617736816406, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8697259426116943, "num_tokens": 574782236.0, "step": 15143 }, { "epoch": 1.926472458974685, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7754030227661133, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8825974464416504, "num_tokens": 574826457.0, "step": 15144 }, { "epoch": 1.9265996692532756, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5794966220855713, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8626456260681152, "num_tokens": 574870362.0, "step": 15145 }, { "epoch": 1.9267268795318662, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7766776084899902, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8767932653427124, "num_tokens": 574907984.0, "step": 15146 }, { "epoch": 1.9268540898104567, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5816867351531982, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8645834922790527, "num_tokens": 574950813.0, "step": 15147 }, { "epoch": 1.9269813000890472, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.81105899810791, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8744762539863586, "num_tokens": 574985210.0, "step": 15148 }, { "epoch": 1.9271085103676378, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5783228874206543, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.877303957939148, "num_tokens": 575020263.0, "step": 15149 }, { "epoch": 1.927235720646228, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7920119762420654, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8682082891464233, "num_tokens": 575060632.0, "step": 15150 }, { "epoch": 1.9273629309248186, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.57228422164917, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8681674599647522, "num_tokens": 575102725.0, "step": 15151 }, { "epoch": 1.9274901412034091, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7851297855377197, "learning_rate": 1e-05, "loss": 0.3942, "mean_token_accuracy": 0.8872513771057129, "num_tokens": 575143069.0, "step": 15152 }, { "epoch": 1.9276173514819996, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5763323307037354, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8737232685089111, "num_tokens": 575182618.0, "step": 15153 }, { "epoch": 1.9277445617605902, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.799393892288208, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8714010715484619, "num_tokens": 575224576.0, "step": 15154 }, { "epoch": 1.9278717720391807, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5805907249450684, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8814186453819275, "num_tokens": 575260152.0, "step": 15155 }, { "epoch": 1.9279989823177712, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.8056981563568115, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8690022826194763, "num_tokens": 575292308.0, "step": 15156 }, { "epoch": 1.9281261925963618, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.59139084815979, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8651936054229736, "num_tokens": 575326305.0, "step": 15157 }, { "epoch": 1.9282534028749523, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7970309257507324, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8675947189331055, "num_tokens": 575361775.0, "step": 15158 }, { "epoch": 1.9283806131535428, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5832276344299316, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8802909851074219, "num_tokens": 575397508.0, "step": 15159 }, { "epoch": 1.9285078234321333, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.781886577606201, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8690328598022461, "num_tokens": 575437770.0, "step": 15160 }, { "epoch": 1.9286350337107239, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5917041301727295, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8761328458786011, "num_tokens": 575478322.0, "step": 15161 }, { "epoch": 1.9287622439893144, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.78355073928833, "learning_rate": 1e-05, "loss": 0.4896, "mean_token_accuracy": 0.8594675660133362, "num_tokens": 575518816.0, "step": 15162 }, { "epoch": 1.928889454267905, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.601435422897339, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8628753423690796, "num_tokens": 575554276.0, "step": 15163 }, { "epoch": 1.9290166645464955, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7843921184539795, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8658571839332581, "num_tokens": 575591331.0, "step": 15164 }, { "epoch": 1.929143874825086, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6043248176574707, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8687002658843994, "num_tokens": 575629449.0, "step": 15165 }, { "epoch": 1.9292710851036765, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7662134170532227, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.858106791973114, "num_tokens": 575668169.0, "step": 15166 }, { "epoch": 1.929398295382267, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5977611541748047, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8612164258956909, "num_tokens": 575708481.0, "step": 15167 }, { "epoch": 1.9295255056608573, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.777287006378174, "learning_rate": 1e-05, "loss": 0.5083, "mean_token_accuracy": 0.8501013517379761, "num_tokens": 575744879.0, "step": 15168 }, { "epoch": 1.9296527159394479, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.580933094024658, "learning_rate": 1e-05, "loss": 0.3783, "mean_token_accuracy": 0.8922187089920044, "num_tokens": 575781587.0, "step": 15169 }, { "epoch": 1.9297799262180384, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.780927896499634, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8699793815612793, "num_tokens": 575816662.0, "step": 15170 }, { "epoch": 1.929907136496629, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6025779247283936, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8653437495231628, "num_tokens": 575855169.0, "step": 15171 }, { "epoch": 1.9300343467752195, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7500240802764893, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8823550343513489, "num_tokens": 575896939.0, "step": 15172 }, { "epoch": 1.93016155705381, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.595663547515869, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8760459423065186, "num_tokens": 575930587.0, "step": 15173 }, { "epoch": 1.9302887673324005, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7640063762664795, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8561819195747375, "num_tokens": 575964432.0, "step": 15174 }, { "epoch": 1.9304159776109908, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6098525524139404, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8739631175994873, "num_tokens": 576004916.0, "step": 15175 }, { "epoch": 1.9305431878895813, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.740630865097046, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8514419794082642, "num_tokens": 576047857.0, "step": 15176 }, { "epoch": 1.9306703981681719, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5971779823303223, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8622759580612183, "num_tokens": 576089087.0, "step": 15177 }, { "epoch": 1.9307976084467624, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.721956729888916, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8740457892417908, "num_tokens": 576134106.0, "step": 15178 }, { "epoch": 1.930924818725353, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.59302020072937, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8683895468711853, "num_tokens": 576172566.0, "step": 15179 }, { "epoch": 1.9310520290039435, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7396624088287354, "learning_rate": 1e-05, "loss": 0.537, "mean_token_accuracy": 0.8468296527862549, "num_tokens": 576214469.0, "step": 15180 }, { "epoch": 1.931179239282534, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.579407215118408, "learning_rate": 1e-05, "loss": 0.3711, "mean_token_accuracy": 0.8929634094238281, "num_tokens": 576249316.0, "step": 15181 }, { "epoch": 1.9313064495611245, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.747218608856201, "learning_rate": 1e-05, "loss": 0.5019, "mean_token_accuracy": 0.8509352803230286, "num_tokens": 576290748.0, "step": 15182 }, { "epoch": 1.931433659839715, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5819125175476074, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8725032806396484, "num_tokens": 576329137.0, "step": 15183 }, { "epoch": 1.9315608701183056, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7412736415863037, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8843766450881958, "num_tokens": 576367147.0, "step": 15184 }, { "epoch": 1.931688080396896, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5834484100341797, "learning_rate": 1e-05, "loss": 0.4919, "mean_token_accuracy": 0.8626044392585754, "num_tokens": 576406068.0, "step": 15185 }, { "epoch": 1.9318152906754866, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7599267959594727, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8658226728439331, "num_tokens": 576438150.0, "step": 15186 }, { "epoch": 1.9319425009540772, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.582628011703491, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8754975199699402, "num_tokens": 576470630.0, "step": 15187 }, { "epoch": 1.9320697112326677, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.752140522003174, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8776654005050659, "num_tokens": 576505896.0, "step": 15188 }, { "epoch": 1.9321969215112582, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.565936803817749, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8810452818870544, "num_tokens": 576545204.0, "step": 15189 }, { "epoch": 1.9323241317898487, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7628612518310547, "learning_rate": 1e-05, "loss": 0.5076, "mean_token_accuracy": 0.8593369126319885, "num_tokens": 576590746.0, "step": 15190 }, { "epoch": 1.9324513420684393, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5706088542938232, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8686171770095825, "num_tokens": 576631185.0, "step": 15191 }, { "epoch": 1.9325785523470298, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7808308601379395, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8701863288879395, "num_tokens": 576668945.0, "step": 15192 }, { "epoch": 1.93270576262562, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5733306407928467, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8723219633102417, "num_tokens": 576706052.0, "step": 15193 }, { "epoch": 1.9328329729042106, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.788120985031128, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8722355961799622, "num_tokens": 576740668.0, "step": 15194 }, { "epoch": 1.9329601831828012, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.569550037384033, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8613147735595703, "num_tokens": 576780312.0, "step": 15195 }, { "epoch": 1.9330873934613917, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7833917140960693, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.882838785648346, "num_tokens": 576817130.0, "step": 15196 }, { "epoch": 1.9332146037399822, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5732839107513428, "learning_rate": 1e-05, "loss": 0.3879, "mean_token_accuracy": 0.886864185333252, "num_tokens": 576849823.0, "step": 15197 }, { "epoch": 1.9333418140185727, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.793879985809326, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8627157211303711, "num_tokens": 576890890.0, "step": 15198 }, { "epoch": 1.933469024297163, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5805370807647705, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8565167188644409, "num_tokens": 576927926.0, "step": 15199 }, { "epoch": 1.9335962345757536, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8054521083831787, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8604612946510315, "num_tokens": 576966194.0, "step": 15200 }, { "epoch": 1.933723444854344, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.577587842941284, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8593783974647522, "num_tokens": 577007218.0, "step": 15201 }, { "epoch": 1.9338506551329346, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7932584285736084, "learning_rate": 1e-05, "loss": 0.3902, "mean_token_accuracy": 0.8891331553459167, "num_tokens": 577040551.0, "step": 15202 }, { "epoch": 1.9339778654115252, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5753438472747803, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8717304468154907, "num_tokens": 577082676.0, "step": 15203 }, { "epoch": 1.9341050756901157, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.791870355606079, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8770517110824585, "num_tokens": 577124572.0, "step": 15204 }, { "epoch": 1.9342322859687062, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.567199468612671, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8834244608879089, "num_tokens": 577167571.0, "step": 15205 }, { "epoch": 1.9343594962472968, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.800635814666748, "learning_rate": 1e-05, "loss": 0.4977, "mean_token_accuracy": 0.8523234724998474, "num_tokens": 577211976.0, "step": 15206 }, { "epoch": 1.9344867065258873, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.578566551208496, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8768964409828186, "num_tokens": 577252572.0, "step": 15207 }, { "epoch": 1.9346139168044778, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.814890146255493, "learning_rate": 1e-05, "loss": 0.4839, "mean_token_accuracy": 0.8571289777755737, "num_tokens": 577286923.0, "step": 15208 }, { "epoch": 1.9347411270830683, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5789339542388916, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8682912588119507, "num_tokens": 577324520.0, "step": 15209 }, { "epoch": 1.9348683373616589, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8076462745666504, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8674978017807007, "num_tokens": 577361100.0, "step": 15210 }, { "epoch": 1.9349955476402494, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5807135105133057, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8685768246650696, "num_tokens": 577402714.0, "step": 15211 }, { "epoch": 1.93512275791884, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.827066659927368, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8590054512023926, "num_tokens": 577437226.0, "step": 15212 }, { "epoch": 1.9352499681974304, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.587326765060425, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8742517828941345, "num_tokens": 577475564.0, "step": 15213 }, { "epoch": 1.935377178476021, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8255691528320312, "learning_rate": 1e-05, "loss": 0.5219, "mean_token_accuracy": 0.8521831631660461, "num_tokens": 577512623.0, "step": 15214 }, { "epoch": 1.9355043887546115, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.587942361831665, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8717507123947144, "num_tokens": 577545515.0, "step": 15215 }, { "epoch": 1.935631599033202, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8194470405578613, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.871706485748291, "num_tokens": 577577043.0, "step": 15216 }, { "epoch": 1.9357588093117923, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5834057331085205, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8782175183296204, "num_tokens": 577616311.0, "step": 15217 }, { "epoch": 1.9358860195903829, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8024723529815674, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.875068187713623, "num_tokens": 577654963.0, "step": 15218 }, { "epoch": 1.9360132298689734, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.595857858657837, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8813048601150513, "num_tokens": 577688283.0, "step": 15219 }, { "epoch": 1.936140440147564, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7904844284057617, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8622576594352722, "num_tokens": 577726855.0, "step": 15220 }, { "epoch": 1.9362676504261545, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6034483909606934, "learning_rate": 1e-05, "loss": 0.4959, "mean_token_accuracy": 0.8587010502815247, "num_tokens": 577765699.0, "step": 15221 }, { "epoch": 1.936394860704745, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7796828746795654, "learning_rate": 1e-05, "loss": 0.3773, "mean_token_accuracy": 0.8934211730957031, "num_tokens": 577797393.0, "step": 15222 }, { "epoch": 1.9365220709833355, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5929324626922607, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8799883127212524, "num_tokens": 577838768.0, "step": 15223 }, { "epoch": 1.9366492812619258, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.765131711959839, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8744279146194458, "num_tokens": 577882455.0, "step": 15224 }, { "epoch": 1.9367764915405163, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.583681583404541, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8854666352272034, "num_tokens": 577922649.0, "step": 15225 }, { "epoch": 1.9369037018191069, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.771712303161621, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8721747398376465, "num_tokens": 577961679.0, "step": 15226 }, { "epoch": 1.9370309120976974, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5902259349823, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8754413723945618, "num_tokens": 578001848.0, "step": 15227 }, { "epoch": 1.937158122376288, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7695066928863525, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8745474815368652, "num_tokens": 578039278.0, "step": 15228 }, { "epoch": 1.9372853326548785, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.595747232437134, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8693697452545166, "num_tokens": 578078152.0, "step": 15229 }, { "epoch": 1.937412542933469, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7684714794158936, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8649047613143921, "num_tokens": 578115927.0, "step": 15230 }, { "epoch": 1.9375397532120595, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5782358646392822, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.874907374382019, "num_tokens": 578157425.0, "step": 15231 }, { "epoch": 1.93766696349065, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.760775327682495, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8778047561645508, "num_tokens": 578201985.0, "step": 15232 }, { "epoch": 1.9377941737692406, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5880956649780273, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.870349645614624, "num_tokens": 578239471.0, "step": 15233 }, { "epoch": 1.937921384047831, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7652158737182617, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8796446323394775, "num_tokens": 578272302.0, "step": 15234 }, { "epoch": 1.9380485943264216, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5826613903045654, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.870202898979187, "num_tokens": 578314585.0, "step": 15235 }, { "epoch": 1.9381758046050122, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7689061164855957, "learning_rate": 1e-05, "loss": 0.5005, "mean_token_accuracy": 0.8505802750587463, "num_tokens": 578355561.0, "step": 15236 }, { "epoch": 1.9383030148836027, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.583603620529175, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8683798313140869, "num_tokens": 578398903.0, "step": 15237 }, { "epoch": 1.9384302251621932, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7756717205047607, "learning_rate": 1e-05, "loss": 0.5322, "mean_token_accuracy": 0.8446064591407776, "num_tokens": 578439293.0, "step": 15238 }, { "epoch": 1.9385574354407837, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5945448875427246, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8539731502532959, "num_tokens": 578477965.0, "step": 15239 }, { "epoch": 1.9386846457193743, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.763700008392334, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8761013746261597, "num_tokens": 578515577.0, "step": 15240 }, { "epoch": 1.9388118559979648, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.579080104827881, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8710010051727295, "num_tokens": 578557415.0, "step": 15241 }, { "epoch": 1.938939066276555, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7657265663146973, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8763516545295715, "num_tokens": 578594556.0, "step": 15242 }, { "epoch": 1.9390662765551456, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.599611282348633, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8810598850250244, "num_tokens": 578627219.0, "step": 15243 }, { "epoch": 1.9391934868337362, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7625539302825928, "learning_rate": 1e-05, "loss": 0.4897, "mean_token_accuracy": 0.8566358089447021, "num_tokens": 578669135.0, "step": 15244 }, { "epoch": 1.9393206971123267, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5993261337280273, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8772637844085693, "num_tokens": 578706174.0, "step": 15245 }, { "epoch": 1.9394479073909172, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7512409687042236, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8685749769210815, "num_tokens": 578738942.0, "step": 15246 }, { "epoch": 1.9395751176695077, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.597907543182373, "learning_rate": 1e-05, "loss": 0.3734, "mean_token_accuracy": 0.8922412991523743, "num_tokens": 578771553.0, "step": 15247 }, { "epoch": 1.939702327948098, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.741140842437744, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8668927550315857, "num_tokens": 578811532.0, "step": 15248 }, { "epoch": 1.9398295382266886, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5911309719085693, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8684384822845459, "num_tokens": 578852858.0, "step": 15249 }, { "epoch": 1.939956748505279, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7441844940185547, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8668531775474548, "num_tokens": 578890340.0, "step": 15250 }, { "epoch": 1.9400839587838696, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5866942405700684, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8795958161354065, "num_tokens": 578928953.0, "step": 15251 }, { "epoch": 1.9402111690624602, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7477667331695557, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8807148337364197, "num_tokens": 578967978.0, "step": 15252 }, { "epoch": 1.9403383793410507, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.602750062942505, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8774083256721497, "num_tokens": 579005484.0, "step": 15253 }, { "epoch": 1.9404655896196412, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.742610216140747, "learning_rate": 1e-05, "loss": 0.4996, "mean_token_accuracy": 0.8541504144668579, "num_tokens": 579045172.0, "step": 15254 }, { "epoch": 1.9405927998982317, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6022212505340576, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8812257051467896, "num_tokens": 579075001.0, "step": 15255 }, { "epoch": 1.9407200101768223, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7411088943481445, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8707367777824402, "num_tokens": 579116029.0, "step": 15256 }, { "epoch": 1.9408472204554128, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.581738233566284, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8696091175079346, "num_tokens": 579157741.0, "step": 15257 }, { "epoch": 1.9409744307340033, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7435922622680664, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.871640682220459, "num_tokens": 579196033.0, "step": 15258 }, { "epoch": 1.9411016410125939, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5728836059570312, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8808125853538513, "num_tokens": 579235060.0, "step": 15259 }, { "epoch": 1.9412288512911844, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7620859146118164, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8736557960510254, "num_tokens": 579272184.0, "step": 15260 }, { "epoch": 1.941356061569775, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5714409351348877, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8725184202194214, "num_tokens": 579309461.0, "step": 15261 }, { "epoch": 1.9414832718483654, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.768092393875122, "learning_rate": 1e-05, "loss": 0.4922, "mean_token_accuracy": 0.8551629185676575, "num_tokens": 579351871.0, "step": 15262 }, { "epoch": 1.941610482126956, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5748956203460693, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8568279147148132, "num_tokens": 579390123.0, "step": 15263 }, { "epoch": 1.9417376924055465, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.778920888900757, "learning_rate": 1e-05, "loss": 0.4806, "mean_token_accuracy": 0.8613656163215637, "num_tokens": 579426015.0, "step": 15264 }, { "epoch": 1.941864902684137, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5674846172332764, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8577562570571899, "num_tokens": 579465853.0, "step": 15265 }, { "epoch": 1.9419921129627273, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7840983867645264, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.867085337638855, "num_tokens": 579499740.0, "step": 15266 }, { "epoch": 1.9421193232413179, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5567026138305664, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8869675397872925, "num_tokens": 579538908.0, "step": 15267 }, { "epoch": 1.9422465335199084, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.797417163848877, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8726190328598022, "num_tokens": 579576026.0, "step": 15268 }, { "epoch": 1.942373743798499, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5576012134552, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8770260214805603, "num_tokens": 579618980.0, "step": 15269 }, { "epoch": 1.9425009540770894, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8142664432525635, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8581643104553223, "num_tokens": 579659205.0, "step": 15270 }, { "epoch": 1.94262816435568, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5625364780426025, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8825893402099609, "num_tokens": 579695598.0, "step": 15271 }, { "epoch": 1.9427553746342705, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.821972131729126, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8725488781929016, "num_tokens": 579730871.0, "step": 15272 }, { "epoch": 1.9428825849128608, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5648906230926514, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8675203323364258, "num_tokens": 579772879.0, "step": 15273 }, { "epoch": 1.9430097951914513, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.821232795715332, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8670253753662109, "num_tokens": 579812854.0, "step": 15274 }, { "epoch": 1.9431370054700419, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5775389671325684, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8652745485305786, "num_tokens": 579857324.0, "step": 15275 }, { "epoch": 1.9432642157486324, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8293657302856445, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8721864223480225, "num_tokens": 579897103.0, "step": 15276 }, { "epoch": 1.943391426027223, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5846970081329346, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8644298911094666, "num_tokens": 579937382.0, "step": 15277 }, { "epoch": 1.9435186363058135, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.817272186279297, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.876413881778717, "num_tokens": 579978591.0, "step": 15278 }, { "epoch": 1.943645846584404, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.593428611755371, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8755649328231812, "num_tokens": 580014354.0, "step": 15279 }, { "epoch": 1.9437730568629945, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7955305576324463, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8775118589401245, "num_tokens": 580055575.0, "step": 15280 }, { "epoch": 1.943900267141585, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5833022594451904, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8652164340019226, "num_tokens": 580105007.0, "step": 15281 }, { "epoch": 1.9440274774201756, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.79310941696167, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8624258041381836, "num_tokens": 580150820.0, "step": 15282 }, { "epoch": 1.944154687698766, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.600804328918457, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8820332884788513, "num_tokens": 580188303.0, "step": 15283 }, { "epoch": 1.9442818979773566, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7971115112304688, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8696840405464172, "num_tokens": 580221824.0, "step": 15284 }, { "epoch": 1.9444091082559471, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.590308666229248, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8689109086990356, "num_tokens": 580262615.0, "step": 15285 }, { "epoch": 1.9445363185345377, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7948672771453857, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8770413398742676, "num_tokens": 580293535.0, "step": 15286 }, { "epoch": 1.9446635288131282, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.619323492050171, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8690894842147827, "num_tokens": 580328645.0, "step": 15287 }, { "epoch": 1.9447907390917187, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7674202919006348, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8716745972633362, "num_tokens": 580365157.0, "step": 15288 }, { "epoch": 1.9449179493703093, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6012539863586426, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8781791925430298, "num_tokens": 580398195.0, "step": 15289 }, { "epoch": 1.9450451596488998, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7457051277160645, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8681490421295166, "num_tokens": 580440238.0, "step": 15290 }, { "epoch": 1.94517236992749, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6136696338653564, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8645713329315186, "num_tokens": 580477997.0, "step": 15291 }, { "epoch": 1.9452995802060806, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7371442317962646, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8726788759231567, "num_tokens": 580516335.0, "step": 15292 }, { "epoch": 1.9454267904846712, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6123619079589844, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8820421695709229, "num_tokens": 580546492.0, "step": 15293 }, { "epoch": 1.9455540007632617, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7230324745178223, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8843353986740112, "num_tokens": 580583804.0, "step": 15294 }, { "epoch": 1.9456812110418522, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.606764078140259, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8603239059448242, "num_tokens": 580618111.0, "step": 15295 }, { "epoch": 1.9458084213204427, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7166173458099365, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8689842224121094, "num_tokens": 580657058.0, "step": 15296 }, { "epoch": 1.945935631599033, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.599339246749878, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.862145185470581, "num_tokens": 580701628.0, "step": 15297 }, { "epoch": 1.9460628418776236, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.713545799255371, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.875386118888855, "num_tokens": 580738766.0, "step": 15298 }, { "epoch": 1.946190052156214, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.594097137451172, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8723313808441162, "num_tokens": 580780110.0, "step": 15299 }, { "epoch": 1.9463172624348046, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7141566276550293, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8785872459411621, "num_tokens": 580825208.0, "step": 15300 }, { "epoch": 1.9464444727133952, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5947747230529785, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.853497326374054, "num_tokens": 580870339.0, "step": 15301 }, { "epoch": 1.9465716829919857, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7305493354797363, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.881807804107666, "num_tokens": 580899509.0, "step": 15302 }, { "epoch": 1.9466988932705762, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.585481882095337, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8736447095870972, "num_tokens": 580941345.0, "step": 15303 }, { "epoch": 1.9468261035491667, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7279646396636963, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.878860592842102, "num_tokens": 580976819.0, "step": 15304 }, { "epoch": 1.9469533138277573, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5905542373657227, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.877084493637085, "num_tokens": 581011971.0, "step": 15305 }, { "epoch": 1.9470805241063478, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7533462047576904, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8565348982810974, "num_tokens": 581045871.0, "step": 15306 }, { "epoch": 1.9472077343849383, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5918049812316895, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8808325529098511, "num_tokens": 581077986.0, "step": 15307 }, { "epoch": 1.9473349446635289, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.748089075088501, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8740975856781006, "num_tokens": 581112988.0, "step": 15308 }, { "epoch": 1.9474621549421194, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5812265872955322, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8636170029640198, "num_tokens": 581155314.0, "step": 15309 }, { "epoch": 1.94758936522071, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.731027364730835, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8772811889648438, "num_tokens": 581203622.0, "step": 15310 }, { "epoch": 1.9477165754993004, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6329381465911865, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8755179643630981, "num_tokens": 581234306.0, "step": 15311 }, { "epoch": 1.947843785777891, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7129440307617188, "learning_rate": 1e-05, "loss": 0.4974, "mean_token_accuracy": 0.8549996614456177, "num_tokens": 581275884.0, "step": 15312 }, { "epoch": 1.9479709960564815, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.59521222114563, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8858786821365356, "num_tokens": 581312674.0, "step": 15313 }, { "epoch": 1.948098206335072, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6974151134490967, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8640277981758118, "num_tokens": 581352400.0, "step": 15314 }, { "epoch": 1.9482254166136623, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6101043224334717, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8668860197067261, "num_tokens": 581390136.0, "step": 15315 }, { "epoch": 1.9483526268922529, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6999573707580566, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8838701248168945, "num_tokens": 581424864.0, "step": 15316 }, { "epoch": 1.9484798371708434, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.6119930744171143, "learning_rate": 1e-05, "loss": 0.5006, "mean_token_accuracy": 0.8551715016365051, "num_tokens": 581467791.0, "step": 15317 }, { "epoch": 1.948607047449434, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6967759132385254, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8716686964035034, "num_tokens": 581505121.0, "step": 15318 }, { "epoch": 1.9487342577280244, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5947418212890625, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8762896060943604, "num_tokens": 581543306.0, "step": 15319 }, { "epoch": 1.948861468006615, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7195088863372803, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8674527406692505, "num_tokens": 581573476.0, "step": 15320 }, { "epoch": 1.9489886782852053, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5905959606170654, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8838996887207031, "num_tokens": 581606693.0, "step": 15321 }, { "epoch": 1.9491158885637958, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.726071357727051, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8710998296737671, "num_tokens": 581637826.0, "step": 15322 }, { "epoch": 1.9492430988423863, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.5823192596435547, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8838768005371094, "num_tokens": 581670590.0, "step": 15323 }, { "epoch": 1.9493703091209769, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7460601329803467, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8712900280952454, "num_tokens": 581704091.0, "step": 15324 }, { "epoch": 1.9494975193995674, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.562760353088379, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8786371946334839, "num_tokens": 581747362.0, "step": 15325 }, { "epoch": 1.949624729678158, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.765564203262329, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8748472332954407, "num_tokens": 581781105.0, "step": 15326 }, { "epoch": 1.9497519399567484, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5578348636627197, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8760162591934204, "num_tokens": 581824800.0, "step": 15327 }, { "epoch": 1.949879150235339, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.777954578399658, "learning_rate": 1e-05, "loss": 0.478, "mean_token_accuracy": 0.8599923849105835, "num_tokens": 581865984.0, "step": 15328 }, { "epoch": 1.9500063605139295, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5617358684539795, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.872605562210083, "num_tokens": 581903098.0, "step": 15329 }, { "epoch": 1.95013357079252, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7896780967712402, "learning_rate": 1e-05, "loss": 0.4014, "mean_token_accuracy": 0.8852319717407227, "num_tokens": 581940704.0, "step": 15330 }, { "epoch": 1.9502607810711106, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.559619903564453, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8606382012367249, "num_tokens": 581983598.0, "step": 15331 }, { "epoch": 1.950387991349701, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8027122020721436, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8745864033699036, "num_tokens": 582020073.0, "step": 15332 }, { "epoch": 1.9505152016282916, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5655624866485596, "learning_rate": 1e-05, "loss": 0.3937, "mean_token_accuracy": 0.8853302001953125, "num_tokens": 582056597.0, "step": 15333 }, { "epoch": 1.9506424119068821, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8128371238708496, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.8864629864692688, "num_tokens": 582090164.0, "step": 15334 }, { "epoch": 1.9507696221854727, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5576956272125244, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8789209127426147, "num_tokens": 582130553.0, "step": 15335 }, { "epoch": 1.9508968324640632, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8406906127929688, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8782280683517456, "num_tokens": 582161175.0, "step": 15336 }, { "epoch": 1.9510240427426537, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5704855918884277, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8733609318733215, "num_tokens": 582194424.0, "step": 15337 }, { "epoch": 1.9511512530212443, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8421132564544678, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8692002296447754, "num_tokens": 582231446.0, "step": 15338 }, { "epoch": 1.9512784632998348, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5814871788024902, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8692476749420166, "num_tokens": 582270486.0, "step": 15339 }, { "epoch": 1.951405673578425, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.842061758041382, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8738742470741272, "num_tokens": 582303583.0, "step": 15340 }, { "epoch": 1.9515328838570156, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5819272994995117, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8712775707244873, "num_tokens": 582344982.0, "step": 15341 }, { "epoch": 1.9516600941356061, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.827618360519409, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8805906772613525, "num_tokens": 582383373.0, "step": 15342 }, { "epoch": 1.9517873044141967, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5819573402404785, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8714144229888916, "num_tokens": 582426671.0, "step": 15343 }, { "epoch": 1.9519145146927872, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.833897113800049, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8643894791603088, "num_tokens": 582464801.0, "step": 15344 }, { "epoch": 1.9520417249713777, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.5846328735351562, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8703544735908508, "num_tokens": 582507627.0, "step": 15345 }, { "epoch": 1.952168935249968, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.82114315032959, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8812724351882935, "num_tokens": 582544072.0, "step": 15346 }, { "epoch": 1.9522961455285586, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5935330390930176, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8676255941390991, "num_tokens": 582587928.0, "step": 15347 }, { "epoch": 1.952423355807149, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.827272653579712, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8715312480926514, "num_tokens": 582622182.0, "step": 15348 }, { "epoch": 1.9525505660857396, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.604712963104248, "learning_rate": 1e-05, "loss": 0.3891, "mean_token_accuracy": 0.8902750015258789, "num_tokens": 582654890.0, "step": 15349 }, { "epoch": 1.9526777763643302, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.808574676513672, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8765499591827393, "num_tokens": 582690874.0, "step": 15350 }, { "epoch": 1.9528049866429207, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5964930057525635, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8764948844909668, "num_tokens": 582726543.0, "step": 15351 }, { "epoch": 1.9529321969215112, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8009610176086426, "learning_rate": 1e-05, "loss": 0.4769, "mean_token_accuracy": 0.8622373342514038, "num_tokens": 582764552.0, "step": 15352 }, { "epoch": 1.9530594072001017, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.601088047027588, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8594815731048584, "num_tokens": 582804613.0, "step": 15353 }, { "epoch": 1.9531866174786923, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.789175510406494, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8761844635009766, "num_tokens": 582841510.0, "step": 15354 }, { "epoch": 1.9533138277572828, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6127960681915283, "learning_rate": 1e-05, "loss": 0.5222, "mean_token_accuracy": 0.8423221111297607, "num_tokens": 582879371.0, "step": 15355 }, { "epoch": 1.9534410380358733, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.773905038833618, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8650836944580078, "num_tokens": 582915354.0, "step": 15356 }, { "epoch": 1.9535682483144639, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.60951566696167, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8783972263336182, "num_tokens": 582949897.0, "step": 15357 }, { "epoch": 1.9536954585930544, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7771451473236084, "learning_rate": 1e-05, "loss": 0.4867, "mean_token_accuracy": 0.8600032329559326, "num_tokens": 582984383.0, "step": 15358 }, { "epoch": 1.953822668871645, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.606018543243408, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8661129474639893, "num_tokens": 583023146.0, "step": 15359 }, { "epoch": 1.9539498791502354, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.746652841567993, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8829303979873657, "num_tokens": 583061485.0, "step": 15360 }, { "epoch": 1.954077089428826, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.610588312149048, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8683863282203674, "num_tokens": 583097278.0, "step": 15361 }, { "epoch": 1.9542042997074165, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.725494623184204, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8777127861976624, "num_tokens": 583140458.0, "step": 15362 }, { "epoch": 1.954331509986007, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.59597110748291, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8785327076911926, "num_tokens": 583184798.0, "step": 15363 }, { "epoch": 1.9544587202645973, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.723313093185425, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8805065155029297, "num_tokens": 583220517.0, "step": 15364 }, { "epoch": 1.9545859305431879, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6121630668640137, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8611365556716919, "num_tokens": 583256615.0, "step": 15365 }, { "epoch": 1.9547131408217784, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.720846652984619, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8665702939033508, "num_tokens": 583294533.0, "step": 15366 }, { "epoch": 1.954840351100369, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.598416328430176, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8818249702453613, "num_tokens": 583333610.0, "step": 15367 }, { "epoch": 1.9549675613789594, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.728904962539673, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8794339299201965, "num_tokens": 583366424.0, "step": 15368 }, { "epoch": 1.95509477165755, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5966689586639404, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.870400607585907, "num_tokens": 583403939.0, "step": 15369 }, { "epoch": 1.9552219819361403, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7228293418884277, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8705639243125916, "num_tokens": 583441053.0, "step": 15370 }, { "epoch": 1.9553491922147308, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5948233604431152, "learning_rate": 1e-05, "loss": 0.3634, "mean_token_accuracy": 0.8932604789733887, "num_tokens": 583471168.0, "step": 15371 }, { "epoch": 1.9554764024933213, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.723806142807007, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8683077692985535, "num_tokens": 583510604.0, "step": 15372 }, { "epoch": 1.9556036127719119, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.596208095550537, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8749697804450989, "num_tokens": 583546186.0, "step": 15373 }, { "epoch": 1.9557308230505024, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7286181449890137, "learning_rate": 1e-05, "loss": 0.4753, "mean_token_accuracy": 0.8586384057998657, "num_tokens": 583581301.0, "step": 15374 }, { "epoch": 1.955858033329093, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5958292484283447, "learning_rate": 1e-05, "loss": 0.5013, "mean_token_accuracy": 0.8558021783828735, "num_tokens": 583624005.0, "step": 15375 }, { "epoch": 1.9559852436076834, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7264328002929688, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8710437417030334, "num_tokens": 583662738.0, "step": 15376 }, { "epoch": 1.956112453886274, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.5949385166168213, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8807860612869263, "num_tokens": 583700742.0, "step": 15377 }, { "epoch": 1.9562396641648645, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.720702886581421, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8788078427314758, "num_tokens": 583740529.0, "step": 15378 }, { "epoch": 1.956366874443455, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.587010145187378, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.8840799927711487, "num_tokens": 583776866.0, "step": 15379 }, { "epoch": 1.9564940847220456, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7298316955566406, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8770099878311157, "num_tokens": 583815561.0, "step": 15380 }, { "epoch": 1.956621295000636, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.910064697265625e-05, "grad_norm": 2.595073699951172, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.878690242767334, "num_tokens": 583846701.0, "step": 15381 }, { "epoch": 1.9567485052792266, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7219183444976807, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.8845744729042053, "num_tokens": 583885561.0, "step": 15382 }, { "epoch": 1.9568757155578171, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.582315683364868, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8653137683868408, "num_tokens": 583928662.0, "step": 15383 }, { "epoch": 1.9570029258364077, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7318596839904785, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8783587217330933, "num_tokens": 583969097.0, "step": 15384 }, { "epoch": 1.9571301361149982, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.58205509185791, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8673509359359741, "num_tokens": 584008547.0, "step": 15385 }, { "epoch": 1.9572573463935887, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7640762329101562, "learning_rate": 1e-05, "loss": 0.517, "mean_token_accuracy": 0.8540285229682922, "num_tokens": 584046982.0, "step": 15386 }, { "epoch": 1.9573845566721793, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.566129446029663, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8852503299713135, "num_tokens": 584081011.0, "step": 15387 }, { "epoch": 1.9575117669507698, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7788755893707275, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8622207641601562, "num_tokens": 584115550.0, "step": 15388 }, { "epoch": 1.95763897722936, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.6388401985168457, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8791168928146362, "num_tokens": 584153733.0, "step": 15389 }, { "epoch": 1.9577661875079506, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7665669918060303, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8677706718444824, "num_tokens": 584182097.0, "step": 15390 }, { "epoch": 1.9578933977865411, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.933906555175781e-05, "grad_norm": 2.574871063232422, "learning_rate": 1e-05, "loss": 0.3732, "mean_token_accuracy": 0.8932703733444214, "num_tokens": 584215089.0, "step": 15391 }, { "epoch": 1.9580206080651317, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7660627365112305, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8890010714530945, "num_tokens": 584244635.0, "step": 15392 }, { "epoch": 1.9581478183437222, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5698468685150146, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.8784894347190857, "num_tokens": 584280267.0, "step": 15393 }, { "epoch": 1.9582750286223127, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.773221492767334, "learning_rate": 1e-05, "loss": 0.3928, "mean_token_accuracy": 0.8876553773880005, "num_tokens": 584311998.0, "step": 15394 }, { "epoch": 1.958402238900903, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5884430408477783, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8715260624885559, "num_tokens": 584342975.0, "step": 15395 }, { "epoch": 1.9585294491794936, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7720377445220947, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8721940517425537, "num_tokens": 584379087.0, "step": 15396 }, { "epoch": 1.958656659458084, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5761523246765137, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8657407760620117, "num_tokens": 584420126.0, "step": 15397 }, { "epoch": 1.9587838697366746, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.776688575744629, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8746362328529358, "num_tokens": 584457742.0, "step": 15398 }, { "epoch": 1.9589110800152651, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.568389892578125, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8769235610961914, "num_tokens": 584497439.0, "step": 15399 }, { "epoch": 1.9590382902938557, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7887823581695557, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8829368948936462, "num_tokens": 584532922.0, "step": 15400 }, { "epoch": 1.9591655005724462, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5635814666748047, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8838663101196289, "num_tokens": 584574834.0, "step": 15401 }, { "epoch": 1.9592927108510367, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7909843921661377, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8726627230644226, "num_tokens": 584615494.0, "step": 15402 }, { "epoch": 1.9594199211296273, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5855934619903564, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8635333776473999, "num_tokens": 584651494.0, "step": 15403 }, { "epoch": 1.9595471314082178, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.788501739501953, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8639576435089111, "num_tokens": 584691556.0, "step": 15404 }, { "epoch": 1.9596743416868083, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5842783451080322, "learning_rate": 1e-05, "loss": 0.4983, "mean_token_accuracy": 0.8528656959533691, "num_tokens": 584735960.0, "step": 15405 }, { "epoch": 1.9598015519653988, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8002748489379883, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8663384318351746, "num_tokens": 584770808.0, "step": 15406 }, { "epoch": 1.9599287622439894, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5695881843566895, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8752703070640564, "num_tokens": 584811451.0, "step": 15407 }, { "epoch": 1.96005597252258, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8123550415039062, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.871403157711029, "num_tokens": 584845050.0, "step": 15408 }, { "epoch": 1.9601831828011704, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.566364288330078, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8738812208175659, "num_tokens": 584883857.0, "step": 15409 }, { "epoch": 1.960310393079761, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8240160942077637, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8679634928703308, "num_tokens": 584918118.0, "step": 15410 }, { "epoch": 1.9604376033583515, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5710244178771973, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8702808022499084, "num_tokens": 584960789.0, "step": 15411 }, { "epoch": 1.960564813636942, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.802358388900757, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8866704702377319, "num_tokens": 585000631.0, "step": 15412 }, { "epoch": 1.9606920239155323, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.57857346534729, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8751279711723328, "num_tokens": 585036615.0, "step": 15413 }, { "epoch": 1.9608192341941229, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7982981204986572, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8847143650054932, "num_tokens": 585075076.0, "step": 15414 }, { "epoch": 1.9609464444727134, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.594061851501465, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8793196082115173, "num_tokens": 585112215.0, "step": 15415 }, { "epoch": 1.961073654751304, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7960360050201416, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8779534101486206, "num_tokens": 585148920.0, "step": 15416 }, { "epoch": 1.9612008650298944, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.592726945877075, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8651131391525269, "num_tokens": 585191128.0, "step": 15417 }, { "epoch": 1.961328075308485, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7828829288482666, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8678296208381653, "num_tokens": 585234130.0, "step": 15418 }, { "epoch": 1.9614552855870753, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6053593158721924, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8688188791275024, "num_tokens": 585267089.0, "step": 15419 }, { "epoch": 1.9615824958656658, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.763362169265747, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8761385679244995, "num_tokens": 585310508.0, "step": 15420 }, { "epoch": 1.9617097061442563, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5999083518981934, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8771413564682007, "num_tokens": 585346579.0, "step": 15421 }, { "epoch": 1.9618369164228469, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7664878368377686, "learning_rate": 1e-05, "loss": 0.4892, "mean_token_accuracy": 0.8668836355209351, "num_tokens": 585383914.0, "step": 15422 }, { "epoch": 1.9619641267014374, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5969998836517334, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8639758825302124, "num_tokens": 585419322.0, "step": 15423 }, { "epoch": 1.962091336980028, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7437477111816406, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8786200284957886, "num_tokens": 585459081.0, "step": 15424 }, { "epoch": 1.9622185472586184, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6015408039093018, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8657659888267517, "num_tokens": 585497116.0, "step": 15425 }, { "epoch": 1.962345757537209, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.743511438369751, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.863646388053894, "num_tokens": 585534704.0, "step": 15426 }, { "epoch": 1.9624729678157995, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5921730995178223, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8804773092269897, "num_tokens": 585575806.0, "step": 15427 }, { "epoch": 1.96260017809439, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.742621421813965, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8788304328918457, "num_tokens": 585612585.0, "step": 15428 }, { "epoch": 1.9627273883729806, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5902111530303955, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8781867027282715, "num_tokens": 585647414.0, "step": 15429 }, { "epoch": 1.962854598651571, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7299225330352783, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8687241077423096, "num_tokens": 585693802.0, "step": 15430 }, { "epoch": 1.9629818089301616, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.596038579940796, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8737545609474182, "num_tokens": 585727955.0, "step": 15431 }, { "epoch": 1.9631090192087521, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.726642370223999, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8614712953567505, "num_tokens": 585769784.0, "step": 15432 }, { "epoch": 1.9632362294873427, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.584735870361328, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8791131973266602, "num_tokens": 585809989.0, "step": 15433 }, { "epoch": 1.9633634397659332, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.728459119796753, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8740613460540771, "num_tokens": 585851054.0, "step": 15434 }, { "epoch": 1.9634906500445237, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5960185527801514, "learning_rate": 1e-05, "loss": 0.4744, "mean_token_accuracy": 0.8618724942207336, "num_tokens": 585886638.0, "step": 15435 }, { "epoch": 1.9636178603231143, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.728902578353882, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8634399175643921, "num_tokens": 585925572.0, "step": 15436 }, { "epoch": 1.9637450706017048, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5881333351135254, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8724055886268616, "num_tokens": 585961939.0, "step": 15437 }, { "epoch": 1.963872280880295, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7380709648132324, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8694612383842468, "num_tokens": 585996804.0, "step": 15438 }, { "epoch": 1.9639994911588856, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5673575401306152, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8813243508338928, "num_tokens": 586040347.0, "step": 15439 }, { "epoch": 1.9641267014374761, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7406811714172363, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8795558214187622, "num_tokens": 586078495.0, "step": 15440 }, { "epoch": 1.9642539117160667, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5822134017944336, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8801165819168091, "num_tokens": 586114743.0, "step": 15441 }, { "epoch": 1.9643811219946572, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7568068504333496, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8802517056465149, "num_tokens": 586150462.0, "step": 15442 }, { "epoch": 1.9645083322732477, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.597935438156128, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8715753555297852, "num_tokens": 586182860.0, "step": 15443 }, { "epoch": 1.964635542551838, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.73512864112854, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8686767816543579, "num_tokens": 586226438.0, "step": 15444 }, { "epoch": 1.9647627528304286, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5819571018218994, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8744058609008789, "num_tokens": 586267022.0, "step": 15445 }, { "epoch": 1.964889963109019, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.748419761657715, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8795849084854126, "num_tokens": 586299415.0, "step": 15446 }, { "epoch": 1.9650171733876096, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.581336259841919, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8736650943756104, "num_tokens": 586337351.0, "step": 15447 }, { "epoch": 1.9651443836662001, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.751009702682495, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8635392189025879, "num_tokens": 586378311.0, "step": 15448 }, { "epoch": 1.9652715939447907, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5799965858459473, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8740403652191162, "num_tokens": 586414137.0, "step": 15449 }, { "epoch": 1.9653988042233812, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.763051986694336, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.866999089717865, "num_tokens": 586451536.0, "step": 15450 }, { "epoch": 1.9655260145019717, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5876874923706055, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8692508339881897, "num_tokens": 586484643.0, "step": 15451 }, { "epoch": 1.9656532247805623, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7623941898345947, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8625967502593994, "num_tokens": 586523926.0, "step": 15452 }, { "epoch": 1.9657804350591528, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.566041946411133, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8754822015762329, "num_tokens": 586564999.0, "step": 15453 }, { "epoch": 1.9659076453377433, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.768144369125366, "learning_rate": 1e-05, "loss": 0.4799, "mean_token_accuracy": 0.8651981353759766, "num_tokens": 586603455.0, "step": 15454 }, { "epoch": 1.9660348556163338, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.564194917678833, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8683518767356873, "num_tokens": 586646345.0, "step": 15455 }, { "epoch": 1.9661620658949244, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7786622047424316, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8686646819114685, "num_tokens": 586688065.0, "step": 15456 }, { "epoch": 1.966289276173515, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5711307525634766, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8785573840141296, "num_tokens": 586722631.0, "step": 15457 }, { "epoch": 1.9664164864521054, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7808191776275635, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.8900673389434814, "num_tokens": 586762630.0, "step": 15458 }, { "epoch": 1.966543696730696, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.585007905960083, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8727455139160156, "num_tokens": 586799514.0, "step": 15459 }, { "epoch": 1.9666709070092865, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.76531720161438, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8830922842025757, "num_tokens": 586839507.0, "step": 15460 }, { "epoch": 1.966798117287877, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5815279483795166, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8631690144538879, "num_tokens": 586879263.0, "step": 15461 }, { "epoch": 1.9669253275664673, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7696826457977295, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8725148439407349, "num_tokens": 586920297.0, "step": 15462 }, { "epoch": 1.9670525378450578, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.580307722091675, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8614045977592468, "num_tokens": 586961171.0, "step": 15463 }, { "epoch": 1.9671797481236484, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7705233097076416, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8740799427032471, "num_tokens": 587003808.0, "step": 15464 }, { "epoch": 1.967306958402239, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.586082935333252, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8617488145828247, "num_tokens": 587042056.0, "step": 15465 }, { "epoch": 1.9674341686808294, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.771599769592285, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8578941822052002, "num_tokens": 587083009.0, "step": 15466 }, { "epoch": 1.96756137895942, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.586156129837036, "learning_rate": 1e-05, "loss": 0.5021, "mean_token_accuracy": 0.8522619009017944, "num_tokens": 587125129.0, "step": 15467 }, { "epoch": 1.9676885892380103, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.777385950088501, "learning_rate": 1e-05, "loss": 0.4887, "mean_token_accuracy": 0.8575221300125122, "num_tokens": 587163314.0, "step": 15468 }, { "epoch": 1.9678157995166008, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.563028335571289, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8699485659599304, "num_tokens": 587208275.0, "step": 15469 }, { "epoch": 1.9679430097951913, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7990081310272217, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8705862164497375, "num_tokens": 587242885.0, "step": 15470 }, { "epoch": 1.9680702200737819, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6061525344848633, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8643310070037842, "num_tokens": 587287787.0, "step": 15471 }, { "epoch": 1.9681974303523724, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7663493156433105, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8849852085113525, "num_tokens": 587325445.0, "step": 15472 }, { "epoch": 1.968324640630963, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5721523761749268, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8734244108200073, "num_tokens": 587366691.0, "step": 15473 }, { "epoch": 1.9684518509095534, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7807934284210205, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8623425960540771, "num_tokens": 587406310.0, "step": 15474 }, { "epoch": 1.968579061188144, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5839321613311768, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8697855472564697, "num_tokens": 587446654.0, "step": 15475 }, { "epoch": 1.9687062714667345, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7839133739471436, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8678401708602905, "num_tokens": 587481089.0, "step": 15476 }, { "epoch": 1.968833481745325, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5793278217315674, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8716132640838623, "num_tokens": 587521293.0, "step": 15477 }, { "epoch": 1.9689606920239155, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.786837100982666, "learning_rate": 1e-05, "loss": 0.4673, "mean_token_accuracy": 0.8646537065505981, "num_tokens": 587564193.0, "step": 15478 }, { "epoch": 1.969087902302506, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5853683948516846, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8638163208961487, "num_tokens": 587603487.0, "step": 15479 }, { "epoch": 1.9692151125810966, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7920548915863037, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8612515926361084, "num_tokens": 587641131.0, "step": 15480 }, { "epoch": 1.9693423228596871, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.590242624282837, "learning_rate": 1e-05, "loss": 0.4987, "mean_token_accuracy": 0.8533663749694824, "num_tokens": 587680012.0, "step": 15481 }, { "epoch": 1.9694695331382777, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7719836235046387, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8757174015045166, "num_tokens": 587721904.0, "step": 15482 }, { "epoch": 1.9695967434168682, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6078009605407715, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8755596876144409, "num_tokens": 587752080.0, "step": 15483 }, { "epoch": 1.9697239536954587, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.758528470993042, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8824888467788696, "num_tokens": 587787838.0, "step": 15484 }, { "epoch": 1.9698511639740492, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6035966873168945, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8725352883338928, "num_tokens": 587826370.0, "step": 15485 }, { "epoch": 1.9699783742526398, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.739314556121826, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.8879064321517944, "num_tokens": 587866716.0, "step": 15486 }, { "epoch": 1.97010558453123, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6099135875701904, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8758399486541748, "num_tokens": 587902766.0, "step": 15487 }, { "epoch": 1.9702327948098206, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.731825828552246, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8603941798210144, "num_tokens": 587944584.0, "step": 15488 }, { "epoch": 1.9703600050884111, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.669173240661621, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8607085943222046, "num_tokens": 587983482.0, "step": 15489 }, { "epoch": 1.9704872153670017, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.696768045425415, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8798853158950806, "num_tokens": 588019621.0, "step": 15490 }, { "epoch": 1.9706144256455922, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.627439260482788, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8782563805580139, "num_tokens": 588056845.0, "step": 15491 }, { "epoch": 1.9707416359241827, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.673158645629883, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8790233135223389, "num_tokens": 588093731.0, "step": 15492 }, { "epoch": 1.970868846202773, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.628774642944336, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.886906623840332, "num_tokens": 588128908.0, "step": 15493 }, { "epoch": 1.9709960564813636, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6769659519195557, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8591443300247192, "num_tokens": 588167729.0, "step": 15494 }, { "epoch": 1.971123266759954, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6540567874908447, "learning_rate": 1e-05, "loss": 0.4898, "mean_token_accuracy": 0.8551927208900452, "num_tokens": 588200337.0, "step": 15495 }, { "epoch": 1.9712504770385446, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6534006595611572, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8705023527145386, "num_tokens": 588240736.0, "step": 15496 }, { "epoch": 1.9713776873171351, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.656489849090576, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8623338937759399, "num_tokens": 588275187.0, "step": 15497 }, { "epoch": 1.9715048975957257, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6611156463623047, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8669962286949158, "num_tokens": 588308800.0, "step": 15498 }, { "epoch": 1.9716321078743162, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6332404613494873, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8855706453323364, "num_tokens": 588347688.0, "step": 15499 }, { "epoch": 1.9717593181529067, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6808645725250244, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8730063438415527, "num_tokens": 588380791.0, "step": 15500 }, { "epoch": 1.9718865284314973, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.643980026245117, "learning_rate": 1e-05, "loss": 0.4714, "mean_token_accuracy": 0.8590356707572937, "num_tokens": 588413667.0, "step": 15501 }, { "epoch": 1.9720137387100878, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.650611162185669, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.87278151512146, "num_tokens": 588453884.0, "step": 15502 }, { "epoch": 1.9721409489886783, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6448605060577393, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8753865957260132, "num_tokens": 588488895.0, "step": 15503 }, { "epoch": 1.9722681592672688, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7930681705474854, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8770419359207153, "num_tokens": 588526807.0, "step": 15504 }, { "epoch": 1.9723953695458594, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.589165449142456, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8762872219085693, "num_tokens": 588562272.0, "step": 15505 }, { "epoch": 1.97252257982445, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7375845909118652, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8685586452484131, "num_tokens": 588598298.0, "step": 15506 }, { "epoch": 1.9726497901030404, "ewc_loss": 0.0673828125, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.56354022026062, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8729598522186279, "num_tokens": 588637833.0, "step": 15507 }, { "epoch": 1.972777000381631, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7462968826293945, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8694347143173218, "num_tokens": 588674933.0, "step": 15508 }, { "epoch": 1.9729042106602215, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5616695880889893, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8739039301872253, "num_tokens": 588711899.0, "step": 15509 }, { "epoch": 1.973031420938812, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.764636516571045, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8714524507522583, "num_tokens": 588749702.0, "step": 15510 }, { "epoch": 1.9731586312174023, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5650815963745117, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8779494166374207, "num_tokens": 588787066.0, "step": 15511 }, { "epoch": 1.9732858414959928, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7834441661834717, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8740370273590088, "num_tokens": 588824062.0, "step": 15512 }, { "epoch": 1.9734130517745834, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.570992946624756, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.870908260345459, "num_tokens": 588864858.0, "step": 15513 }, { "epoch": 1.973540262053174, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.792097568511963, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8689472079277039, "num_tokens": 588900219.0, "step": 15514 }, { "epoch": 1.9736674723317644, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5795505046844482, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8678171634674072, "num_tokens": 588937256.0, "step": 15515 }, { "epoch": 1.973794682610355, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7940869331359863, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.878719687461853, "num_tokens": 588971263.0, "step": 15516 }, { "epoch": 1.9739218928889453, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.578517436981201, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8679688572883606, "num_tokens": 589010710.0, "step": 15517 }, { "epoch": 1.9740491031675358, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002803802490234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.806164264678955, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.877711832523346, "num_tokens": 589043940.0, "step": 15518 }, { "epoch": 1.9741763134461263, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.579577684402466, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8686003684997559, "num_tokens": 589082187.0, "step": 15519 }, { "epoch": 1.9743035237247168, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7857186794281006, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8791227340698242, "num_tokens": 589126870.0, "step": 15520 }, { "epoch": 1.9744307340033074, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.56786847114563, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8660942316055298, "num_tokens": 589168378.0, "step": 15521 }, { "epoch": 1.974557944281898, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.802436590194702, "learning_rate": 1e-05, "loss": 0.4995, "mean_token_accuracy": 0.8567225933074951, "num_tokens": 589213070.0, "step": 15522 }, { "epoch": 1.9746851545604884, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5819177627563477, "learning_rate": 1e-05, "loss": 0.4608, "mean_token_accuracy": 0.8696468472480774, "num_tokens": 589253008.0, "step": 15523 }, { "epoch": 1.974812364839079, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8037781715393066, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8657368421554565, "num_tokens": 589291786.0, "step": 15524 }, { "epoch": 1.9749395751176695, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5742225646972656, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8737819194793701, "num_tokens": 589332045.0, "step": 15525 }, { "epoch": 1.97506678539626, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.810546398162842, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8504935503005981, "num_tokens": 589374983.0, "step": 15526 }, { "epoch": 1.9751939956748505, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.552340269088745, "learning_rate": 1e-05, "loss": 0.3728, "mean_token_accuracy": 0.8957833647727966, "num_tokens": 589419786.0, "step": 15527 }, { "epoch": 1.975321205953441, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.824384927749634, "learning_rate": 1e-05, "loss": 0.3914, "mean_token_accuracy": 0.8901246190071106, "num_tokens": 589454365.0, "step": 15528 }, { "epoch": 1.9754484162320316, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.590589761734009, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8760889172554016, "num_tokens": 589486829.0, "step": 15529 }, { "epoch": 1.9755756265106221, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8145883083343506, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8738586902618408, "num_tokens": 589527500.0, "step": 15530 }, { "epoch": 1.9757028367892127, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5834274291992188, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.8845399618148804, "num_tokens": 589570064.0, "step": 15531 }, { "epoch": 1.9758300470678032, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.830990791320801, "learning_rate": 1e-05, "loss": 0.4934, "mean_token_accuracy": 0.8559103012084961, "num_tokens": 589605327.0, "step": 15532 }, { "epoch": 1.9759572573463937, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5841064453125, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8833863735198975, "num_tokens": 589640886.0, "step": 15533 }, { "epoch": 1.9760844676249842, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8168745040893555, "learning_rate": 1e-05, "loss": 0.4955, "mean_token_accuracy": 0.8586360812187195, "num_tokens": 589681178.0, "step": 15534 }, { "epoch": 1.9762116779035748, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.584698438644409, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8845070600509644, "num_tokens": 589719073.0, "step": 15535 }, { "epoch": 1.976338888182165, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.813366651535034, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8628594875335693, "num_tokens": 589760528.0, "step": 15536 }, { "epoch": 1.9764660984607556, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6094322204589844, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8714420199394226, "num_tokens": 589795946.0, "step": 15537 }, { "epoch": 1.9765933087393461, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.810004472732544, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8704366683959961, "num_tokens": 589829139.0, "step": 15538 }, { "epoch": 1.9767205190179367, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6010849475860596, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8846412301063538, "num_tokens": 589862916.0, "step": 15539 }, { "epoch": 1.9768477292965272, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7909915447235107, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8778771758079529, "num_tokens": 589900230.0, "step": 15540 }, { "epoch": 1.9769749395751177, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.628652572631836, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8716230392456055, "num_tokens": 589938737.0, "step": 15541 }, { "epoch": 1.977102149853708, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7569572925567627, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8781383037567139, "num_tokens": 589977989.0, "step": 15542 }, { "epoch": 1.9772293601322986, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.596353054046631, "learning_rate": 1e-05, "loss": 0.374, "mean_token_accuracy": 0.8887049555778503, "num_tokens": 590015547.0, "step": 15543 }, { "epoch": 1.977356570410889, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7543535232543945, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.873150646686554, "num_tokens": 590055257.0, "step": 15544 }, { "epoch": 1.9774837806894796, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.615750312805176, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8748606443405151, "num_tokens": 590087911.0, "step": 15545 }, { "epoch": 1.9776109909680701, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.733569622039795, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8688104748725891, "num_tokens": 590128889.0, "step": 15546 }, { "epoch": 1.9777382012466607, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.616199016571045, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8593162894248962, "num_tokens": 590168078.0, "step": 15547 }, { "epoch": 1.9778654115252512, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.729731798171997, "learning_rate": 1e-05, "loss": 0.498, "mean_token_accuracy": 0.8545072078704834, "num_tokens": 590211361.0, "step": 15548 }, { "epoch": 1.9779926218038417, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.91557240486145, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8629138469696045, "num_tokens": 590249922.0, "step": 15549 }, { "epoch": 1.9781198320824323, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6001315116882324, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8734660744667053, "num_tokens": 590288284.0, "step": 15550 }, { "epoch": 1.9782470423610228, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.745511531829834, "learning_rate": 1e-05, "loss": 0.4754, "mean_token_accuracy": 0.8593888282775879, "num_tokens": 590324373.0, "step": 15551 }, { "epoch": 1.9783742526396133, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.558209180831909, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8740589618682861, "num_tokens": 590365374.0, "step": 15552 }, { "epoch": 1.9785014629182038, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.788140058517456, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8700301051139832, "num_tokens": 590402712.0, "step": 15553 }, { "epoch": 1.9786286731967944, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5988080501556396, "learning_rate": 1e-05, "loss": 0.5292, "mean_token_accuracy": 0.8542957305908203, "num_tokens": 590438434.0, "step": 15554 }, { "epoch": 1.978755883475385, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7702181339263916, "learning_rate": 1e-05, "loss": 0.4808, "mean_token_accuracy": 0.8627287745475769, "num_tokens": 590482619.0, "step": 15555 }, { "epoch": 1.9788830937539754, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.581930160522461, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8559978008270264, "num_tokens": 590527828.0, "step": 15556 }, { "epoch": 1.979010304032566, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.794255018234253, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8749829530715942, "num_tokens": 590561758.0, "step": 15557 }, { "epoch": 1.9791375143111565, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5833606719970703, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8845405578613281, "num_tokens": 590598323.0, "step": 15558 }, { "epoch": 1.979264724589747, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.774420738220215, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8669690489768982, "num_tokens": 590642469.0, "step": 15559 }, { "epoch": 1.9793919348683373, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.578162670135498, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8668777942657471, "num_tokens": 590683603.0, "step": 15560 }, { "epoch": 1.9795191451469278, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.814713716506958, "learning_rate": 1e-05, "loss": 0.4014, "mean_token_accuracy": 0.8815062046051025, "num_tokens": 590710136.0, "step": 15561 }, { "epoch": 1.9796463554255184, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5797438621520996, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8695463538169861, "num_tokens": 590743104.0, "step": 15562 }, { "epoch": 1.979773565704109, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8086204528808594, "learning_rate": 1e-05, "loss": 0.4583, "mean_token_accuracy": 0.8662419319152832, "num_tokens": 590781851.0, "step": 15563 }, { "epoch": 1.9799007759826994, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.586543321609497, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8611430525779724, "num_tokens": 590819865.0, "step": 15564 }, { "epoch": 1.98002798626129, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.803253173828125, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8805326223373413, "num_tokens": 590856478.0, "step": 15565 }, { "epoch": 1.9801551965398803, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5911967754364014, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.861160397529602, "num_tokens": 590895765.0, "step": 15566 }, { "epoch": 1.9802824068184708, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7985777854919434, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8756226897239685, "num_tokens": 590931238.0, "step": 15567 }, { "epoch": 1.9804096170970613, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5905356407165527, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8753563165664673, "num_tokens": 590969875.0, "step": 15568 }, { "epoch": 1.9805368273756518, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.795807123184204, "learning_rate": 1e-05, "loss": 0.4818, "mean_token_accuracy": 0.8581490516662598, "num_tokens": 591007964.0, "step": 15569 }, { "epoch": 1.9806640376542424, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.579766273498535, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.877395510673523, "num_tokens": 591049339.0, "step": 15570 }, { "epoch": 1.980791247932833, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.797760009765625, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8808428645133972, "num_tokens": 591086429.0, "step": 15571 }, { "epoch": 1.9809184582114234, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.586150646209717, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8758187294006348, "num_tokens": 591128252.0, "step": 15572 }, { "epoch": 1.981045668490014, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.791288137435913, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8844242691993713, "num_tokens": 591165049.0, "step": 15573 }, { "epoch": 1.9811728787686045, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6067659854888916, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8828614950180054, "num_tokens": 591196805.0, "step": 15574 }, { "epoch": 1.981300089047195, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.767554759979248, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8747882843017578, "num_tokens": 591242897.0, "step": 15575 }, { "epoch": 1.9814272993257855, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.594973564147949, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8695735931396484, "num_tokens": 591280684.0, "step": 15576 }, { "epoch": 1.981554509604376, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.776599645614624, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8761111497879028, "num_tokens": 591315264.0, "step": 15577 }, { "epoch": 1.9816817198829666, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5976767539978027, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8693133592605591, "num_tokens": 591358028.0, "step": 15578 }, { "epoch": 1.9818089301615571, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7667109966278076, "learning_rate": 1e-05, "loss": 0.5106, "mean_token_accuracy": 0.8530141711235046, "num_tokens": 591404272.0, "step": 15579 }, { "epoch": 1.9819361404401477, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.588757038116455, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.884963870048523, "num_tokens": 591439896.0, "step": 15580 }, { "epoch": 1.9820633507187382, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.766737461090088, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8771036863327026, "num_tokens": 591477318.0, "step": 15581 }, { "epoch": 1.9821905609973287, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5902440547943115, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8678861856460571, "num_tokens": 591518600.0, "step": 15582 }, { "epoch": 1.9823177712759192, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7964868545532227, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.86881023645401, "num_tokens": 591547414.0, "step": 15583 }, { "epoch": 1.9824449815545098, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.574582815170288, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.8609903454780579, "num_tokens": 591594270.0, "step": 15584 }, { "epoch": 1.9825721918331, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.787949323654175, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8747658729553223, "num_tokens": 591638252.0, "step": 15585 }, { "epoch": 1.9826994021116906, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5907623767852783, "learning_rate": 1e-05, "loss": 0.5108, "mean_token_accuracy": 0.8518451452255249, "num_tokens": 591680268.0, "step": 15586 }, { "epoch": 1.9828266123902811, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 3.8778905868530273, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8753851652145386, "num_tokens": 591717724.0, "step": 15587 }, { "epoch": 1.9829538226688717, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 5.316734313964844e-05, "grad_norm": 2.8824105262756348, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8865047097206116, "num_tokens": 591754562.0, "step": 15588 }, { "epoch": 1.9830810329474622, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.902503728866577, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8765126466751099, "num_tokens": 591794838.0, "step": 15589 }, { "epoch": 1.9832082432260527, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.678657293319702, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8528295755386353, "num_tokens": 591834771.0, "step": 15590 }, { "epoch": 1.983335453504643, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 2.9357080459594727, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8604079484939575, "num_tokens": 591872764.0, "step": 15591 }, { "epoch": 1.9834626637832335, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.6802470684051514, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8672037720680237, "num_tokens": 591907035.0, "step": 15592 }, { "epoch": 1.983589874061824, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.8890111446380615, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8769494295120239, "num_tokens": 591940797.0, "step": 15593 }, { "epoch": 1.9837170843404146, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6565704345703125, "learning_rate": 1e-05, "loss": 0.3909, "mean_token_accuracy": 0.8871318101882935, "num_tokens": 591976872.0, "step": 15594 }, { "epoch": 1.9838442946190051, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.935264587402344e-05, "grad_norm": 2.869694471359253, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8579100370407104, "num_tokens": 592015933.0, "step": 15595 }, { "epoch": 1.9839715048975957, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6537954807281494, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.8793723583221436, "num_tokens": 592046796.0, "step": 15596 }, { "epoch": 1.9840987151761862, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.834733724594116, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8713580369949341, "num_tokens": 592092841.0, "step": 15597 }, { "epoch": 1.9842259254547767, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6458792686462402, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8548338413238525, "num_tokens": 592134118.0, "step": 15598 }, { "epoch": 1.9843531357333672, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.820096731185913, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8663248419761658, "num_tokens": 592167292.0, "step": 15599 }, { "epoch": 1.9844803460119578, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.624321699142456, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8840692639350891, "num_tokens": 592205616.0, "step": 15600 }, { "epoch": 1.9846075562905483, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.808145761489868, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8782822489738464, "num_tokens": 592246650.0, "step": 15601 }, { "epoch": 1.9847347665691388, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.614132881164551, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8761571049690247, "num_tokens": 592284729.0, "step": 15602 }, { "epoch": 1.9848619768477294, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.80380916595459, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.874679446220398, "num_tokens": 592318423.0, "step": 15603 }, { "epoch": 1.9849891871263199, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.687119960784912, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8703001737594604, "num_tokens": 592351304.0, "step": 15604 }, { "epoch": 1.9851163974049104, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7486350536346436, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8742393255233765, "num_tokens": 592388353.0, "step": 15605 }, { "epoch": 1.985243607683501, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.621643304824829, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8780824542045593, "num_tokens": 592427651.0, "step": 15606 }, { "epoch": 1.9853708179620915, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.738917589187622, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8725063800811768, "num_tokens": 592463638.0, "step": 15607 }, { "epoch": 1.985498028240682, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6243503093719482, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8624184727668762, "num_tokens": 592501508.0, "step": 15608 }, { "epoch": 1.9856252385192723, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.726468563079834, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8794071674346924, "num_tokens": 592538418.0, "step": 15609 }, { "epoch": 1.9857524487978628, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.60628080368042, "learning_rate": 1e-05, "loss": 0.3921, "mean_token_accuracy": 0.88541579246521, "num_tokens": 592573552.0, "step": 15610 }, { "epoch": 1.9858796590764534, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.730283260345459, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8742351531982422, "num_tokens": 592607212.0, "step": 15611 }, { "epoch": 1.986006869355044, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.600254774093628, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.866654098033905, "num_tokens": 592649910.0, "step": 15612 }, { "epoch": 1.9861340796336344, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7252039909362793, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8733350038528442, "num_tokens": 592689555.0, "step": 15613 }, { "epoch": 1.986261289912225, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.591827392578125, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8758405447006226, "num_tokens": 592731921.0, "step": 15614 }, { "epoch": 1.9863885001908153, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.759896755218506, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8763314485549927, "num_tokens": 592763631.0, "step": 15615 }, { "epoch": 1.9865157104694058, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.582777500152588, "learning_rate": 1e-05, "loss": 0.3789, "mean_token_accuracy": 0.8908356428146362, "num_tokens": 592796961.0, "step": 15616 }, { "epoch": 1.9866429207479963, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.749079465866089, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.879151463508606, "num_tokens": 592831380.0, "step": 15617 }, { "epoch": 1.9867701310265868, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.577150344848633, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.864874005317688, "num_tokens": 592873606.0, "step": 15618 }, { "epoch": 1.9868973413051774, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.753725290298462, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8645454049110413, "num_tokens": 592919311.0, "step": 15619 }, { "epoch": 1.987024551583768, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5902926921844482, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8500643968582153, "num_tokens": 592961818.0, "step": 15620 }, { "epoch": 1.9871517618623584, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7605650424957275, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.87053382396698, "num_tokens": 593003320.0, "step": 15621 }, { "epoch": 1.987278972140949, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5640790462493896, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8617055416107178, "num_tokens": 593050887.0, "step": 15622 }, { "epoch": 1.9874061824195395, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7801105976104736, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8625180125236511, "num_tokens": 593086296.0, "step": 15623 }, { "epoch": 1.98753339269813, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5866124629974365, "learning_rate": 1e-05, "loss": 0.4937, "mean_token_accuracy": 0.8549654483795166, "num_tokens": 593122135.0, "step": 15624 }, { "epoch": 1.9876606029767205, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7783994674682617, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8629257678985596, "num_tokens": 593158845.0, "step": 15625 }, { "epoch": 1.987787813255311, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5634031295776367, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8781850337982178, "num_tokens": 593199386.0, "step": 15626 }, { "epoch": 1.9879150235339016, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.78694486618042, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8813275694847107, "num_tokens": 593237061.0, "step": 15627 }, { "epoch": 1.9880422338124921, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5805916786193848, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8721700310707092, "num_tokens": 593275220.0, "step": 15628 }, { "epoch": 1.9881694440910826, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7924153804779053, "learning_rate": 1e-05, "loss": 0.5033, "mean_token_accuracy": 0.8514389991760254, "num_tokens": 593314576.0, "step": 15629 }, { "epoch": 1.9882966543696732, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5625810623168945, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8795571327209473, "num_tokens": 593362905.0, "step": 15630 }, { "epoch": 1.9884238646482637, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8013110160827637, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8795977234840393, "num_tokens": 593402486.0, "step": 15631 }, { "epoch": 1.9885510749268542, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5731701850891113, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8821744918823242, "num_tokens": 593440555.0, "step": 15632 }, { "epoch": 1.9886782852054448, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.811335563659668, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8713266849517822, "num_tokens": 593480488.0, "step": 15633 }, { "epoch": 1.988805495484035, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.57317852973938, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8775057792663574, "num_tokens": 593518615.0, "step": 15634 }, { "epoch": 1.9889327057626256, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.814113140106201, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8780899047851562, "num_tokens": 593556990.0, "step": 15635 }, { "epoch": 1.9890599160412161, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5785677433013916, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8629230260848999, "num_tokens": 593599438.0, "step": 15636 }, { "epoch": 1.9891871263198067, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.824340343475342, "learning_rate": 1e-05, "loss": 0.4833, "mean_token_accuracy": 0.8589398264884949, "num_tokens": 593635991.0, "step": 15637 }, { "epoch": 1.9893143365983972, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.594484329223633, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8725232481956482, "num_tokens": 593676600.0, "step": 15638 }, { "epoch": 1.9894415468769877, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8147764205932617, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8818956017494202, "num_tokens": 593709069.0, "step": 15639 }, { "epoch": 1.989568757155578, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.589684247970581, "learning_rate": 1e-05, "loss": 0.3744, "mean_token_accuracy": 0.8969203233718872, "num_tokens": 593740361.0, "step": 15640 }, { "epoch": 1.9896959674341685, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.794854164123535, "learning_rate": 1e-05, "loss": 0.3826, "mean_token_accuracy": 0.8869901895523071, "num_tokens": 593775424.0, "step": 15641 }, { "epoch": 1.989823177712759, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5999488830566406, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.8649735450744629, "num_tokens": 593816562.0, "step": 15642 }, { "epoch": 1.9899503879913496, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.797342538833618, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8735955953598022, "num_tokens": 593853272.0, "step": 15643 }, { "epoch": 1.9900775982699401, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.604454278945923, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8778245449066162, "num_tokens": 593885308.0, "step": 15644 }, { "epoch": 1.9902048085485307, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.8042490482330322, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.8844293355941772, "num_tokens": 593918150.0, "step": 15645 }, { "epoch": 1.9903320188271212, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5969457626342773, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8669470548629761, "num_tokens": 593961947.0, "step": 15646 }, { "epoch": 1.9904592291057117, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.790520668029785, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8573797941207886, "num_tokens": 594000659.0, "step": 15647 }, { "epoch": 1.9905864393843022, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5957114696502686, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8724762201309204, "num_tokens": 594040033.0, "step": 15648 }, { "epoch": 1.9907136496628928, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7794289588928223, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.877913236618042, "num_tokens": 594080534.0, "step": 15649 }, { "epoch": 1.9908408599414833, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5979628562927246, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8679792881011963, "num_tokens": 594118015.0, "step": 15650 }, { "epoch": 1.9909680702200738, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7725906372070312, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8756304383277893, "num_tokens": 594153839.0, "step": 15651 }, { "epoch": 1.9910952804986644, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.582841396331787, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8828918933868408, "num_tokens": 594195672.0, "step": 15652 }, { "epoch": 1.9912224907772549, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7812864780426025, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8668644428253174, "num_tokens": 594239178.0, "step": 15653 }, { "epoch": 1.9913497010558454, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5983786582946777, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8747395277023315, "num_tokens": 594274938.0, "step": 15654 }, { "epoch": 1.991476911334436, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.765601634979248, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8824014663696289, "num_tokens": 594310632.0, "step": 15655 }, { "epoch": 1.9916041216130265, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.607910633087158, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8664100170135498, "num_tokens": 594344308.0, "step": 15656 }, { "epoch": 1.991731331891617, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.750342845916748, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8701472878456116, "num_tokens": 594385287.0, "step": 15657 }, { "epoch": 1.9918585421702073, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.597837448120117, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8820117712020874, "num_tokens": 594422521.0, "step": 15658 }, { "epoch": 1.9919857524487978, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.739323616027832, "learning_rate": 1e-05, "loss": 0.4942, "mean_token_accuracy": 0.85793536901474, "num_tokens": 594466837.0, "step": 15659 }, { "epoch": 1.9921129627273884, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6144461631774902, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8719161748886108, "num_tokens": 594501283.0, "step": 15660 }, { "epoch": 1.9922401730059789, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.746640920639038, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.859237551689148, "num_tokens": 594544627.0, "step": 15661 }, { "epoch": 1.9923673832845694, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.00028228759765625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5991337299346924, "learning_rate": 1e-05, "loss": 0.5044, "mean_token_accuracy": 0.8511741161346436, "num_tokens": 594586363.0, "step": 15662 }, { "epoch": 1.99249459356316, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7398860454559326, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8696079850196838, "num_tokens": 594624950.0, "step": 15663 }, { "epoch": 1.9926218038417502, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.60148024559021, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8574622869491577, "num_tokens": 594665103.0, "step": 15664 }, { "epoch": 1.9927490141203408, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7424986362457275, "learning_rate": 1e-05, "loss": 0.4021, "mean_token_accuracy": 0.8822064995765686, "num_tokens": 594699506.0, "step": 15665 }, { "epoch": 1.9928762243989313, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.588200569152832, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8675101399421692, "num_tokens": 594741771.0, "step": 15666 }, { "epoch": 1.9930034346775218, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7335286140441895, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8804298639297485, "num_tokens": 594784705.0, "step": 15667 }, { "epoch": 1.9931306449561124, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.58015775680542, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8773992657661438, "num_tokens": 594827800.0, "step": 15668 }, { "epoch": 1.993257855234703, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.754430055618286, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8693377375602722, "num_tokens": 594865150.0, "step": 15669 }, { "epoch": 1.9933850655132934, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.5987608432769775, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8586985468864441, "num_tokens": 594903440.0, "step": 15670 }, { "epoch": 1.993512275791884, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.753741502761841, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8760443925857544, "num_tokens": 594936761.0, "step": 15671 }, { "epoch": 1.9936394860704745, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.601750612258911, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8593117594718933, "num_tokens": 594972298.0, "step": 15672 }, { "epoch": 1.993766696349065, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.746096134185791, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8727251291275024, "num_tokens": 595009530.0, "step": 15673 }, { "epoch": 1.9938939066276555, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5841116905212402, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8748149871826172, "num_tokens": 595048147.0, "step": 15674 }, { "epoch": 1.994021116906246, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.749985933303833, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8839667439460754, "num_tokens": 595083580.0, "step": 15675 }, { "epoch": 1.9941483271848366, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5937204360961914, "learning_rate": 1e-05, "loss": 0.5054, "mean_token_accuracy": 0.8513950109481812, "num_tokens": 595122989.0, "step": 15676 }, { "epoch": 1.9942755374634271, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.736845016479492, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8677706718444824, "num_tokens": 595165378.0, "step": 15677 }, { "epoch": 1.9944027477420176, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.604434013366699, "learning_rate": 1e-05, "loss": 0.5153, "mean_token_accuracy": 0.8500502109527588, "num_tokens": 595203802.0, "step": 15678 }, { "epoch": 1.9945299580206082, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.735142946243286, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.866836667060852, "num_tokens": 595242056.0, "step": 15679 }, { "epoch": 1.9946571682991987, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6094324588775635, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.8568309545516968, "num_tokens": 595276381.0, "step": 15680 }, { "epoch": 1.9947843785777892, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.712229013442993, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8720080852508545, "num_tokens": 595318551.0, "step": 15681 }, { "epoch": 1.9949115888563798, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6072123050689697, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8624573349952698, "num_tokens": 595361802.0, "step": 15682 }, { "epoch": 1.99503879913497, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7092835903167725, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.873368501663208, "num_tokens": 595402853.0, "step": 15683 }, { "epoch": 1.9951660094135606, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.604475975036621, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8767745494842529, "num_tokens": 595443774.0, "step": 15684 }, { "epoch": 1.9952932196921511, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.710660457611084, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8661237955093384, "num_tokens": 595485395.0, "step": 15685 }, { "epoch": 1.9954204299707416, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5972235202789307, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.866870641708374, "num_tokens": 595527724.0, "step": 15686 }, { "epoch": 1.9955476402493322, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7184417247772217, "learning_rate": 1e-05, "loss": 0.4814, "mean_token_accuracy": 0.8600765466690063, "num_tokens": 595569081.0, "step": 15687 }, { "epoch": 1.9956748505279227, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5903561115264893, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.862427830696106, "num_tokens": 595611400.0, "step": 15688 }, { "epoch": 1.995802060806513, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7317776679992676, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8698149919509888, "num_tokens": 595648063.0, "step": 15689 }, { "epoch": 1.9959292710851035, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5865256786346436, "learning_rate": 1e-05, "loss": 0.4856, "mean_token_accuracy": 0.8579287528991699, "num_tokens": 595688679.0, "step": 15690 }, { "epoch": 1.996056481363694, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7465708255767822, "learning_rate": 1e-05, "loss": 0.4713, "mean_token_accuracy": 0.8612769842147827, "num_tokens": 595726120.0, "step": 15691 }, { "epoch": 1.9961836916422846, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.574826955795288, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.871792197227478, "num_tokens": 595762020.0, "step": 15692 }, { "epoch": 1.9963109019208751, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7607791423797607, "learning_rate": 1e-05, "loss": 0.5027, "mean_token_accuracy": 0.8507636785507202, "num_tokens": 595797909.0, "step": 15693 }, { "epoch": 1.9964381121994657, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.575075387954712, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8672593832015991, "num_tokens": 595837542.0, "step": 15694 }, { "epoch": 1.9965653224780562, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.8821842670440674, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8752329349517822, "num_tokens": 595874006.0, "step": 15695 }, { "epoch": 1.9966925327566467, "ewc_loss": 0.06787109375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 3.9577484130859375e-05, "grad_norm": 2.572114944458008, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8650000095367432, "num_tokens": 595904098.0, "step": 15696 }, { "epoch": 1.9968197430352372, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.826916456222534, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8761290311813354, "num_tokens": 595945725.0, "step": 15697 }, { "epoch": 1.9969469533138278, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5731632709503174, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8706831336021423, "num_tokens": 595977826.0, "step": 15698 }, { "epoch": 1.9970741635924183, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8532912731170654, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8633647561073303, "num_tokens": 596011350.0, "step": 15699 }, { "epoch": 1.9972013738710088, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5608291625976562, "learning_rate": 1e-05, "loss": 0.3943, "mean_token_accuracy": 0.8853722810745239, "num_tokens": 596050688.0, "step": 15700 }, { "epoch": 1.9973285841495994, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8527309894561768, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8809974789619446, "num_tokens": 596090711.0, "step": 15701 }, { "epoch": 1.9974557944281899, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.594109535217285, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8701112270355225, "num_tokens": 596130346.0, "step": 15702 }, { "epoch": 1.9975830047067804, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8507308959960938, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8624334931373596, "num_tokens": 596173035.0, "step": 15703 }, { "epoch": 1.997710214985371, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.612182378768921, "learning_rate": 1e-05, "loss": 0.4894, "mean_token_accuracy": 0.8535758256912231, "num_tokens": 596208105.0, "step": 15704 }, { "epoch": 1.9978374252639615, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8143675327301025, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8849879503250122, "num_tokens": 596247367.0, "step": 15705 }, { "epoch": 1.997964635542552, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.60880708694458, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8691861629486084, "num_tokens": 596288226.0, "step": 15706 }, { "epoch": 1.9980918458211423, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.804469108581543, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8596681356430054, "num_tokens": 596329422.0, "step": 15707 }, { "epoch": 1.9982190560997328, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.611992597579956, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8753129243850708, "num_tokens": 596367447.0, "step": 15708 }, { "epoch": 1.9983462663783234, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.807626962661743, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.86810302734375, "num_tokens": 596401564.0, "step": 15709 }, { "epoch": 1.9984734766569139, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.60923171043396, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8649722337722778, "num_tokens": 596441076.0, "step": 15710 }, { "epoch": 1.9986006869355044, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7761824131011963, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8797122836112976, "num_tokens": 596480360.0, "step": 15711 }, { "epoch": 1.998727897214095, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.630842924118042, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8609873056411743, "num_tokens": 596512836.0, "step": 15712 }, { "epoch": 1.9988551074926852, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7502193450927734, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8844884634017944, "num_tokens": 596549154.0, "step": 15713 }, { "epoch": 1.9989823177712758, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6024956703186035, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8780826330184937, "num_tokens": 596590528.0, "step": 15714 }, { "epoch": 1.9991095280498663, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.754608631134033, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.871628999710083, "num_tokens": 596623926.0, "step": 15715 }, { "epoch": 1.9992367383284568, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5999834537506104, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8733384609222412, "num_tokens": 596661102.0, "step": 15716 }, { "epoch": 1.9993639486070474, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.732933521270752, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8845618963241577, "num_tokens": 596701113.0, "step": 15717 }, { "epoch": 1.9994911588856379, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.60912823677063, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8815560340881348, "num_tokens": 596735130.0, "step": 15718 }, { "epoch": 1.9996183691642284, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.736765146255493, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8583186864852905, "num_tokens": 596774612.0, "step": 15719 }, { "epoch": 1.999745579442819, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6077358722686768, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8708661794662476, "num_tokens": 596811033.0, "step": 15720 }, { "epoch": 1.9998727897214095, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.733200788497925, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8601345419883728, "num_tokens": 596849169.0, "step": 15721 }, { "epoch": 2.0, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.598206043243408, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8775404691696167, "num_tokens": 596886661.0, "step": 15722 }, { "epoch": 2.0001272102785905, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.725149393081665, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8872792720794678, "num_tokens": 596927769.0, "step": 15723 }, { "epoch": 2.000254420557181, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5930049419403076, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8819423913955688, "num_tokens": 596966558.0, "step": 15724 }, { "epoch": 2.0003816308357716, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.720170259475708, "learning_rate": 1e-05, "loss": 0.3875, "mean_token_accuracy": 0.8876211643218994, "num_tokens": 597007424.0, "step": 15725 }, { "epoch": 2.000508841114362, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5999984741210938, "learning_rate": 1e-05, "loss": 0.3988, "mean_token_accuracy": 0.8819853067398071, "num_tokens": 597043544.0, "step": 15726 }, { "epoch": 2.0006360513929526, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.715280771255493, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8810310363769531, "num_tokens": 597081079.0, "step": 15727 }, { "epoch": 2.000763261671543, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 3.981590270996094e-05, "grad_norm": 2.581812858581543, "learning_rate": 1e-05, "loss": 0.3677, "mean_token_accuracy": 0.8904467821121216, "num_tokens": 597122188.0, "step": 15728 }, { "epoch": 2.0008904719501337, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.726717948913574, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.879529595375061, "num_tokens": 597164835.0, "step": 15729 }, { "epoch": 2.0010176822287242, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.594947576522827, "learning_rate": 1e-05, "loss": 0.3868, "mean_token_accuracy": 0.8873723149299622, "num_tokens": 597201266.0, "step": 15730 }, { "epoch": 2.0011448925073148, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7291924953460693, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.8857565522193909, "num_tokens": 597244095.0, "step": 15731 }, { "epoch": 2.0012721027859053, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5921759605407715, "learning_rate": 1e-05, "loss": 0.4037, "mean_token_accuracy": 0.8807421326637268, "num_tokens": 597285462.0, "step": 15732 }, { "epoch": 2.001399313064496, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7488880157470703, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8878638744354248, "num_tokens": 597323436.0, "step": 15733 }, { "epoch": 2.0015265233430863, "ewc_loss": 0.068359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.5882656574249268, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8768393993377686, "num_tokens": 597361775.0, "step": 15734 }, { "epoch": 2.0016537336216764, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.8293588161468506, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8722876906394958, "num_tokens": 597402772.0, "step": 15735 }, { "epoch": 2.001780943900267, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5614678859710693, "learning_rate": 1e-05, "loss": 0.3906, "mean_token_accuracy": 0.8866907358169556, "num_tokens": 597444245.0, "step": 15736 }, { "epoch": 2.0019081541788575, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8091516494750977, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8825831413269043, "num_tokens": 597481742.0, "step": 15737 }, { "epoch": 2.002035364457448, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.576843023300171, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8735264539718628, "num_tokens": 597520339.0, "step": 15738 }, { "epoch": 2.0021625747360385, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.822718381881714, "learning_rate": 1e-05, "loss": 0.3877, "mean_token_accuracy": 0.8882126808166504, "num_tokens": 597552756.0, "step": 15739 }, { "epoch": 2.002289785014629, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.588336229324341, "learning_rate": 1e-05, "loss": 0.3769, "mean_token_accuracy": 0.8895778059959412, "num_tokens": 597582224.0, "step": 15740 }, { "epoch": 2.0024169952932196, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8171546459198, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8828966617584229, "num_tokens": 597619121.0, "step": 15741 }, { "epoch": 2.00254420557181, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5993242263793945, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8718514442443848, "num_tokens": 597654074.0, "step": 15742 }, { "epoch": 2.0026714158504006, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8020637035369873, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8850553035736084, "num_tokens": 597695873.0, "step": 15743 }, { "epoch": 2.002798626128991, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.597257614135742, "learning_rate": 1e-05, "loss": 0.3888, "mean_token_accuracy": 0.8862016201019287, "num_tokens": 597733072.0, "step": 15744 }, { "epoch": 2.0029258364075817, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7912867069244385, "learning_rate": 1e-05, "loss": 0.3481, "mean_token_accuracy": 0.9021698236465454, "num_tokens": 597772777.0, "step": 15745 }, { "epoch": 2.0030530466861722, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6300220489501953, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8722337484359741, "num_tokens": 597809276.0, "step": 15746 }, { "epoch": 2.0031802569647628, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7681915760040283, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8761846423149109, "num_tokens": 597848956.0, "step": 15747 }, { "epoch": 2.0033074672433533, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6124093532562256, "learning_rate": 1e-05, "loss": 0.3932, "mean_token_accuracy": 0.887031078338623, "num_tokens": 597890664.0, "step": 15748 }, { "epoch": 2.003434677521944, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7739930152893066, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8798539638519287, "num_tokens": 597931017.0, "step": 15749 }, { "epoch": 2.0035618878005343, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6012182235717773, "learning_rate": 1e-05, "loss": 0.3649, "mean_token_accuracy": 0.8955355882644653, "num_tokens": 597971552.0, "step": 15750 }, { "epoch": 2.003689098079125, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7775771617889404, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8726728558540344, "num_tokens": 598011554.0, "step": 15751 }, { "epoch": 2.0038163083577154, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6033575534820557, "learning_rate": 1e-05, "loss": 0.4014, "mean_token_accuracy": 0.8824366927146912, "num_tokens": 598050633.0, "step": 15752 }, { "epoch": 2.003943518636306, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7691268920898438, "learning_rate": 1e-05, "loss": 0.3827, "mean_token_accuracy": 0.8891172409057617, "num_tokens": 598088022.0, "step": 15753 }, { "epoch": 2.0040707289148965, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6010918617248535, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.880806565284729, "num_tokens": 598126333.0, "step": 15754 }, { "epoch": 2.004197939193487, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7817442417144775, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.879954993724823, "num_tokens": 598161725.0, "step": 15755 }, { "epoch": 2.0043251494720775, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5973026752471924, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8863100409507751, "num_tokens": 598203158.0, "step": 15756 }, { "epoch": 2.004452359750668, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7800559997558594, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8854837417602539, "num_tokens": 598241093.0, "step": 15757 }, { "epoch": 2.0045795700292586, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.607438087463379, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8823059797286987, "num_tokens": 598276389.0, "step": 15758 }, { "epoch": 2.0047067803078487, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.80790638923645, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8848457336425781, "num_tokens": 598316645.0, "step": 15759 }, { "epoch": 2.004833990586439, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5902633666992188, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8794350624084473, "num_tokens": 598361443.0, "step": 15760 }, { "epoch": 2.0049612008650297, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7821192741394043, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8835429549217224, "num_tokens": 598402934.0, "step": 15761 }, { "epoch": 2.0050884111436202, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6030328273773193, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8777392506599426, "num_tokens": 598436315.0, "step": 15762 }, { "epoch": 2.0052156214222108, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.781114339828491, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8891212940216064, "num_tokens": 598476562.0, "step": 15763 }, { "epoch": 2.0053428317008013, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5919547080993652, "learning_rate": 1e-05, "loss": 0.3673, "mean_token_accuracy": 0.8949069976806641, "num_tokens": 598515180.0, "step": 15764 }, { "epoch": 2.005470041979392, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.780017614364624, "learning_rate": 1e-05, "loss": 0.3742, "mean_token_accuracy": 0.8935791850090027, "num_tokens": 598554177.0, "step": 15765 }, { "epoch": 2.0055972522579824, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.609975814819336, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.879989743232727, "num_tokens": 598589091.0, "step": 15766 }, { "epoch": 2.005724462536573, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.781548261642456, "learning_rate": 1e-05, "loss": 0.3875, "mean_token_accuracy": 0.8891280889511108, "num_tokens": 598624570.0, "step": 15767 }, { "epoch": 2.0058516728151634, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.607652425765991, "learning_rate": 1e-05, "loss": 0.3405, "mean_token_accuracy": 0.901370108127594, "num_tokens": 598660350.0, "step": 15768 }, { "epoch": 2.005978883093754, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7804067134857178, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8691330552101135, "num_tokens": 598701823.0, "step": 15769 }, { "epoch": 2.0061060933723445, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002841949462890625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6186609268188477, "learning_rate": 1e-05, "loss": 0.3816, "mean_token_accuracy": 0.8883551359176636, "num_tokens": 598738637.0, "step": 15770 }, { "epoch": 2.006233303650935, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.77459454536438, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8850514888763428, "num_tokens": 598773459.0, "step": 15771 }, { "epoch": 2.0063605139295255, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6055822372436523, "learning_rate": 1e-05, "loss": 0.3848, "mean_token_accuracy": 0.8888671398162842, "num_tokens": 598810098.0, "step": 15772 }, { "epoch": 2.006487724208116, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.760927200317383, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8744105100631714, "num_tokens": 598850701.0, "step": 15773 }, { "epoch": 2.0066149344867066, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.601595878601074, "learning_rate": 1e-05, "loss": 0.3592, "mean_token_accuracy": 0.8971784114837646, "num_tokens": 598887745.0, "step": 15774 }, { "epoch": 2.006742144765297, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.755922555923462, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8852984309196472, "num_tokens": 598928360.0, "step": 15775 }, { "epoch": 2.0068693550438876, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6035985946655273, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8841292858123779, "num_tokens": 598965628.0, "step": 15776 }, { "epoch": 2.006996565322478, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7560949325561523, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8817269802093506, "num_tokens": 599008445.0, "step": 15777 }, { "epoch": 2.0071237756010687, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.614593029022217, "learning_rate": 1e-05, "loss": 0.3613, "mean_token_accuracy": 0.8992980718612671, "num_tokens": 599041730.0, "step": 15778 }, { "epoch": 2.007250985879659, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.741666316986084, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8865675330162048, "num_tokens": 599082481.0, "step": 15779 }, { "epoch": 2.0073781961582498, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6120193004608154, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8815088272094727, "num_tokens": 599122503.0, "step": 15780 }, { "epoch": 2.0075054064368403, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7481529712677, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8699994683265686, "num_tokens": 599158558.0, "step": 15781 }, { "epoch": 2.007632616715431, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6070640087127686, "learning_rate": 1e-05, "loss": 0.3927, "mean_token_accuracy": 0.8870565295219421, "num_tokens": 599196769.0, "step": 15782 }, { "epoch": 2.0077598269940213, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7468390464782715, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8848508596420288, "num_tokens": 599231050.0, "step": 15783 }, { "epoch": 2.0078870372726114, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.601088523864746, "learning_rate": 1e-05, "loss": 0.3806, "mean_token_accuracy": 0.8900223970413208, "num_tokens": 599271826.0, "step": 15784 }, { "epoch": 2.008014247551202, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.752816677093506, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8779124617576599, "num_tokens": 599311134.0, "step": 15785 }, { "epoch": 2.0081414578297925, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6218364238739014, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8851161599159241, "num_tokens": 599349240.0, "step": 15786 }, { "epoch": 2.008268668108383, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7474911212921143, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8845840692520142, "num_tokens": 599381624.0, "step": 15787 }, { "epoch": 2.0083958783869735, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.604570150375366, "learning_rate": 1e-05, "loss": 0.3899, "mean_token_accuracy": 0.8840070366859436, "num_tokens": 599421524.0, "step": 15788 }, { "epoch": 2.008523088665564, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.743279457092285, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8753784894943237, "num_tokens": 599459285.0, "step": 15789 }, { "epoch": 2.0086502989441546, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.589618444442749, "learning_rate": 1e-05, "loss": 0.3564, "mean_token_accuracy": 0.897220253944397, "num_tokens": 599501682.0, "step": 15790 }, { "epoch": 2.008777509222745, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7494993209838867, "learning_rate": 1e-05, "loss": 0.3775, "mean_token_accuracy": 0.8935580849647522, "num_tokens": 599538518.0, "step": 15791 }, { "epoch": 2.0089047195013356, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.607100248336792, "learning_rate": 1e-05, "loss": 0.3924, "mean_token_accuracy": 0.8908275961875916, "num_tokens": 599572802.0, "step": 15792 }, { "epoch": 2.009031929779926, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7435250282287598, "learning_rate": 1e-05, "loss": 0.3875, "mean_token_accuracy": 0.8895715475082397, "num_tokens": 599614366.0, "step": 15793 }, { "epoch": 2.0091591400585167, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5967817306518555, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8816463947296143, "num_tokens": 599654814.0, "step": 15794 }, { "epoch": 2.0092863503371072, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7583041191101074, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8839214444160461, "num_tokens": 599689678.0, "step": 15795 }, { "epoch": 2.0094135606156978, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.608314275741577, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8797997236251831, "num_tokens": 599725932.0, "step": 15796 }, { "epoch": 2.0095407708942883, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7484755516052246, "learning_rate": 1e-05, "loss": 0.3744, "mean_token_accuracy": 0.8945400714874268, "num_tokens": 599762041.0, "step": 15797 }, { "epoch": 2.009667981172879, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5992417335510254, "learning_rate": 1e-05, "loss": 0.3868, "mean_token_accuracy": 0.8879381418228149, "num_tokens": 599800706.0, "step": 15798 }, { "epoch": 2.0097951914514693, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.772322177886963, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.869343101978302, "num_tokens": 599833483.0, "step": 15799 }, { "epoch": 2.00992240173006, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5985164642333984, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8842860460281372, "num_tokens": 599872172.0, "step": 15800 }, { "epoch": 2.0100496120086504, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.756831407546997, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8787668943405151, "num_tokens": 599914149.0, "step": 15801 }, { "epoch": 2.010176822287241, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.601815700531006, "learning_rate": 1e-05, "loss": 0.3922, "mean_token_accuracy": 0.8842957019805908, "num_tokens": 599950603.0, "step": 15802 }, { "epoch": 2.0103040325658315, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.760810136795044, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8770052194595337, "num_tokens": 599991844.0, "step": 15803 }, { "epoch": 2.010431242844422, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5952236652374268, "learning_rate": 1e-05, "loss": 0.3881, "mean_token_accuracy": 0.8886027336120605, "num_tokens": 600033303.0, "step": 15804 }, { "epoch": 2.0105584531230125, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.771340847015381, "learning_rate": 1e-05, "loss": 0.3615, "mean_token_accuracy": 0.8979037404060364, "num_tokens": 600065633.0, "step": 15805 }, { "epoch": 2.010685663401603, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5885610580444336, "learning_rate": 1e-05, "loss": 0.3968, "mean_token_accuracy": 0.8824890851974487, "num_tokens": 600104333.0, "step": 15806 }, { "epoch": 2.0108128736801936, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7833974361419678, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8782587051391602, "num_tokens": 600141329.0, "step": 15807 }, { "epoch": 2.0109400839587837, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.579796075820923, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.872976541519165, "num_tokens": 600184171.0, "step": 15808 }, { "epoch": 2.011067294237374, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.793912172317505, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8693177700042725, "num_tokens": 600220249.0, "step": 15809 }, { "epoch": 2.0111945045159647, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.000286102294921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6095428466796875, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8795159459114075, "num_tokens": 600256015.0, "step": 15810 }, { "epoch": 2.0113217147945552, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8171725273132324, "learning_rate": 1e-05, "loss": 0.3986, "mean_token_accuracy": 0.8824496865272522, "num_tokens": 600282893.0, "step": 15811 }, { "epoch": 2.0114489250731458, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.589599609375, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8640795946121216, "num_tokens": 600320940.0, "step": 15812 }, { "epoch": 2.0115761353517363, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8012781143188477, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8655785322189331, "num_tokens": 600362054.0, "step": 15813 }, { "epoch": 2.011703345630327, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6093976497650146, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8690381646156311, "num_tokens": 600395080.0, "step": 15814 }, { "epoch": 2.0118305559089174, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7970757484436035, "learning_rate": 1e-05, "loss": 0.381, "mean_token_accuracy": 0.8908618688583374, "num_tokens": 600428060.0, "step": 15815 }, { "epoch": 2.011957766187508, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.585174083709717, "learning_rate": 1e-05, "loss": 0.3858, "mean_token_accuracy": 0.889488935470581, "num_tokens": 600469833.0, "step": 15816 }, { "epoch": 2.0120849764660984, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8207054138183594, "learning_rate": 1e-05, "loss": 0.3898, "mean_token_accuracy": 0.8853855133056641, "num_tokens": 600498006.0, "step": 15817 }, { "epoch": 2.012212186744689, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.592144250869751, "learning_rate": 1e-05, "loss": 0.3923, "mean_token_accuracy": 0.8848247528076172, "num_tokens": 600534744.0, "step": 15818 }, { "epoch": 2.0123393970232795, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8172929286956787, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8813008069992065, "num_tokens": 600570282.0, "step": 15819 }, { "epoch": 2.01246660730187, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.604827404022217, "learning_rate": 1e-05, "loss": 0.3756, "mean_token_accuracy": 0.8909096717834473, "num_tokens": 600603858.0, "step": 15820 }, { "epoch": 2.0125938175804605, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.81589412689209, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8690139055252075, "num_tokens": 600642078.0, "step": 15821 }, { "epoch": 2.012721027859051, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6052801609039307, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8782580494880676, "num_tokens": 600679424.0, "step": 15822 }, { "epoch": 2.0128482381376416, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7996320724487305, "learning_rate": 1e-05, "loss": 0.3835, "mean_token_accuracy": 0.8912546634674072, "num_tokens": 600716310.0, "step": 15823 }, { "epoch": 2.012975448416232, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6109519004821777, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.871167778968811, "num_tokens": 600757031.0, "step": 15824 }, { "epoch": 2.0131026586948226, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.793290376663208, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8816584348678589, "num_tokens": 600795414.0, "step": 15825 }, { "epoch": 2.013229868973413, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.630601406097412, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8701915740966797, "num_tokens": 600829239.0, "step": 15826 }, { "epoch": 2.0133570792520037, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7660446166992188, "learning_rate": 1e-05, "loss": 0.3762, "mean_token_accuracy": 0.8910326957702637, "num_tokens": 600868155.0, "step": 15827 }, { "epoch": 2.013484289530594, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6202104091644287, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8762932419776917, "num_tokens": 600906797.0, "step": 15828 }, { "epoch": 2.0136114998091847, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.769014358520508, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8741057515144348, "num_tokens": 600948399.0, "step": 15829 }, { "epoch": 2.0137387100877753, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.60486102104187, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8795201182365417, "num_tokens": 600989682.0, "step": 15830 }, { "epoch": 2.013865920366366, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7643964290618896, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8790848255157471, "num_tokens": 601025387.0, "step": 15831 }, { "epoch": 2.0139931306449563, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6045095920562744, "learning_rate": 1e-05, "loss": 0.3698, "mean_token_accuracy": 0.8939387798309326, "num_tokens": 601061124.0, "step": 15832 }, { "epoch": 2.0141203409235464, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.757194995880127, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8775126338005066, "num_tokens": 601105789.0, "step": 15833 }, { "epoch": 2.014247551202137, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.627044200897217, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.868622899055481, "num_tokens": 601138697.0, "step": 15834 }, { "epoch": 2.0143747614807275, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.750699043273926, "learning_rate": 1e-05, "loss": 0.3747, "mean_token_accuracy": 0.8928505778312683, "num_tokens": 601172875.0, "step": 15835 }, { "epoch": 2.014501971759318, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.616474151611328, "learning_rate": 1e-05, "loss": 0.3646, "mean_token_accuracy": 0.8967981934547424, "num_tokens": 601206303.0, "step": 15836 }, { "epoch": 2.0146291820379085, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7418203353881836, "learning_rate": 1e-05, "loss": 0.3591, "mean_token_accuracy": 0.8984718322753906, "num_tokens": 601243879.0, "step": 15837 }, { "epoch": 2.014756392316499, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6130425930023193, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8833895921707153, "num_tokens": 601282639.0, "step": 15838 }, { "epoch": 2.0148836025950896, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7341175079345703, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8791276812553406, "num_tokens": 601326204.0, "step": 15839 }, { "epoch": 2.01501081287368, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.622828722000122, "learning_rate": 1e-05, "loss": 0.3829, "mean_token_accuracy": 0.8870073556900024, "num_tokens": 601358936.0, "step": 15840 }, { "epoch": 2.0151380231522706, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.721165418624878, "learning_rate": 1e-05, "loss": 0.3645, "mean_token_accuracy": 0.8949013948440552, "num_tokens": 601398253.0, "step": 15841 }, { "epoch": 2.015265233430861, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6083292961120605, "learning_rate": 1e-05, "loss": 0.3237, "mean_token_accuracy": 0.9071482419967651, "num_tokens": 601434806.0, "step": 15842 }, { "epoch": 2.0153924437094517, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7336137294769287, "learning_rate": 1e-05, "loss": 0.38, "mean_token_accuracy": 0.8870192170143127, "num_tokens": 601471834.0, "step": 15843 }, { "epoch": 2.0155196539880422, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.606586217880249, "learning_rate": 1e-05, "loss": 0.3765, "mean_token_accuracy": 0.8914573192596436, "num_tokens": 601511256.0, "step": 15844 }, { "epoch": 2.0156468642666328, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.731039047241211, "learning_rate": 1e-05, "loss": 0.3728, "mean_token_accuracy": 0.8913112878799438, "num_tokens": 601547434.0, "step": 15845 }, { "epoch": 2.0157740745452233, "ewc_loss": 0.06884765625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.6061196327209473, "learning_rate": 1e-05, "loss": 0.3718, "mean_token_accuracy": 0.8914517164230347, "num_tokens": 601581375.0, "step": 15846 }, { "epoch": 2.015901284823814, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.746370553970337, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.879009485244751, "num_tokens": 601615067.0, "step": 15847 }, { "epoch": 2.0160284951024043, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.586994171142578, "learning_rate": 1e-05, "loss": 0.338, "mean_token_accuracy": 0.9005041122436523, "num_tokens": 601654168.0, "step": 15848 }, { "epoch": 2.016155705380995, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7586164474487305, "learning_rate": 1e-05, "loss": 0.3719, "mean_token_accuracy": 0.8948917984962463, "num_tokens": 601687113.0, "step": 15849 }, { "epoch": 2.0162829156595854, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5926995277404785, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8758127689361572, "num_tokens": 601723993.0, "step": 15850 }, { "epoch": 2.016410125938176, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7758710384368896, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8667316436767578, "num_tokens": 601758522.0, "step": 15851 }, { "epoch": 2.0165373362167665, "ewc_loss": 0.0693359375, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.577057361602783, "learning_rate": 1e-05, "loss": 0.3391, "mean_token_accuracy": 0.9043899774551392, "num_tokens": 601794221.0, "step": 15852 }, { "epoch": 2.016664546495357, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7735462188720703, "learning_rate": 1e-05, "loss": 0.3811, "mean_token_accuracy": 0.8930280208587646, "num_tokens": 601830430.0, "step": 15853 }, { "epoch": 2.0167917567739475, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5751054286956787, "learning_rate": 1e-05, "loss": 0.3926, "mean_token_accuracy": 0.8845022916793823, "num_tokens": 601871834.0, "step": 15854 }, { "epoch": 2.016918967052538, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7836074829101562, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8784886598587036, "num_tokens": 601911432.0, "step": 15855 }, { "epoch": 2.0170461773311286, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5724785327911377, "learning_rate": 1e-05, "loss": 0.3713, "mean_token_accuracy": 0.8930826783180237, "num_tokens": 601955592.0, "step": 15856 }, { "epoch": 2.0171733876097186, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0002880096435546875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.797642946243286, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.8790014982223511, "num_tokens": 601996843.0, "step": 15857 }, { "epoch": 2.017300597888309, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.584658622741699, "learning_rate": 1e-05, "loss": 0.3773, "mean_token_accuracy": 0.8915494680404663, "num_tokens": 602034963.0, "step": 15858 }, { "epoch": 2.0174278081668997, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.817814350128174, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8827637434005737, "num_tokens": 602073460.0, "step": 15859 }, { "epoch": 2.0175550184454902, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5863633155822754, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8695665597915649, "num_tokens": 602117442.0, "step": 15860 }, { "epoch": 2.0176822287240808, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8062520027160645, "learning_rate": 1e-05, "loss": 0.3855, "mean_token_accuracy": 0.8884440660476685, "num_tokens": 602161253.0, "step": 15861 }, { "epoch": 2.0178094390026713, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.580289363861084, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8752744197845459, "num_tokens": 602204324.0, "step": 15862 }, { "epoch": 2.017936649281262, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.828873872756958, "learning_rate": 1e-05, "loss": 0.3844, "mean_token_accuracy": 0.887660026550293, "num_tokens": 602236335.0, "step": 15863 }, { "epoch": 2.0180638595598523, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6162941455841064, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8763518929481506, "num_tokens": 602271302.0, "step": 15864 }, { "epoch": 2.018191069838443, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8162503242492676, "learning_rate": 1e-05, "loss": 0.3819, "mean_token_accuracy": 0.8905947208404541, "num_tokens": 602304103.0, "step": 15865 }, { "epoch": 2.0183182801170334, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5942862033843994, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8841414451599121, "num_tokens": 602343293.0, "step": 15866 }, { "epoch": 2.018445490395624, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8025994300842285, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8770352005958557, "num_tokens": 602383650.0, "step": 15867 }, { "epoch": 2.0185727006742145, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5997426509857178, "learning_rate": 1e-05, "loss": 0.3798, "mean_token_accuracy": 0.8932104110717773, "num_tokens": 602423436.0, "step": 15868 }, { "epoch": 2.018699910952805, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.802196979522705, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8837897777557373, "num_tokens": 602460909.0, "step": 15869 }, { "epoch": 2.0188271212313955, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6114461421966553, "learning_rate": 1e-05, "loss": 0.4061, "mean_token_accuracy": 0.8804473876953125, "num_tokens": 602499610.0, "step": 15870 }, { "epoch": 2.018954331509986, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.784634590148926, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.879629373550415, "num_tokens": 602540783.0, "step": 15871 }, { "epoch": 2.0190815417885766, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.61002516746521, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8770310878753662, "num_tokens": 602579205.0, "step": 15872 }, { "epoch": 2.019208752067167, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7826297283172607, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8819268941879272, "num_tokens": 602613720.0, "step": 15873 }, { "epoch": 2.0193359623457576, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6066501140594482, "learning_rate": 1e-05, "loss": 0.3886, "mean_token_accuracy": 0.8895928859710693, "num_tokens": 602651954.0, "step": 15874 }, { "epoch": 2.019463172624348, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7807319164276123, "learning_rate": 1e-05, "loss": 0.3734, "mean_token_accuracy": 0.8947765231132507, "num_tokens": 602685553.0, "step": 15875 }, { "epoch": 2.0195903829029387, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.602952003479004, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8773311376571655, "num_tokens": 602727854.0, "step": 15876 }, { "epoch": 2.019717593181529, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7773873805999756, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8840458393096924, "num_tokens": 602765993.0, "step": 15877 }, { "epoch": 2.0198448034601197, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5958926677703857, "learning_rate": 1e-05, "loss": 0.3652, "mean_token_accuracy": 0.8943319916725159, "num_tokens": 602807450.0, "step": 15878 }, { "epoch": 2.0199720137387103, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7805092334747314, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8668785095214844, "num_tokens": 602848689.0, "step": 15879 }, { "epoch": 2.020099224017301, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.592852830886841, "learning_rate": 1e-05, "loss": 0.3759, "mean_token_accuracy": 0.8962668180465698, "num_tokens": 602888010.0, "step": 15880 }, { "epoch": 2.0202264342958913, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.799144744873047, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8684682846069336, "num_tokens": 602923345.0, "step": 15881 }, { "epoch": 2.0203536445744814, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5913140773773193, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8853086233139038, "num_tokens": 602967320.0, "step": 15882 }, { "epoch": 2.020480854853072, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.790074110031128, "learning_rate": 1e-05, "loss": 0.364, "mean_token_accuracy": 0.8937289714813232, "num_tokens": 603005177.0, "step": 15883 }, { "epoch": 2.0206080651316625, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.598557710647583, "learning_rate": 1e-05, "loss": 0.3655, "mean_token_accuracy": 0.8942297697067261, "num_tokens": 603039562.0, "step": 15884 }, { "epoch": 2.020735275410253, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7870233058929443, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.890133261680603, "num_tokens": 603079075.0, "step": 15885 }, { "epoch": 2.0208624856888435, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6098740100860596, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.882850706577301, "num_tokens": 603118077.0, "step": 15886 }, { "epoch": 2.020989695967434, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7856318950653076, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8744643926620483, "num_tokens": 603156921.0, "step": 15887 }, { "epoch": 2.0211169062460246, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5941038131713867, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8846499919891357, "num_tokens": 603202062.0, "step": 15888 }, { "epoch": 2.021244116524615, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7848334312438965, "learning_rate": 1e-05, "loss": 0.3876, "mean_token_accuracy": 0.8903409838676453, "num_tokens": 603236240.0, "step": 15889 }, { "epoch": 2.0213713268032056, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6182992458343506, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8789905905723572, "num_tokens": 603277376.0, "step": 15890 }, { "epoch": 2.021498537081796, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7636234760284424, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.8834700584411621, "num_tokens": 603314740.0, "step": 15891 }, { "epoch": 2.0216257473603867, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.607680559158325, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.8843111991882324, "num_tokens": 603354992.0, "step": 15892 }, { "epoch": 2.021752957638977, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7709884643554688, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.8790616393089294, "num_tokens": 603387271.0, "step": 15893 }, { "epoch": 2.0218801679175677, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.600864887237549, "learning_rate": 1e-05, "loss": 0.379, "mean_token_accuracy": 0.8921979665756226, "num_tokens": 603427092.0, "step": 15894 }, { "epoch": 2.0220073781961583, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7702596187591553, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8723424077033997, "num_tokens": 603465188.0, "step": 15895 }, { "epoch": 2.022134588474749, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.607508420944214, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8815470933914185, "num_tokens": 603499290.0, "step": 15896 }, { "epoch": 2.0222617987533393, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.813974142074585, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8871612548828125, "num_tokens": 603530594.0, "step": 15897 }, { "epoch": 2.02238900903193, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.617617130279541, "learning_rate": 1e-05, "loss": 0.3919, "mean_token_accuracy": 0.8851757049560547, "num_tokens": 603569337.0, "step": 15898 }, { "epoch": 2.0225162193105204, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.765535593032837, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8853726387023926, "num_tokens": 603609704.0, "step": 15899 }, { "epoch": 2.022643429589111, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6077959537506104, "learning_rate": 1e-05, "loss": 0.4008, "mean_token_accuracy": 0.8830920457839966, "num_tokens": 603647623.0, "step": 15900 }, { "epoch": 2.0227706398677014, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7722811698913574, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8866423964500427, "num_tokens": 603682237.0, "step": 15901 }, { "epoch": 2.022897850146292, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6042938232421875, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8779743313789368, "num_tokens": 603720348.0, "step": 15902 }, { "epoch": 2.0230250604248825, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7681405544281006, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8853572607040405, "num_tokens": 603759857.0, "step": 15903 }, { "epoch": 2.023152270703473, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.616460084915161, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8792095184326172, "num_tokens": 603796970.0, "step": 15904 }, { "epoch": 2.0232794809820636, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.764453411102295, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8807064890861511, "num_tokens": 603838385.0, "step": 15905 }, { "epoch": 2.0234066912606536, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5937612056732178, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.8818198442459106, "num_tokens": 603884262.0, "step": 15906 }, { "epoch": 2.023533901539244, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7663185596466064, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8784624338150024, "num_tokens": 603925013.0, "step": 15907 }, { "epoch": 2.0236611118178347, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.589815378189087, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8738322257995605, "num_tokens": 603971173.0, "step": 15908 }, { "epoch": 2.0237883220964252, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7793521881103516, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8776752948760986, "num_tokens": 604009091.0, "step": 15909 }, { "epoch": 2.0239155323750158, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.604095697402954, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8827298879623413, "num_tokens": 604045163.0, "step": 15910 }, { "epoch": 2.0240427426536063, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7593777179718018, "learning_rate": 1e-05, "loss": 0.3866, "mean_token_accuracy": 0.8871332406997681, "num_tokens": 604091379.0, "step": 15911 }, { "epoch": 2.024169952932197, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6045210361480713, "learning_rate": 1e-05, "loss": 0.4098, "mean_token_accuracy": 0.8807615041732788, "num_tokens": 604132159.0, "step": 15912 }, { "epoch": 2.0242971632107873, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.782681941986084, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8725123405456543, "num_tokens": 604169689.0, "step": 15913 }, { "epoch": 2.024424373489378, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6015946865081787, "learning_rate": 1e-05, "loss": 0.3434, "mean_token_accuracy": 0.9024428129196167, "num_tokens": 604201913.0, "step": 15914 }, { "epoch": 2.0245515837679684, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7795774936676025, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8766636848449707, "num_tokens": 604238275.0, "step": 15915 }, { "epoch": 2.024678794046559, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.585132360458374, "learning_rate": 1e-05, "loss": 0.3745, "mean_token_accuracy": 0.8906440138816833, "num_tokens": 604279862.0, "step": 15916 }, { "epoch": 2.0248060043251495, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.777644395828247, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8812752366065979, "num_tokens": 604319091.0, "step": 15917 }, { "epoch": 2.02493321460374, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.591135025024414, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.8755452036857605, "num_tokens": 604363614.0, "step": 15918 }, { "epoch": 2.0250604248823305, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8059263229370117, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8806725740432739, "num_tokens": 604394422.0, "step": 15919 }, { "epoch": 2.025187635160921, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5976758003234863, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8803635835647583, "num_tokens": 604432844.0, "step": 15920 }, { "epoch": 2.0253148454395116, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7889256477355957, "learning_rate": 1e-05, "loss": 0.4012, "mean_token_accuracy": 0.8843730688095093, "num_tokens": 604472074.0, "step": 15921 }, { "epoch": 2.025442055718102, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5892322063446045, "learning_rate": 1e-05, "loss": 0.347, "mean_token_accuracy": 0.8994932770729065, "num_tokens": 604507827.0, "step": 15922 }, { "epoch": 2.0255692659966926, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.791180372238159, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8779001832008362, "num_tokens": 604548600.0, "step": 15923 }, { "epoch": 2.025696476275283, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002899169921875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.609467029571533, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8683255314826965, "num_tokens": 604582701.0, "step": 15924 }, { "epoch": 2.0258236865538737, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7781779766082764, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.874567985534668, "num_tokens": 604624089.0, "step": 15925 }, { "epoch": 2.025950896832464, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6146538257598877, "learning_rate": 1e-05, "loss": 0.3769, "mean_token_accuracy": 0.8918784260749817, "num_tokens": 604655764.0, "step": 15926 }, { "epoch": 2.0260781071110547, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7755484580993652, "learning_rate": 1e-05, "loss": 0.3779, "mean_token_accuracy": 0.8921669125556946, "num_tokens": 604688000.0, "step": 15927 }, { "epoch": 2.0262053173896453, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.612769842147827, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8775003552436829, "num_tokens": 604724602.0, "step": 15928 }, { "epoch": 2.026332527668236, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7622969150543213, "learning_rate": 1e-05, "loss": 0.3945, "mean_token_accuracy": 0.8863328695297241, "num_tokens": 604761794.0, "step": 15929 }, { "epoch": 2.0264597379468263, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5956075191497803, "learning_rate": 1e-05, "loss": 0.3546, "mean_token_accuracy": 0.89902663230896, "num_tokens": 604800239.0, "step": 15930 }, { "epoch": 2.0265869482254164, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7686684131622314, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8837786912918091, "num_tokens": 604837143.0, "step": 15931 }, { "epoch": 2.026714158504007, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.599346876144409, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8835021257400513, "num_tokens": 604874074.0, "step": 15932 }, { "epoch": 2.0268413687825975, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.779158353805542, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8767498731613159, "num_tokens": 604911639.0, "step": 15933 }, { "epoch": 2.026968579061188, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.596104145050049, "learning_rate": 1e-05, "loss": 0.3695, "mean_token_accuracy": 0.8948500752449036, "num_tokens": 604944963.0, "step": 15934 }, { "epoch": 2.0270957893397785, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7692747116088867, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8835991621017456, "num_tokens": 604986048.0, "step": 15935 }, { "epoch": 2.027222999618369, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6018788814544678, "learning_rate": 1e-05, "loss": 0.3804, "mean_token_accuracy": 0.8892693519592285, "num_tokens": 605023995.0, "step": 15936 }, { "epoch": 2.0273502098969596, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8820602893829346, "learning_rate": 1e-05, "loss": 0.387, "mean_token_accuracy": 0.8887795805931091, "num_tokens": 605064612.0, "step": 15937 }, { "epoch": 2.02747742017555, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5965487957000732, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8780162930488586, "num_tokens": 605099332.0, "step": 15938 }, { "epoch": 2.0276046304541406, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.816072702407837, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8843920230865479, "num_tokens": 605137616.0, "step": 15939 }, { "epoch": 2.027731840732731, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5888893604278564, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8797620534896851, "num_tokens": 605177974.0, "step": 15940 }, { "epoch": 2.0278590510113217, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8100697994232178, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8814185857772827, "num_tokens": 605219156.0, "step": 15941 }, { "epoch": 2.027986261289912, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5810348987579346, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.8844205141067505, "num_tokens": 605261131.0, "step": 15942 }, { "epoch": 2.0281134715685027, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.803992509841919, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8786198496818542, "num_tokens": 605303640.0, "step": 15943 }, { "epoch": 2.0282406818470933, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.600459098815918, "learning_rate": 1e-05, "loss": 0.3682, "mean_token_accuracy": 0.8924969434738159, "num_tokens": 605341823.0, "step": 15944 }, { "epoch": 2.028367892125684, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.7930221557617188, "learning_rate": 1e-05, "loss": 0.3798, "mean_token_accuracy": 0.8923248648643494, "num_tokens": 605380282.0, "step": 15945 }, { "epoch": 2.0284951024042743, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.60794997215271, "learning_rate": 1e-05, "loss": 0.3779, "mean_token_accuracy": 0.8903173804283142, "num_tokens": 605415077.0, "step": 15946 }, { "epoch": 2.028622312682865, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.787071943283081, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8829442262649536, "num_tokens": 605453805.0, "step": 15947 }, { "epoch": 2.0287495229614554, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.626969814300537, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8826338052749634, "num_tokens": 605488740.0, "step": 15948 }, { "epoch": 2.028876733240046, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7675516605377197, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8740330934524536, "num_tokens": 605529155.0, "step": 15949 }, { "epoch": 2.0290039435186364, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.615679979324341, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8825955390930176, "num_tokens": 605567610.0, "step": 15950 }, { "epoch": 2.029131153797227, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002918243408203125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7574477195739746, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8783621788024902, "num_tokens": 605609485.0, "step": 15951 }, { "epoch": 2.0292583640758175, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.606602907180786, "learning_rate": 1e-05, "loss": 0.3703, "mean_token_accuracy": 0.8938463926315308, "num_tokens": 605644718.0, "step": 15952 }, { "epoch": 2.029385574354408, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7448337078094482, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.884373664855957, "num_tokens": 605686893.0, "step": 15953 }, { "epoch": 2.0295127846329986, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6167261600494385, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8842787742614746, "num_tokens": 605724007.0, "step": 15954 }, { "epoch": 2.0296399949115886, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7528769969940186, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8757277727127075, "num_tokens": 605761213.0, "step": 15955 }, { "epoch": 2.029767205190179, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5960865020751953, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8855313062667847, "num_tokens": 605803274.0, "step": 15956 }, { "epoch": 2.0298944154687697, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7631947994232178, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8798366785049438, "num_tokens": 605838326.0, "step": 15957 }, { "epoch": 2.0300216257473602, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6019670963287354, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8851514458656311, "num_tokens": 605878882.0, "step": 15958 }, { "epoch": 2.0301488360259508, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.771929979324341, "learning_rate": 1e-05, "loss": 0.3844, "mean_token_accuracy": 0.8882600665092468, "num_tokens": 605913574.0, "step": 15959 }, { "epoch": 2.0302760463045413, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.614182710647583, "learning_rate": 1e-05, "loss": 0.3733, "mean_token_accuracy": 0.8920996189117432, "num_tokens": 605944845.0, "step": 15960 }, { "epoch": 2.030403256583132, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7608535289764404, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8629449009895325, "num_tokens": 605982604.0, "step": 15961 }, { "epoch": 2.0305304668617223, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5984692573547363, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8779182434082031, "num_tokens": 606023506.0, "step": 15962 }, { "epoch": 2.030657677140313, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.753593683242798, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8736483454704285, "num_tokens": 606063319.0, "step": 15963 }, { "epoch": 2.0307848874189034, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.595576524734497, "learning_rate": 1e-05, "loss": 0.3587, "mean_token_accuracy": 0.8973639607429504, "num_tokens": 606103393.0, "step": 15964 }, { "epoch": 2.030912097697494, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7496142387390137, "learning_rate": 1e-05, "loss": 0.3763, "mean_token_accuracy": 0.893639326095581, "num_tokens": 606144191.0, "step": 15965 }, { "epoch": 2.0310393079760845, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6024341583251953, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8865898251533508, "num_tokens": 606183219.0, "step": 15966 }, { "epoch": 2.031166518254675, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.766023635864258, "learning_rate": 1e-05, "loss": 0.41, "mean_token_accuracy": 0.8793411254882812, "num_tokens": 606221429.0, "step": 15967 }, { "epoch": 2.0312937285332655, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5986084938049316, "learning_rate": 1e-05, "loss": 0.3816, "mean_token_accuracy": 0.8880786895751953, "num_tokens": 606260810.0, "step": 15968 }, { "epoch": 2.031420938811856, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.768467664718628, "learning_rate": 1e-05, "loss": 0.3864, "mean_token_accuracy": 0.888434648513794, "num_tokens": 606299602.0, "step": 15969 }, { "epoch": 2.0315481490904466, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5843539237976074, "learning_rate": 1e-05, "loss": 0.3883, "mean_token_accuracy": 0.8865962028503418, "num_tokens": 606341130.0, "step": 15970 }, { "epoch": 2.031675359369037, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.789353370666504, "learning_rate": 1e-05, "loss": 0.3946, "mean_token_accuracy": 0.8870538473129272, "num_tokens": 606378341.0, "step": 15971 }, { "epoch": 2.0318025696476276, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.612339496612549, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8649137020111084, "num_tokens": 606413799.0, "step": 15972 }, { "epoch": 2.031929779926218, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7788937091827393, "learning_rate": 1e-05, "loss": 0.3766, "mean_token_accuracy": 0.8907795548439026, "num_tokens": 606447428.0, "step": 15973 }, { "epoch": 2.0320569902048087, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.604665756225586, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8832466006278992, "num_tokens": 606481652.0, "step": 15974 }, { "epoch": 2.032184200483399, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7654895782470703, "learning_rate": 1e-05, "loss": 0.3535, "mean_token_accuracy": 0.9000667929649353, "num_tokens": 606519279.0, "step": 15975 }, { "epoch": 2.0323114107619897, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.599954605102539, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.8841467499732971, "num_tokens": 606556399.0, "step": 15976 }, { "epoch": 2.0324386210405803, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7769076824188232, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8687214255332947, "num_tokens": 606596807.0, "step": 15977 }, { "epoch": 2.032565831319171, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.612407684326172, "learning_rate": 1e-05, "loss": 0.3611, "mean_token_accuracy": 0.8961397409439087, "num_tokens": 606629231.0, "step": 15978 }, { "epoch": 2.032693041597761, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7618584632873535, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8788264393806458, "num_tokens": 606669042.0, "step": 15979 }, { "epoch": 2.0328202518763514, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.625575065612793, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8738678693771362, "num_tokens": 606705550.0, "step": 15980 }, { "epoch": 2.032947462154942, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.75591778755188, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8771040439605713, "num_tokens": 606739866.0, "step": 15981 }, { "epoch": 2.0330746724335325, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5994815826416016, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.882073163986206, "num_tokens": 606782923.0, "step": 15982 }, { "epoch": 2.033201882712123, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7529704570770264, "learning_rate": 1e-05, "loss": 0.3922, "mean_token_accuracy": 0.8877584934234619, "num_tokens": 606822831.0, "step": 15983 }, { "epoch": 2.0333290929907135, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6142160892486572, "learning_rate": 1e-05, "loss": 0.3916, "mean_token_accuracy": 0.887032151222229, "num_tokens": 606861129.0, "step": 15984 }, { "epoch": 2.033456303269304, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.743821382522583, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8669015169143677, "num_tokens": 606902744.0, "step": 15985 }, { "epoch": 2.0335835135478946, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6341896057128906, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8756757974624634, "num_tokens": 606935137.0, "step": 15986 }, { "epoch": 2.033710723826485, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.724862813949585, "learning_rate": 1e-05, "loss": 0.3844, "mean_token_accuracy": 0.8870911598205566, "num_tokens": 606972215.0, "step": 15987 }, { "epoch": 2.0338379341050756, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6307051181793213, "learning_rate": 1e-05, "loss": 0.4037, "mean_token_accuracy": 0.8855288624763489, "num_tokens": 607004489.0, "step": 15988 }, { "epoch": 2.033965144383666, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7316911220550537, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8786143064498901, "num_tokens": 607039457.0, "step": 15989 }, { "epoch": 2.0340923546622567, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6114931106567383, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8780730962753296, "num_tokens": 607084863.0, "step": 15990 }, { "epoch": 2.034219564940847, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.728865623474121, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8723925352096558, "num_tokens": 607125013.0, "step": 15991 }, { "epoch": 2.0343467752194377, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.623960018157959, "learning_rate": 1e-05, "loss": 0.3916, "mean_token_accuracy": 0.8864151239395142, "num_tokens": 607160501.0, "step": 15992 }, { "epoch": 2.0344739854980283, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.724156141281128, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8764805793762207, "num_tokens": 607204621.0, "step": 15993 }, { "epoch": 2.034601195776619, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6037068367004395, "learning_rate": 1e-05, "loss": 0.3848, "mean_token_accuracy": 0.8905884623527527, "num_tokens": 607245904.0, "step": 15994 }, { "epoch": 2.0347284060552093, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.725881338119507, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.885003387928009, "num_tokens": 607286907.0, "step": 15995 }, { "epoch": 2.0348556163338, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6219851970672607, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8713539838790894, "num_tokens": 607325990.0, "step": 15996 }, { "epoch": 2.0349828266123904, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.741746425628662, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8749727606773376, "num_tokens": 607363000.0, "step": 15997 }, { "epoch": 2.035110036890981, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6057472229003906, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8827401399612427, "num_tokens": 607405913.0, "step": 15998 }, { "epoch": 2.0352372471695714, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7470645904541016, "learning_rate": 1e-05, "loss": 0.4033, "mean_token_accuracy": 0.8833187818527222, "num_tokens": 607440242.0, "step": 15999 }, { "epoch": 2.035364457448162, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5882632732391357, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8826689124107361, "num_tokens": 607485705.0, "step": 16000 }, { "epoch": 2.0354916677267525, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7522811889648438, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.87559574842453, "num_tokens": 607524026.0, "step": 16001 }, { "epoch": 2.035618878005343, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6071839332580566, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.876295268535614, "num_tokens": 607562925.0, "step": 16002 }, { "epoch": 2.0357460882839336, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.770479440689087, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8719655275344849, "num_tokens": 607598126.0, "step": 16003 }, { "epoch": 2.0358732985625236, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.592266798019409, "learning_rate": 1e-05, "loss": 0.374, "mean_token_accuracy": 0.8948059678077698, "num_tokens": 607631911.0, "step": 16004 }, { "epoch": 2.036000508841114, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7728281021118164, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8731679916381836, "num_tokens": 607669817.0, "step": 16005 }, { "epoch": 2.0361277191197047, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.588486909866333, "learning_rate": 1e-05, "loss": 0.391, "mean_token_accuracy": 0.8865423202514648, "num_tokens": 607710713.0, "step": 16006 }, { "epoch": 2.036254929398295, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.764235019683838, "learning_rate": 1e-05, "loss": 0.3575, "mean_token_accuracy": 0.8986853361129761, "num_tokens": 607751540.0, "step": 16007 }, { "epoch": 2.0363821396768857, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6086623668670654, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.8834699988365173, "num_tokens": 607792831.0, "step": 16008 }, { "epoch": 2.0365093499554763, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.790928363800049, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8828047513961792, "num_tokens": 607824175.0, "step": 16009 }, { "epoch": 2.036636560234067, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5916292667388916, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8732061982154846, "num_tokens": 607864302.0, "step": 16010 }, { "epoch": 2.0367637705126573, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.785595178604126, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8777626752853394, "num_tokens": 607903231.0, "step": 16011 }, { "epoch": 2.036890980791248, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5831124782562256, "learning_rate": 1e-05, "loss": 0.389, "mean_token_accuracy": 0.8886393308639526, "num_tokens": 607942795.0, "step": 16012 }, { "epoch": 2.0370181910698384, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7922298908233643, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8811050057411194, "num_tokens": 607978034.0, "step": 16013 }, { "epoch": 2.037145401348429, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.605530261993408, "learning_rate": 1e-05, "loss": 0.3787, "mean_token_accuracy": 0.8923736810684204, "num_tokens": 608014260.0, "step": 16014 }, { "epoch": 2.0372726116270194, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7824063301086426, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8727585077285767, "num_tokens": 608052637.0, "step": 16015 }, { "epoch": 2.03739982190561, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.620435953140259, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8774749040603638, "num_tokens": 608085877.0, "step": 16016 }, { "epoch": 2.0375270321842005, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7727391719818115, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8808779716491699, "num_tokens": 608125059.0, "step": 16017 }, { "epoch": 2.037654242462791, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.601942777633667, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.888084888458252, "num_tokens": 608161357.0, "step": 16018 }, { "epoch": 2.0377814527413816, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7773289680480957, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8809770345687866, "num_tokens": 608196534.0, "step": 16019 }, { "epoch": 2.037908663019972, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5944082736968994, "learning_rate": 1e-05, "loss": 0.3987, "mean_token_accuracy": 0.8816576600074768, "num_tokens": 608237130.0, "step": 16020 }, { "epoch": 2.0380358732985626, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7914366722106934, "learning_rate": 1e-05, "loss": 0.3615, "mean_token_accuracy": 0.9001270532608032, "num_tokens": 608279064.0, "step": 16021 }, { "epoch": 2.038163083577153, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5917880535125732, "learning_rate": 1e-05, "loss": 0.3714, "mean_token_accuracy": 0.8922672271728516, "num_tokens": 608316019.0, "step": 16022 }, { "epoch": 2.0382902938557437, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7752621173858643, "learning_rate": 1e-05, "loss": 0.3743, "mean_token_accuracy": 0.8938770294189453, "num_tokens": 608357199.0, "step": 16023 }, { "epoch": 2.038417504134334, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6073577404022217, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8816163539886475, "num_tokens": 608392149.0, "step": 16024 }, { "epoch": 2.0385447144129247, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.774430751800537, "learning_rate": 1e-05, "loss": 0.3874, "mean_token_accuracy": 0.8911054134368896, "num_tokens": 608431200.0, "step": 16025 }, { "epoch": 2.0386719246915153, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.604936122894287, "learning_rate": 1e-05, "loss": 0.3846, "mean_token_accuracy": 0.8883827328681946, "num_tokens": 608471832.0, "step": 16026 }, { "epoch": 2.038799134970106, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.825350522994995, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8723469376564026, "num_tokens": 608508362.0, "step": 16027 }, { "epoch": 2.0389263452486963, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.593334197998047, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8777443170547485, "num_tokens": 608547988.0, "step": 16028 }, { "epoch": 2.0390535555272864, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.802973747253418, "learning_rate": 1e-05, "loss": 0.3812, "mean_token_accuracy": 0.8924015760421753, "num_tokens": 608580448.0, "step": 16029 }, { "epoch": 2.039180765805877, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.601016044616699, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8735851049423218, "num_tokens": 608618573.0, "step": 16030 }, { "epoch": 2.0393079760844675, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7887508869171143, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8804208040237427, "num_tokens": 608658161.0, "step": 16031 }, { "epoch": 2.039435186363058, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6057534217834473, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8757088780403137, "num_tokens": 608696599.0, "step": 16032 }, { "epoch": 2.0395623966416485, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.772425651550293, "learning_rate": 1e-05, "loss": 0.386, "mean_token_accuracy": 0.8905550241470337, "num_tokens": 608738347.0, "step": 16033 }, { "epoch": 2.039689606920239, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6001367568969727, "learning_rate": 1e-05, "loss": 0.3895, "mean_token_accuracy": 0.8865780234336853, "num_tokens": 608779483.0, "step": 16034 }, { "epoch": 2.0398168171988296, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.771043062210083, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8806741237640381, "num_tokens": 608821369.0, "step": 16035 }, { "epoch": 2.03994402747742, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6216859817504883, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8818498849868774, "num_tokens": 608854749.0, "step": 16036 }, { "epoch": 2.0400712377560106, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7479019165039062, "learning_rate": 1e-05, "loss": 0.3784, "mean_token_accuracy": 0.8902192115783691, "num_tokens": 608893692.0, "step": 16037 }, { "epoch": 2.040198448034601, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6176154613494873, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8776757717132568, "num_tokens": 608930721.0, "step": 16038 }, { "epoch": 2.0403256583131917, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7537901401519775, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8754545450210571, "num_tokens": 608967022.0, "step": 16039 }, { "epoch": 2.040452868591782, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6143908500671387, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8643736243247986, "num_tokens": 609009995.0, "step": 16040 }, { "epoch": 2.0405800788703727, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7544970512390137, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8793011903762817, "num_tokens": 609044636.0, "step": 16041 }, { "epoch": 2.0407072891489633, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.62423038482666, "learning_rate": 1e-05, "loss": 0.3831, "mean_token_accuracy": 0.8892861604690552, "num_tokens": 609076328.0, "step": 16042 }, { "epoch": 2.040834499427554, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7370245456695557, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8665698170661926, "num_tokens": 609118955.0, "step": 16043 }, { "epoch": 2.0409617097061443, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6227495670318604, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8711366057395935, "num_tokens": 609157958.0, "step": 16044 }, { "epoch": 2.041088919984735, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.717787504196167, "learning_rate": 1e-05, "loss": 0.3885, "mean_token_accuracy": 0.8866321444511414, "num_tokens": 609197636.0, "step": 16045 }, { "epoch": 2.0412161302633254, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.621304512023926, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.864153265953064, "num_tokens": 609239648.0, "step": 16046 }, { "epoch": 2.041343340541916, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7051503658294678, "learning_rate": 1e-05, "loss": 0.3677, "mean_token_accuracy": 0.8937191963195801, "num_tokens": 609281012.0, "step": 16047 }, { "epoch": 2.0414705508205064, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6240642070770264, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8813480138778687, "num_tokens": 609323775.0, "step": 16048 }, { "epoch": 2.041597761099097, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.711611270904541, "learning_rate": 1e-05, "loss": 0.3664, "mean_token_accuracy": 0.8960525989532471, "num_tokens": 609358960.0, "step": 16049 }, { "epoch": 2.0417249713776875, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.626020908355713, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8714910745620728, "num_tokens": 609395447.0, "step": 16050 }, { "epoch": 2.041852181656278, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7081804275512695, "learning_rate": 1e-05, "loss": 0.3665, "mean_token_accuracy": 0.8949167728424072, "num_tokens": 609429811.0, "step": 16051 }, { "epoch": 2.0419793919348685, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6269428730010986, "learning_rate": 1e-05, "loss": 0.3883, "mean_token_accuracy": 0.887731671333313, "num_tokens": 609466104.0, "step": 16052 }, { "epoch": 2.0421066022134586, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6943204402923584, "learning_rate": 1e-05, "loss": 0.3726, "mean_token_accuracy": 0.8906420469284058, "num_tokens": 609503513.0, "step": 16053 }, { "epoch": 2.042233812492049, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6138088703155518, "learning_rate": 1e-05, "loss": 0.3674, "mean_token_accuracy": 0.893884003162384, "num_tokens": 609546899.0, "step": 16054 }, { "epoch": 2.0423610227706397, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7006916999816895, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8757544159889221, "num_tokens": 609589297.0, "step": 16055 }, { "epoch": 2.04248823304923, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6159093379974365, "learning_rate": 1e-05, "loss": 0.3784, "mean_token_accuracy": 0.8905544281005859, "num_tokens": 609624839.0, "step": 16056 }, { "epoch": 2.0426154433278207, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7102811336517334, "learning_rate": 1e-05, "loss": 0.3748, "mean_token_accuracy": 0.8899146318435669, "num_tokens": 609661708.0, "step": 16057 }, { "epoch": 2.0427426536064113, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.622046709060669, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8724867701530457, "num_tokens": 609704133.0, "step": 16058 }, { "epoch": 2.042869863885002, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7126543521881104, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.8819593191146851, "num_tokens": 609740743.0, "step": 16059 }, { "epoch": 2.0429970741635923, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000293731689453125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6256842613220215, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8725869059562683, "num_tokens": 609777520.0, "step": 16060 }, { "epoch": 2.043124284442183, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7230875492095947, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8827772736549377, "num_tokens": 609812974.0, "step": 16061 }, { "epoch": 2.0432514947207734, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6117422580718994, "learning_rate": 1e-05, "loss": 0.3901, "mean_token_accuracy": 0.8888474702835083, "num_tokens": 609853287.0, "step": 16062 }, { "epoch": 2.043378704999364, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.720130681991577, "learning_rate": 1e-05, "loss": 0.3662, "mean_token_accuracy": 0.8936846256256104, "num_tokens": 609888768.0, "step": 16063 }, { "epoch": 2.0435059152779544, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6127796173095703, "learning_rate": 1e-05, "loss": 0.3824, "mean_token_accuracy": 0.88833087682724, "num_tokens": 609927976.0, "step": 16064 }, { "epoch": 2.043633125556545, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7210323810577393, "learning_rate": 1e-05, "loss": 0.3799, "mean_token_accuracy": 0.8892338275909424, "num_tokens": 609964404.0, "step": 16065 }, { "epoch": 2.0437603358351355, "ewc_loss": 0.06982421875, "ewc_loss_diag": 0.0002956390380859375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.605929136276245, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8827725648880005, "num_tokens": 610005188.0, "step": 16066 }, { "epoch": 2.043887546113726, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7341318130493164, "learning_rate": 1e-05, "loss": 0.3916, "mean_token_accuracy": 0.8896209001541138, "num_tokens": 610039118.0, "step": 16067 }, { "epoch": 2.0440147563923166, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6106884479522705, "learning_rate": 1e-05, "loss": 0.3738, "mean_token_accuracy": 0.8924236297607422, "num_tokens": 610074867.0, "step": 16068 }, { "epoch": 2.044141966670907, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.733222484588623, "learning_rate": 1e-05, "loss": 0.3945, "mean_token_accuracy": 0.8866286277770996, "num_tokens": 610111382.0, "step": 16069 }, { "epoch": 2.0442691769494976, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5982279777526855, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.8801888823509216, "num_tokens": 610148026.0, "step": 16070 }, { "epoch": 2.044396387228088, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.748014211654663, "learning_rate": 1e-05, "loss": 0.3784, "mean_token_accuracy": 0.8895409107208252, "num_tokens": 610179841.0, "step": 16071 }, { "epoch": 2.0445235975066787, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.602786064147949, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8789434432983398, "num_tokens": 610214224.0, "step": 16072 }, { "epoch": 2.044650807785269, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.84618878364563, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.884992241859436, "num_tokens": 610257640.0, "step": 16073 }, { "epoch": 2.0447780180638597, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5652916431427, "learning_rate": 1e-05, "loss": 0.3692, "mean_token_accuracy": 0.8943573832511902, "num_tokens": 610295438.0, "step": 16074 }, { "epoch": 2.0449052283424503, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7991416454315186, "learning_rate": 1e-05, "loss": 0.3822, "mean_token_accuracy": 0.8909643292427063, "num_tokens": 610336390.0, "step": 16075 }, { "epoch": 2.045032438621041, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5685760974884033, "learning_rate": 1e-05, "loss": 0.3853, "mean_token_accuracy": 0.8876575827598572, "num_tokens": 610369860.0, "step": 16076 }, { "epoch": 2.045159648899631, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8290722370147705, "learning_rate": 1e-05, "loss": 0.3972, "mean_token_accuracy": 0.8864887952804565, "num_tokens": 610408116.0, "step": 16077 }, { "epoch": 2.0452868591782214, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.608715295791626, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8741432428359985, "num_tokens": 610444317.0, "step": 16078 }, { "epoch": 2.045414069456812, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.80300235748291, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8749445676803589, "num_tokens": 610485675.0, "step": 16079 }, { "epoch": 2.0455412797354025, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.592909812927246, "learning_rate": 1e-05, "loss": 0.3856, "mean_token_accuracy": 0.8893525004386902, "num_tokens": 610524697.0, "step": 16080 }, { "epoch": 2.045668490013993, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.796804428100586, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8865221738815308, "num_tokens": 610567470.0, "step": 16081 }, { "epoch": 2.0457957002925835, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.603464126586914, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.8848321437835693, "num_tokens": 610603168.0, "step": 16082 }, { "epoch": 2.045922910571174, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.796346664428711, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8823825716972351, "num_tokens": 610643416.0, "step": 16083 }, { "epoch": 2.0460501208497646, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6113007068634033, "learning_rate": 1e-05, "loss": 0.3905, "mean_token_accuracy": 0.8861262798309326, "num_tokens": 610676513.0, "step": 16084 }, { "epoch": 2.046177331128355, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.796943426132202, "learning_rate": 1e-05, "loss": 0.3904, "mean_token_accuracy": 0.8868381977081299, "num_tokens": 610710842.0, "step": 16085 }, { "epoch": 2.0463045414069456, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.610090970993042, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.87058424949646, "num_tokens": 610747770.0, "step": 16086 }, { "epoch": 2.046431751685536, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7698259353637695, "learning_rate": 1e-05, "loss": 0.3899, "mean_token_accuracy": 0.888408899307251, "num_tokens": 610788780.0, "step": 16087 }, { "epoch": 2.0465589619641267, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.600930690765381, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.886770486831665, "num_tokens": 610832930.0, "step": 16088 }, { "epoch": 2.046686172242717, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7694647312164307, "learning_rate": 1e-05, "loss": 0.3808, "mean_token_accuracy": 0.8892073631286621, "num_tokens": 610875828.0, "step": 16089 }, { "epoch": 2.0468133825213077, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6069517135620117, "learning_rate": 1e-05, "loss": 0.3783, "mean_token_accuracy": 0.891305685043335, "num_tokens": 610913920.0, "step": 16090 }, { "epoch": 2.0469405927998983, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7619540691375732, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.884350597858429, "num_tokens": 610956887.0, "step": 16091 }, { "epoch": 2.047067803078489, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6101202964782715, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8741892576217651, "num_tokens": 611000889.0, "step": 16092 }, { "epoch": 2.0471950133570793, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7651774883270264, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8834595680236816, "num_tokens": 611040573.0, "step": 16093 }, { "epoch": 2.04732222363567, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.610320806503296, "learning_rate": 1e-05, "loss": 0.3768, "mean_token_accuracy": 0.8908110857009888, "num_tokens": 611075925.0, "step": 16094 }, { "epoch": 2.0474494339142604, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.759239435195923, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8755372762680054, "num_tokens": 611116492.0, "step": 16095 }, { "epoch": 2.047576644192851, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6035780906677246, "learning_rate": 1e-05, "loss": 0.3839, "mean_token_accuracy": 0.8886013031005859, "num_tokens": 611155683.0, "step": 16096 }, { "epoch": 2.0477038544714414, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.743523597717285, "learning_rate": 1e-05, "loss": 0.3246, "mean_token_accuracy": 0.9094581007957458, "num_tokens": 611194352.0, "step": 16097 }, { "epoch": 2.047831064750032, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6325864791870117, "learning_rate": 1e-05, "loss": 0.3836, "mean_token_accuracy": 0.8877759575843811, "num_tokens": 611229919.0, "step": 16098 }, { "epoch": 2.0479582750286225, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7314047813415527, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8817442059516907, "num_tokens": 611267394.0, "step": 16099 }, { "epoch": 2.048085485307213, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6186161041259766, "learning_rate": 1e-05, "loss": 0.4868, "mean_token_accuracy": 0.855048418045044, "num_tokens": 611312009.0, "step": 16100 }, { "epoch": 2.0482126955858035, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7966363430023193, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8695777654647827, "num_tokens": 611346723.0, "step": 16101 }, { "epoch": 2.0483399058643936, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.608149290084839, "learning_rate": 1e-05, "loss": 0.3924, "mean_token_accuracy": 0.8872228860855103, "num_tokens": 611381803.0, "step": 16102 }, { "epoch": 2.048467116142984, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7670435905456543, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8838471174240112, "num_tokens": 611419579.0, "step": 16103 }, { "epoch": 2.0485943264215747, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5882678031921387, "learning_rate": 1e-05, "loss": 0.3975, "mean_token_accuracy": 0.884357750415802, "num_tokens": 611456904.0, "step": 16104 }, { "epoch": 2.048721536700165, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.767707109451294, "learning_rate": 1e-05, "loss": 0.37, "mean_token_accuracy": 0.8932027816772461, "num_tokens": 611495597.0, "step": 16105 }, { "epoch": 2.0488487469787557, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.60740327835083, "learning_rate": 1e-05, "loss": 0.3766, "mean_token_accuracy": 0.8895177841186523, "num_tokens": 611531307.0, "step": 16106 }, { "epoch": 2.0489759572573463, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7685208320617676, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8692518472671509, "num_tokens": 611567657.0, "step": 16107 }, { "epoch": 2.049103167535937, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6440439224243164, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8803635835647583, "num_tokens": 611608107.0, "step": 16108 }, { "epoch": 2.0492303778145273, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7513587474823, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8808308839797974, "num_tokens": 611644695.0, "step": 16109 }, { "epoch": 2.049357588093118, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.617572069168091, "learning_rate": 1e-05, "loss": 0.3883, "mean_token_accuracy": 0.887349009513855, "num_tokens": 611679330.0, "step": 16110 }, { "epoch": 2.0494847983717084, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.735818862915039, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.870169460773468, "num_tokens": 611718792.0, "step": 16111 }, { "epoch": 2.049612008650299, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6313724517822266, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8739660978317261, "num_tokens": 611752559.0, "step": 16112 }, { "epoch": 2.0497392189288894, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7142221927642822, "learning_rate": 1e-05, "loss": 0.3874, "mean_token_accuracy": 0.8879879713058472, "num_tokens": 611791616.0, "step": 16113 }, { "epoch": 2.04986642920748, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.64611554145813, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8725640773773193, "num_tokens": 611826577.0, "step": 16114 }, { "epoch": 2.0499936394860705, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7136471271514893, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.880656361579895, "num_tokens": 611861777.0, "step": 16115 }, { "epoch": 2.050120849764661, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.643974542617798, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8764398097991943, "num_tokens": 611897933.0, "step": 16116 }, { "epoch": 2.0502480600432516, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.699709892272949, "learning_rate": 1e-05, "loss": 0.4033, "mean_token_accuracy": 0.8831396102905273, "num_tokens": 611935285.0, "step": 16117 }, { "epoch": 2.050375270321842, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.639143943786621, "learning_rate": 1e-05, "loss": 0.3755, "mean_token_accuracy": 0.890855073928833, "num_tokens": 611969879.0, "step": 16118 }, { "epoch": 2.0505024806004326, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7019989490509033, "learning_rate": 1e-05, "loss": 0.3868, "mean_token_accuracy": 0.8880911469459534, "num_tokens": 612003608.0, "step": 16119 }, { "epoch": 2.050629690879023, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.677253007888794, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8794081211090088, "num_tokens": 612044247.0, "step": 16120 }, { "epoch": 2.0507569011576137, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6895201206207275, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8751809597015381, "num_tokens": 612080776.0, "step": 16121 }, { "epoch": 2.050884111436204, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.7278378009796143, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8724566698074341, "num_tokens": 612119162.0, "step": 16122 }, { "epoch": 2.0510113217147947, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6239542961120605, "learning_rate": 1e-05, "loss": 0.37, "mean_token_accuracy": 0.892890214920044, "num_tokens": 612157774.0, "step": 16123 }, { "epoch": 2.0511385319933853, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.697740077972412, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8848350048065186, "num_tokens": 612192169.0, "step": 16124 }, { "epoch": 2.051265742271976, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6219334602355957, "learning_rate": 1e-05, "loss": 0.3823, "mean_token_accuracy": 0.8882001042366028, "num_tokens": 612225936.0, "step": 16125 }, { "epoch": 2.0513929525505663, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.702300786972046, "learning_rate": 1e-05, "loss": 0.3887, "mean_token_accuracy": 0.8868482112884521, "num_tokens": 612262509.0, "step": 16126 }, { "epoch": 2.0515201628291564, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.616419792175293, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8795861005783081, "num_tokens": 612297952.0, "step": 16127 }, { "epoch": 2.051647373107747, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7142388820648193, "learning_rate": 1e-05, "loss": 0.3926, "mean_token_accuracy": 0.885431170463562, "num_tokens": 612332512.0, "step": 16128 }, { "epoch": 2.0517745833863374, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.611611843109131, "learning_rate": 1e-05, "loss": 0.3817, "mean_token_accuracy": 0.8897190690040588, "num_tokens": 612370350.0, "step": 16129 }, { "epoch": 2.051901793664928, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7309319972991943, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8725451231002808, "num_tokens": 612404775.0, "step": 16130 }, { "epoch": 2.0520290039435185, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6024584770202637, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8826180696487427, "num_tokens": 612447226.0, "step": 16131 }, { "epoch": 2.052156214222109, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7511887550354004, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8710859417915344, "num_tokens": 612481358.0, "step": 16132 }, { "epoch": 2.0522834245006996, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5912232398986816, "learning_rate": 1e-05, "loss": 0.3735, "mean_token_accuracy": 0.8929380178451538, "num_tokens": 612519132.0, "step": 16133 }, { "epoch": 2.05241063477929, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.77166485786438, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8745014667510986, "num_tokens": 612559187.0, "step": 16134 }, { "epoch": 2.0525378450578806, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.580192804336548, "learning_rate": 1e-05, "loss": 0.3825, "mean_token_accuracy": 0.8874212503433228, "num_tokens": 612598208.0, "step": 16135 }, { "epoch": 2.052665055336471, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.779740333557129, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.884658932685852, "num_tokens": 612633951.0, "step": 16136 }, { "epoch": 2.0527922656150617, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.581507682800293, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8778867125511169, "num_tokens": 612673978.0, "step": 16137 }, { "epoch": 2.052919475893652, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.778263807296753, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.884379506111145, "num_tokens": 612720488.0, "step": 16138 }, { "epoch": 2.0530466861722427, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5922446250915527, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8798556327819824, "num_tokens": 612755009.0, "step": 16139 }, { "epoch": 2.0531738964508333, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7939453125, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8791602849960327, "num_tokens": 612790653.0, "step": 16140 }, { "epoch": 2.053301106729424, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.586771011352539, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.8871500492095947, "num_tokens": 612829406.0, "step": 16141 }, { "epoch": 2.0534283170080143, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.79563570022583, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.873051643371582, "num_tokens": 612872304.0, "step": 16142 }, { "epoch": 2.053555527286605, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5886590480804443, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.8807446956634521, "num_tokens": 612907909.0, "step": 16143 }, { "epoch": 2.0536827375651954, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.797272205352783, "learning_rate": 1e-05, "loss": 0.3751, "mean_token_accuracy": 0.8941519260406494, "num_tokens": 612943794.0, "step": 16144 }, { "epoch": 2.053809947843786, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5786707401275635, "learning_rate": 1e-05, "loss": 0.3849, "mean_token_accuracy": 0.887220561504364, "num_tokens": 612987125.0, "step": 16145 }, { "epoch": 2.0539371581223764, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.806568145751953, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8747695684432983, "num_tokens": 613028232.0, "step": 16146 }, { "epoch": 2.054064368400967, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5891640186309814, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8846439123153687, "num_tokens": 613066651.0, "step": 16147 }, { "epoch": 2.0541915786795575, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00029754638671875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8004086017608643, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.8896547555923462, "num_tokens": 613105153.0, "step": 16148 }, { "epoch": 2.054318788958148, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.583911180496216, "learning_rate": 1e-05, "loss": 0.3733, "mean_token_accuracy": 0.8904953002929688, "num_tokens": 613146817.0, "step": 16149 }, { "epoch": 2.0544459992367385, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.797996759414673, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8756008148193359, "num_tokens": 613192644.0, "step": 16150 }, { "epoch": 2.0545732095153286, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.597193956375122, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8798434734344482, "num_tokens": 613232283.0, "step": 16151 }, { "epoch": 2.054700419793919, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7984445095062256, "learning_rate": 1e-05, "loss": 0.3887, "mean_token_accuracy": 0.8892116546630859, "num_tokens": 613270916.0, "step": 16152 }, { "epoch": 2.0548276300725097, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6087121963500977, "learning_rate": 1e-05, "loss": 0.3799, "mean_token_accuracy": 0.8913167715072632, "num_tokens": 613303095.0, "step": 16153 }, { "epoch": 2.0549548403511, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.792693614959717, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8778590559959412, "num_tokens": 613341551.0, "step": 16154 }, { "epoch": 2.0550820506296907, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.601259708404541, "learning_rate": 1e-05, "loss": 0.373, "mean_token_accuracy": 0.8915095925331116, "num_tokens": 613377511.0, "step": 16155 }, { "epoch": 2.0552092609082813, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.79443097114563, "learning_rate": 1e-05, "loss": 0.3852, "mean_token_accuracy": 0.8901422619819641, "num_tokens": 613408534.0, "step": 16156 }, { "epoch": 2.055336471186872, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6615798473358154, "learning_rate": 1e-05, "loss": 0.3652, "mean_token_accuracy": 0.8957490921020508, "num_tokens": 613443461.0, "step": 16157 }, { "epoch": 2.0554636814654623, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.765706777572632, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8712169528007507, "num_tokens": 613482438.0, "step": 16158 }, { "epoch": 2.055590891744053, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6138155460357666, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.87380051612854, "num_tokens": 613521937.0, "step": 16159 }, { "epoch": 2.0557181020226434, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7443273067474365, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.871720552444458, "num_tokens": 613563471.0, "step": 16160 }, { "epoch": 2.055845312301234, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6119072437286377, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8820962905883789, "num_tokens": 613603753.0, "step": 16161 }, { "epoch": 2.0559725225798244, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7450339794158936, "learning_rate": 1e-05, "loss": 0.3855, "mean_token_accuracy": 0.8903564214706421, "num_tokens": 613637602.0, "step": 16162 }, { "epoch": 2.056099732858415, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6159281730651855, "learning_rate": 1e-05, "loss": 0.3987, "mean_token_accuracy": 0.8816019296646118, "num_tokens": 613673807.0, "step": 16163 }, { "epoch": 2.0562269431370055, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7520275115966797, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8807358741760254, "num_tokens": 613710702.0, "step": 16164 }, { "epoch": 2.056354153415596, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6065421104431152, "learning_rate": 1e-05, "loss": 0.4037, "mean_token_accuracy": 0.8807952404022217, "num_tokens": 613748890.0, "step": 16165 }, { "epoch": 2.0564813636941865, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7502410411834717, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8804926872253418, "num_tokens": 613791039.0, "step": 16166 }, { "epoch": 2.056608573972777, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6289126873016357, "learning_rate": 1e-05, "loss": 0.3883, "mean_token_accuracy": 0.8874292373657227, "num_tokens": 613821345.0, "step": 16167 }, { "epoch": 2.0567357842513676, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.757011651992798, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8565361499786377, "num_tokens": 613857770.0, "step": 16168 }, { "epoch": 2.056862994529958, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.600846767425537, "learning_rate": 1e-05, "loss": 0.3785, "mean_token_accuracy": 0.8879459500312805, "num_tokens": 613898844.0, "step": 16169 }, { "epoch": 2.0569902048085487, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7521402835845947, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8735924959182739, "num_tokens": 613940884.0, "step": 16170 }, { "epoch": 2.057117415087139, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6018009185791016, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8834438920021057, "num_tokens": 613981664.0, "step": 16171 }, { "epoch": 2.0572446253657297, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7806098461151123, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8765212297439575, "num_tokens": 614013965.0, "step": 16172 }, { "epoch": 2.0573718356443202, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6103098392486572, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8841627836227417, "num_tokens": 614048841.0, "step": 16173 }, { "epoch": 2.0574990459229108, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.760354995727539, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8836494088172913, "num_tokens": 614087677.0, "step": 16174 }, { "epoch": 2.057626256201501, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.59389328956604, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8824148178100586, "num_tokens": 614128546.0, "step": 16175 }, { "epoch": 2.0577534664800914, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7795345783233643, "learning_rate": 1e-05, "loss": 0.3863, "mean_token_accuracy": 0.8894810080528259, "num_tokens": 614163947.0, "step": 16176 }, { "epoch": 2.057880676758682, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5995025634765625, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8796786665916443, "num_tokens": 614202353.0, "step": 16177 }, { "epoch": 2.0580078870372724, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7722396850585938, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8782621622085571, "num_tokens": 614244781.0, "step": 16178 }, { "epoch": 2.058135097315863, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6112353801727295, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.8842695355415344, "num_tokens": 614278335.0, "step": 16179 }, { "epoch": 2.0582623075944535, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.769181966781616, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8789627552032471, "num_tokens": 614317726.0, "step": 16180 }, { "epoch": 2.058389517873044, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5984325408935547, "learning_rate": 1e-05, "loss": 0.3842, "mean_token_accuracy": 0.8890694379806519, "num_tokens": 614355403.0, "step": 16181 }, { "epoch": 2.0585167281516346, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7546207904815674, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8694275617599487, "num_tokens": 614404370.0, "step": 16182 }, { "epoch": 2.058643938430225, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.598581552505493, "learning_rate": 1e-05, "loss": 0.3725, "mean_token_accuracy": 0.8903912305831909, "num_tokens": 614439613.0, "step": 16183 }, { "epoch": 2.0587711487088156, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7612359523773193, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8759579658508301, "num_tokens": 614481634.0, "step": 16184 }, { "epoch": 2.058898358987406, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6072609424591064, "learning_rate": 1e-05, "loss": 0.3838, "mean_token_accuracy": 0.8875837326049805, "num_tokens": 614517178.0, "step": 16185 }, { "epoch": 2.0590255692659967, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7789628505706787, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8688215613365173, "num_tokens": 614551531.0, "step": 16186 }, { "epoch": 2.059152779544587, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6082370281219482, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8771122694015503, "num_tokens": 614589173.0, "step": 16187 }, { "epoch": 2.0592799898231777, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7529757022857666, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8798743486404419, "num_tokens": 614633657.0, "step": 16188 }, { "epoch": 2.0594072001017683, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6157097816467285, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8787322044372559, "num_tokens": 614670013.0, "step": 16189 }, { "epoch": 2.059534410380359, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.830688238143921, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8755143284797668, "num_tokens": 614706414.0, "step": 16190 }, { "epoch": 2.0596616206589493, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.592033624649048, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8781245350837708, "num_tokens": 614746397.0, "step": 16191 }, { "epoch": 2.05978883093754, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.78544282913208, "learning_rate": 1e-05, "loss": 0.386, "mean_token_accuracy": 0.8899141550064087, "num_tokens": 614785587.0, "step": 16192 }, { "epoch": 2.0599160412161304, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.0002994537353515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.582176446914673, "learning_rate": 1e-05, "loss": 0.4003, "mean_token_accuracy": 0.8857027292251587, "num_tokens": 614823304.0, "step": 16193 }, { "epoch": 2.060043251494721, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8169455528259277, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8837889432907104, "num_tokens": 614857544.0, "step": 16194 }, { "epoch": 2.0601704617733114, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5977702140808105, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8778191804885864, "num_tokens": 614893095.0, "step": 16195 }, { "epoch": 2.060297672051902, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.790344476699829, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8771171569824219, "num_tokens": 614932634.0, "step": 16196 }, { "epoch": 2.0604248823304925, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.605433702468872, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8790270090103149, "num_tokens": 614975039.0, "step": 16197 }, { "epoch": 2.060552092609083, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7817413806915283, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8840949535369873, "num_tokens": 615019092.0, "step": 16198 }, { "epoch": 2.0606793028876735, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5888609886169434, "learning_rate": 1e-05, "loss": 0.3748, "mean_token_accuracy": 0.8924344182014465, "num_tokens": 615060183.0, "step": 16199 }, { "epoch": 2.0608065131662636, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7919037342071533, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8807849884033203, "num_tokens": 615098376.0, "step": 16200 }, { "epoch": 2.060933723444854, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6118526458740234, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8890633583068848, "num_tokens": 615134401.0, "step": 16201 }, { "epoch": 2.0610609337234447, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.775190830230713, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8848168849945068, "num_tokens": 615176145.0, "step": 16202 }, { "epoch": 2.061188144002035, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6281778812408447, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.87685626745224, "num_tokens": 615210281.0, "step": 16203 }, { "epoch": 2.0613153542806257, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.750330924987793, "learning_rate": 1e-05, "loss": 0.3614, "mean_token_accuracy": 0.8991292119026184, "num_tokens": 615246952.0, "step": 16204 }, { "epoch": 2.0614425645592163, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.606306552886963, "learning_rate": 1e-05, "loss": 0.3715, "mean_token_accuracy": 0.8952820301055908, "num_tokens": 615285807.0, "step": 16205 }, { "epoch": 2.061569774837807, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.755814790725708, "learning_rate": 1e-05, "loss": 0.3657, "mean_token_accuracy": 0.8953444957733154, "num_tokens": 615319641.0, "step": 16206 }, { "epoch": 2.0616969851163973, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6263577938079834, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8699027299880981, "num_tokens": 615358072.0, "step": 16207 }, { "epoch": 2.061824195394988, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7547810077667236, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.8829609155654907, "num_tokens": 615393924.0, "step": 16208 }, { "epoch": 2.0619514056735784, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6056742668151855, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8826634883880615, "num_tokens": 615436771.0, "step": 16209 }, { "epoch": 2.062078615952169, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.748222827911377, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8771224021911621, "num_tokens": 615473577.0, "step": 16210 }, { "epoch": 2.0622058262307594, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.62954044342041, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8663033246994019, "num_tokens": 615507585.0, "step": 16211 }, { "epoch": 2.06233303650935, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.738204002380371, "learning_rate": 1e-05, "loss": 0.3833, "mean_token_accuracy": 0.889191746711731, "num_tokens": 615542638.0, "step": 16212 }, { "epoch": 2.0624602467879405, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.625507116317749, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8787343502044678, "num_tokens": 615579567.0, "step": 16213 }, { "epoch": 2.062587457066531, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.722959041595459, "learning_rate": 1e-05, "loss": 0.3786, "mean_token_accuracy": 0.8908039927482605, "num_tokens": 615616563.0, "step": 16214 }, { "epoch": 2.0627146673451215, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.620230197906494, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8724815845489502, "num_tokens": 615653243.0, "step": 16215 }, { "epoch": 2.062841877623712, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7129173278808594, "learning_rate": 1e-05, "loss": 0.3764, "mean_token_accuracy": 0.8893685340881348, "num_tokens": 615689552.0, "step": 16216 }, { "epoch": 2.0629690879023026, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6403584480285645, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8717065453529358, "num_tokens": 615723803.0, "step": 16217 }, { "epoch": 2.063096298180893, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7248363494873047, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8829972743988037, "num_tokens": 615761181.0, "step": 16218 }, { "epoch": 2.0632235084594837, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.607034206390381, "learning_rate": 1e-05, "loss": 0.3795, "mean_token_accuracy": 0.8911139965057373, "num_tokens": 615802200.0, "step": 16219 }, { "epoch": 2.063350718738074, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7980213165283203, "learning_rate": 1e-05, "loss": 0.3895, "mean_token_accuracy": 0.8869304656982422, "num_tokens": 615843299.0, "step": 16220 }, { "epoch": 2.0634779290166647, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.5807688236236572, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8803924322128296, "num_tokens": 615886148.0, "step": 16221 }, { "epoch": 2.0636051392952552, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7650485038757324, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8755350112915039, "num_tokens": 615930595.0, "step": 16222 }, { "epoch": 2.0637323495738458, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5871450901031494, "learning_rate": 1e-05, "loss": 0.4012, "mean_token_accuracy": 0.8825691342353821, "num_tokens": 615966873.0, "step": 16223 }, { "epoch": 2.0638595598524363, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7817471027374268, "learning_rate": 1e-05, "loss": 0.3608, "mean_token_accuracy": 0.8949770331382751, "num_tokens": 616002021.0, "step": 16224 }, { "epoch": 2.0639867701310264, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.591596841812134, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8829169869422913, "num_tokens": 616043692.0, "step": 16225 }, { "epoch": 2.064113980409617, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.777965784072876, "learning_rate": 1e-05, "loss": 0.3817, "mean_token_accuracy": 0.8931298851966858, "num_tokens": 616083811.0, "step": 16226 }, { "epoch": 2.0642411906882074, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.584681987762451, "learning_rate": 1e-05, "loss": 0.3633, "mean_token_accuracy": 0.8944011926651001, "num_tokens": 616121714.0, "step": 16227 }, { "epoch": 2.064368400966798, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.800344467163086, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.879560649394989, "num_tokens": 616156571.0, "step": 16228 }, { "epoch": 2.0644956112453885, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.59987211227417, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8771692514419556, "num_tokens": 616197198.0, "step": 16229 }, { "epoch": 2.064622821523979, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.799224853515625, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8705232739448547, "num_tokens": 616233495.0, "step": 16230 }, { "epoch": 2.0647500318025696, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5908091068267822, "learning_rate": 1e-05, "loss": 0.368, "mean_token_accuracy": 0.8960345983505249, "num_tokens": 616268894.0, "step": 16231 }, { "epoch": 2.06487724208116, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.800535202026367, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8829892873764038, "num_tokens": 616302899.0, "step": 16232 }, { "epoch": 2.0650044523597506, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6059181690216064, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8632699847221375, "num_tokens": 616342937.0, "step": 16233 }, { "epoch": 2.065131662638341, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7758517265319824, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8833066821098328, "num_tokens": 616391581.0, "step": 16234 }, { "epoch": 2.0652588729169317, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.60280442237854, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8849271535873413, "num_tokens": 616430303.0, "step": 16235 }, { "epoch": 2.065386083195522, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8116791248321533, "learning_rate": 1e-05, "loss": 0.3745, "mean_token_accuracy": 0.8925701379776001, "num_tokens": 616468294.0, "step": 16236 }, { "epoch": 2.0655132934741127, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.605607032775879, "learning_rate": 1e-05, "loss": 0.3844, "mean_token_accuracy": 0.8866156339645386, "num_tokens": 616503551.0, "step": 16237 }, { "epoch": 2.0656405037527032, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7854156494140625, "learning_rate": 1e-05, "loss": 0.3886, "mean_token_accuracy": 0.890774130821228, "num_tokens": 616541374.0, "step": 16238 }, { "epoch": 2.065767714031294, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.599493980407715, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.886353611946106, "num_tokens": 616575480.0, "step": 16239 }, { "epoch": 2.0658949243098843, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7807302474975586, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8810682892799377, "num_tokens": 616620323.0, "step": 16240 }, { "epoch": 2.066022134588475, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6010868549346924, "learning_rate": 1e-05, "loss": 0.3707, "mean_token_accuracy": 0.8961157202720642, "num_tokens": 616659843.0, "step": 16241 }, { "epoch": 2.0661493448670654, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7938106060028076, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8875608444213867, "num_tokens": 616696065.0, "step": 16242 }, { "epoch": 2.066276555145656, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.614661455154419, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8872731328010559, "num_tokens": 616729444.0, "step": 16243 }, { "epoch": 2.0664037654242464, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7703702449798584, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8832111358642578, "num_tokens": 616773351.0, "step": 16244 }, { "epoch": 2.066530975702837, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.614633321762085, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8807421922683716, "num_tokens": 616812161.0, "step": 16245 }, { "epoch": 2.0666581859814275, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.794339179992676, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.8665016889572144, "num_tokens": 616847623.0, "step": 16246 }, { "epoch": 2.066785396260018, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6098079681396484, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8802953362464905, "num_tokens": 616884078.0, "step": 16247 }, { "epoch": 2.0669126065386085, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7757980823516846, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8812527060508728, "num_tokens": 616924291.0, "step": 16248 }, { "epoch": 2.0670398168171986, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5965328216552734, "learning_rate": 1e-05, "loss": 0.3856, "mean_token_accuracy": 0.8883187770843506, "num_tokens": 616964580.0, "step": 16249 }, { "epoch": 2.067167027095789, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7754311561584473, "learning_rate": 1e-05, "loss": 0.377, "mean_token_accuracy": 0.8935848474502563, "num_tokens": 617000574.0, "step": 16250 }, { "epoch": 2.0672942373743797, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.598694086074829, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8827351331710815, "num_tokens": 617043323.0, "step": 16251 }, { "epoch": 2.06742144765297, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7831366062164307, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8866358995437622, "num_tokens": 617082585.0, "step": 16252 }, { "epoch": 2.0675486579315607, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6097803115844727, "learning_rate": 1e-05, "loss": 0.3899, "mean_token_accuracy": 0.8858205080032349, "num_tokens": 617118384.0, "step": 16253 }, { "epoch": 2.0676758682101513, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7668380737304688, "learning_rate": 1e-05, "loss": 0.361, "mean_token_accuracy": 0.8983707427978516, "num_tokens": 617153354.0, "step": 16254 }, { "epoch": 2.067803078488742, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.603827953338623, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8847276568412781, "num_tokens": 617192829.0, "step": 16255 }, { "epoch": 2.0679302887673323, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7792041301727295, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8851850032806396, "num_tokens": 617230824.0, "step": 16256 }, { "epoch": 2.068057499045923, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6119580268859863, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8730087280273438, "num_tokens": 617268791.0, "step": 16257 }, { "epoch": 2.0681847093245134, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.768562078475952, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8823760151863098, "num_tokens": 617305366.0, "step": 16258 }, { "epoch": 2.068311919603104, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.621540069580078, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8820069432258606, "num_tokens": 617342294.0, "step": 16259 }, { "epoch": 2.0684391298816944, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.760403633117676, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8835089802742004, "num_tokens": 617380456.0, "step": 16260 }, { "epoch": 2.068566340160285, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.599318027496338, "learning_rate": 1e-05, "loss": 0.3802, "mean_token_accuracy": 0.8898534774780273, "num_tokens": 617420705.0, "step": 16261 }, { "epoch": 2.0686935504388755, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7521111965179443, "learning_rate": 1e-05, "loss": 0.3906, "mean_token_accuracy": 0.8864903450012207, "num_tokens": 617462740.0, "step": 16262 }, { "epoch": 2.068820760717466, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6138222217559814, "learning_rate": 1e-05, "loss": 0.3749, "mean_token_accuracy": 0.8903137445449829, "num_tokens": 617497496.0, "step": 16263 }, { "epoch": 2.0689479709960565, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7465176582336426, "learning_rate": 1e-05, "loss": 0.371, "mean_token_accuracy": 0.8937002420425415, "num_tokens": 617535297.0, "step": 16264 }, { "epoch": 2.069075181274647, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.629230499267578, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8820092678070068, "num_tokens": 617568332.0, "step": 16265 }, { "epoch": 2.0692023915532376, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.732815980911255, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8795745372772217, "num_tokens": 617608934.0, "step": 16266 }, { "epoch": 2.069329601831828, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.638171434402466, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8743772506713867, "num_tokens": 617643794.0, "step": 16267 }, { "epoch": 2.0694568121104187, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7340455055236816, "learning_rate": 1e-05, "loss": 0.3942, "mean_token_accuracy": 0.8879944682121277, "num_tokens": 617676448.0, "step": 16268 }, { "epoch": 2.069584022389009, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6448607444763184, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8600440621376038, "num_tokens": 617713460.0, "step": 16269 }, { "epoch": 2.0697112326675997, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7414047718048096, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8754194974899292, "num_tokens": 617749767.0, "step": 16270 }, { "epoch": 2.0698384429461902, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6354527473449707, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8660266399383545, "num_tokens": 617784991.0, "step": 16271 }, { "epoch": 2.0699656532247808, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7087655067443848, "learning_rate": 1e-05, "loss": 0.3719, "mean_token_accuracy": 0.8915938138961792, "num_tokens": 617820241.0, "step": 16272 }, { "epoch": 2.070092863503371, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.627009630203247, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.871700644493103, "num_tokens": 617861838.0, "step": 16273 }, { "epoch": 2.0702200737819614, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7218172550201416, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8840177059173584, "num_tokens": 617893001.0, "step": 16274 }, { "epoch": 2.070347284060552, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6061484813690186, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8783448934555054, "num_tokens": 617936361.0, "step": 16275 }, { "epoch": 2.0704744943391424, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.739501476287842, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8802800178527832, "num_tokens": 617968267.0, "step": 16276 }, { "epoch": 2.070601704617733, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6071600914001465, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8809818029403687, "num_tokens": 618007710.0, "step": 16277 }, { "epoch": 2.0707289148963235, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7477736473083496, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.8828381896018982, "num_tokens": 618043185.0, "step": 16278 }, { "epoch": 2.070856125174914, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5887041091918945, "learning_rate": 1e-05, "loss": 0.3709, "mean_token_accuracy": 0.8920857906341553, "num_tokens": 618085460.0, "step": 16279 }, { "epoch": 2.0709833354535045, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7656912803649902, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8911992311477661, "num_tokens": 618117155.0, "step": 16280 }, { "epoch": 2.071110545732095, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5994322299957275, "learning_rate": 1e-05, "loss": 0.3879, "mean_token_accuracy": 0.8904500603675842, "num_tokens": 618153380.0, "step": 16281 }, { "epoch": 2.0712377560106856, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.773183822631836, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8827650547027588, "num_tokens": 618190125.0, "step": 16282 }, { "epoch": 2.071364966289276, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5981051921844482, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8842487335205078, "num_tokens": 618225281.0, "step": 16283 }, { "epoch": 2.0714921765678667, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7758185863494873, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.8885806798934937, "num_tokens": 618262213.0, "step": 16284 }, { "epoch": 2.071619386846457, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.624955415725708, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8648201823234558, "num_tokens": 618292780.0, "step": 16285 }, { "epoch": 2.0717465971250477, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7556545734405518, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8676450252532959, "num_tokens": 618332968.0, "step": 16286 }, { "epoch": 2.0718738074036382, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.600635051727295, "learning_rate": 1e-05, "loss": 0.3663, "mean_token_accuracy": 0.8926457166671753, "num_tokens": 618369585.0, "step": 16287 }, { "epoch": 2.0720010176822288, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7566049098968506, "learning_rate": 1e-05, "loss": 0.3618, "mean_token_accuracy": 0.8974416255950928, "num_tokens": 618410388.0, "step": 16288 }, { "epoch": 2.0721282279608193, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5825703144073486, "learning_rate": 1e-05, "loss": 0.3565, "mean_token_accuracy": 0.8978451490402222, "num_tokens": 618453591.0, "step": 16289 }, { "epoch": 2.07225543823941, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.777538537979126, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8866020441055298, "num_tokens": 618489757.0, "step": 16290 }, { "epoch": 2.0723826485180004, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5930652618408203, "learning_rate": 1e-05, "loss": 0.3818, "mean_token_accuracy": 0.8880861401557922, "num_tokens": 618524733.0, "step": 16291 }, { "epoch": 2.072509858796591, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.774266481399536, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8790702819824219, "num_tokens": 618564734.0, "step": 16292 }, { "epoch": 2.0726370690751814, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.592651605606079, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8854460120201111, "num_tokens": 618602503.0, "step": 16293 }, { "epoch": 2.072764279353772, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7760870456695557, "learning_rate": 1e-05, "loss": 0.3696, "mean_token_accuracy": 0.8954062461853027, "num_tokens": 618639907.0, "step": 16294 }, { "epoch": 2.0728914896323625, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.600308895111084, "learning_rate": 1e-05, "loss": 0.393, "mean_token_accuracy": 0.8872562646865845, "num_tokens": 618681366.0, "step": 16295 }, { "epoch": 2.073018699910953, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.765237808227539, "learning_rate": 1e-05, "loss": 0.3552, "mean_token_accuracy": 0.8979891538619995, "num_tokens": 618718610.0, "step": 16296 }, { "epoch": 2.0731459101895435, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.608907699584961, "learning_rate": 1e-05, "loss": 0.3774, "mean_token_accuracy": 0.8893378973007202, "num_tokens": 618757741.0, "step": 16297 }, { "epoch": 2.0732731204681336, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7608439922332764, "learning_rate": 1e-05, "loss": 0.3796, "mean_token_accuracy": 0.890362024307251, "num_tokens": 618798189.0, "step": 16298 }, { "epoch": 2.073400330746724, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.605414390563965, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8810609579086304, "num_tokens": 618837308.0, "step": 16299 }, { "epoch": 2.0735275410253147, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7660977840423584, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8691160678863525, "num_tokens": 618876058.0, "step": 16300 }, { "epoch": 2.073654751303905, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5999491214752197, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8740373849868774, "num_tokens": 618916246.0, "step": 16301 }, { "epoch": 2.0737819615824957, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7552056312561035, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8762699961662292, "num_tokens": 618957639.0, "step": 16302 }, { "epoch": 2.0739091718610863, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.590362071990967, "learning_rate": 1e-05, "loss": 0.3831, "mean_token_accuracy": 0.891315758228302, "num_tokens": 618999505.0, "step": 16303 }, { "epoch": 2.074036382139677, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.767749547958374, "learning_rate": 1e-05, "loss": 0.3824, "mean_token_accuracy": 0.8914211988449097, "num_tokens": 619034807.0, "step": 16304 }, { "epoch": 2.0741635924182673, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5891616344451904, "learning_rate": 1e-05, "loss": 0.3767, "mean_token_accuracy": 0.8926311731338501, "num_tokens": 619073262.0, "step": 16305 }, { "epoch": 2.074290802696858, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7746431827545166, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.875037431716919, "num_tokens": 619112706.0, "step": 16306 }, { "epoch": 2.0744180129754484, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.944624423980713, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8749313354492188, "num_tokens": 619144444.0, "step": 16307 }, { "epoch": 2.074545223254039, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.636134624481201, "learning_rate": 1e-05, "loss": 0.378, "mean_token_accuracy": 0.8915545344352722, "num_tokens": 619183291.0, "step": 16308 }, { "epoch": 2.0746724335326294, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.722388982772827, "learning_rate": 1e-05, "loss": 0.3803, "mean_token_accuracy": 0.8917582035064697, "num_tokens": 619223865.0, "step": 16309 }, { "epoch": 2.07479964381122, "ewc_loss": 0.07080078125, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6095130443573, "learning_rate": 1e-05, "loss": 0.3692, "mean_token_accuracy": 0.8917982578277588, "num_tokens": 619257418.0, "step": 16310 }, { "epoch": 2.0749268540898105, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.745973825454712, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8776565790176392, "num_tokens": 619298724.0, "step": 16311 }, { "epoch": 2.075054064368401, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.616595983505249, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8848610520362854, "num_tokens": 619332497.0, "step": 16312 }, { "epoch": 2.0751812746469915, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.756387233734131, "learning_rate": 1e-05, "loss": 0.3945, "mean_token_accuracy": 0.8832758665084839, "num_tokens": 619363371.0, "step": 16313 }, { "epoch": 2.075308484925582, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.983379602432251, "learning_rate": 1e-05, "loss": 0.3857, "mean_token_accuracy": 0.8868486881256104, "num_tokens": 619396068.0, "step": 16314 }, { "epoch": 2.0754356952041726, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6688575744628906, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8812927007675171, "num_tokens": 619433728.0, "step": 16315 }, { "epoch": 2.075562905482763, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7215933799743652, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8732832074165344, "num_tokens": 619476185.0, "step": 16316 }, { "epoch": 2.0756901157613536, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.587622880935669, "learning_rate": 1e-05, "loss": 0.3442, "mean_token_accuracy": 0.9017142653465271, "num_tokens": 619513184.0, "step": 16317 }, { "epoch": 2.075817326039944, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.768242597579956, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8765701055526733, "num_tokens": 619550168.0, "step": 16318 }, { "epoch": 2.0759445363185347, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.598031997680664, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.879650354385376, "num_tokens": 619588960.0, "step": 16319 }, { "epoch": 2.0760717465971252, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000301361083984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.769542932510376, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8849605321884155, "num_tokens": 619624907.0, "step": 16320 }, { "epoch": 2.0761989568757158, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5823347568511963, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.8896345496177673, "num_tokens": 619670576.0, "step": 16321 }, { "epoch": 2.0763261671543063, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7792928218841553, "learning_rate": 1e-05, "loss": 0.3822, "mean_token_accuracy": 0.8898254632949829, "num_tokens": 619710452.0, "step": 16322 }, { "epoch": 2.0764533774328964, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.600806474685669, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8838635683059692, "num_tokens": 619750104.0, "step": 16323 }, { "epoch": 2.076580587711487, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7780544757843018, "learning_rate": 1e-05, "loss": 0.3932, "mean_token_accuracy": 0.8890790939331055, "num_tokens": 619791364.0, "step": 16324 }, { "epoch": 2.0767077979900774, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6016488075256348, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8789367079734802, "num_tokens": 619831138.0, "step": 16325 }, { "epoch": 2.076835008268668, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7986865043640137, "learning_rate": 1e-05, "loss": 0.4003, "mean_token_accuracy": 0.8863033056259155, "num_tokens": 619862897.0, "step": 16326 }, { "epoch": 2.0769622185472585, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003032684326171875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5992581844329834, "learning_rate": 1e-05, "loss": 0.376, "mean_token_accuracy": 0.8930214643478394, "num_tokens": 619897096.0, "step": 16327 }, { "epoch": 2.077089428825849, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7810072898864746, "learning_rate": 1e-05, "loss": 0.371, "mean_token_accuracy": 0.8962579369544983, "num_tokens": 619937458.0, "step": 16328 }, { "epoch": 2.0772166391044395, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.615145683288574, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8718868494033813, "num_tokens": 619974735.0, "step": 16329 }, { "epoch": 2.07734384938303, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.784656286239624, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8791395425796509, "num_tokens": 620012497.0, "step": 16330 }, { "epoch": 2.0774710596616206, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.602022886276245, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8773646354675293, "num_tokens": 620051120.0, "step": 16331 }, { "epoch": 2.077598269940211, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7716896533966064, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.882683277130127, "num_tokens": 620091479.0, "step": 16332 }, { "epoch": 2.0777254802188017, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.609321117401123, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8838233947753906, "num_tokens": 620126509.0, "step": 16333 }, { "epoch": 2.077852690497392, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7726848125457764, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8856964111328125, "num_tokens": 620164278.0, "step": 16334 }, { "epoch": 2.0779799007759827, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6097018718719482, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.874457061290741, "num_tokens": 620202388.0, "step": 16335 }, { "epoch": 2.0781071110545732, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.767889976501465, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8865621089935303, "num_tokens": 620236981.0, "step": 16336 }, { "epoch": 2.0782343213331638, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6011500358581543, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8865023851394653, "num_tokens": 620275946.0, "step": 16337 }, { "epoch": 2.0783615316117543, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.764392375946045, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8693737387657166, "num_tokens": 620317385.0, "step": 16338 }, { "epoch": 2.078488741890345, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.597790241241455, "learning_rate": 1e-05, "loss": 0.38, "mean_token_accuracy": 0.8891794681549072, "num_tokens": 620353625.0, "step": 16339 }, { "epoch": 2.0786159521689354, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.791527032852173, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8646321296691895, "num_tokens": 620389421.0, "step": 16340 }, { "epoch": 2.078743162447526, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.587644338607788, "learning_rate": 1e-05, "loss": 0.3905, "mean_token_accuracy": 0.8867975473403931, "num_tokens": 620428233.0, "step": 16341 }, { "epoch": 2.0788703727261164, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8113303184509277, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8642762303352356, "num_tokens": 620465156.0, "step": 16342 }, { "epoch": 2.078997583004707, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.592947244644165, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8821481466293335, "num_tokens": 620501731.0, "step": 16343 }, { "epoch": 2.0791247932832975, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.806487798690796, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8793390393257141, "num_tokens": 620538507.0, "step": 16344 }, { "epoch": 2.079252003561888, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5798871517181396, "learning_rate": 1e-05, "loss": 0.3542, "mean_token_accuracy": 0.8965840339660645, "num_tokens": 620577355.0, "step": 16345 }, { "epoch": 2.0793792138404785, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8195605278015137, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.877905011177063, "num_tokens": 620613320.0, "step": 16346 }, { "epoch": 2.0795064241190686, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5739023685455322, "learning_rate": 1e-05, "loss": 0.3618, "mean_token_accuracy": 0.8942063450813293, "num_tokens": 620655167.0, "step": 16347 }, { "epoch": 2.079633634397659, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.830629825592041, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.867658257484436, "num_tokens": 620697898.0, "step": 16348 }, { "epoch": 2.0797608446762497, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5920774936676025, "learning_rate": 1e-05, "loss": 0.3914, "mean_token_accuracy": 0.8874303102493286, "num_tokens": 620736548.0, "step": 16349 }, { "epoch": 2.07988805495484, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8221607208251953, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8718789219856262, "num_tokens": 620781368.0, "step": 16350 }, { "epoch": 2.0800152652334307, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6117541790008545, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.8832833766937256, "num_tokens": 620821749.0, "step": 16351 }, { "epoch": 2.0801424755120212, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8136355876922607, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8799399137496948, "num_tokens": 620857906.0, "step": 16352 }, { "epoch": 2.0802696857906118, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6147890090942383, "learning_rate": 1e-05, "loss": 0.4061, "mean_token_accuracy": 0.8842819333076477, "num_tokens": 620897124.0, "step": 16353 }, { "epoch": 2.0803968960692023, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8195595741271973, "learning_rate": 1e-05, "loss": 0.4571, "mean_token_accuracy": 0.8724397420883179, "num_tokens": 620932123.0, "step": 16354 }, { "epoch": 2.080524106347793, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.623356342315674, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8839153051376343, "num_tokens": 620970906.0, "step": 16355 }, { "epoch": 2.0806513166263834, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7974774837493896, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8742728233337402, "num_tokens": 621007387.0, "step": 16356 }, { "epoch": 2.080778526904974, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6176323890686035, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8848850727081299, "num_tokens": 621043723.0, "step": 16357 }, { "epoch": 2.0809057371835644, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.771383762359619, "learning_rate": 1e-05, "loss": 0.3771, "mean_token_accuracy": 0.8913261890411377, "num_tokens": 621083686.0, "step": 16358 }, { "epoch": 2.081032947462155, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6192758083343506, "learning_rate": 1e-05, "loss": 0.4012, "mean_token_accuracy": 0.883026659488678, "num_tokens": 621125758.0, "step": 16359 }, { "epoch": 2.0811601577407455, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7672905921936035, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.883709192276001, "num_tokens": 621163407.0, "step": 16360 }, { "epoch": 2.081287368019336, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.617649793624878, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8768491744995117, "num_tokens": 621203321.0, "step": 16361 }, { "epoch": 2.0814145782979265, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7646377086639404, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8787314891815186, "num_tokens": 621239777.0, "step": 16362 }, { "epoch": 2.081541788576517, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.627159357070923, "learning_rate": 1e-05, "loss": 0.3874, "mean_token_accuracy": 0.8872498273849487, "num_tokens": 621271841.0, "step": 16363 }, { "epoch": 2.0816689988551076, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.741774559020996, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8853682279586792, "num_tokens": 621309393.0, "step": 16364 }, { "epoch": 2.081796209133698, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6382009983062744, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8704473972320557, "num_tokens": 621344362.0, "step": 16365 }, { "epoch": 2.0819234194122886, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7415239810943604, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8649007081985474, "num_tokens": 621383999.0, "step": 16366 }, { "epoch": 2.082050629690879, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.612668752670288, "learning_rate": 1e-05, "loss": 0.3829, "mean_token_accuracy": 0.8888993263244629, "num_tokens": 621424475.0, "step": 16367 }, { "epoch": 2.0821778399694697, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.743199348449707, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8714256882667542, "num_tokens": 621458241.0, "step": 16368 }, { "epoch": 2.0823050502480602, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.606201648712158, "learning_rate": 1e-05, "loss": 0.3768, "mean_token_accuracy": 0.8919179439544678, "num_tokens": 621495424.0, "step": 16369 }, { "epoch": 2.0824322605266508, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.752480983734131, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8848830461502075, "num_tokens": 621531256.0, "step": 16370 }, { "epoch": 2.082559470805241, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.593427896499634, "learning_rate": 1e-05, "loss": 0.399, "mean_token_accuracy": 0.8835570216178894, "num_tokens": 621571665.0, "step": 16371 }, { "epoch": 2.0826866810838314, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.77000093460083, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8873473405838013, "num_tokens": 621604552.0, "step": 16372 }, { "epoch": 2.082813891362422, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5931663513183594, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8640079498291016, "num_tokens": 621651677.0, "step": 16373 }, { "epoch": 2.0829411016410124, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7618324756622314, "learning_rate": 1e-05, "loss": 0.3845, "mean_token_accuracy": 0.8909583687782288, "num_tokens": 621695108.0, "step": 16374 }, { "epoch": 2.083068311919603, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6079397201538086, "learning_rate": 1e-05, "loss": 0.3758, "mean_token_accuracy": 0.8927236795425415, "num_tokens": 621725768.0, "step": 16375 }, { "epoch": 2.0831955221981935, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.754673480987549, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8742411732673645, "num_tokens": 621768559.0, "step": 16376 }, { "epoch": 2.083322732476784, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6125564575195312, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8821152448654175, "num_tokens": 621805464.0, "step": 16377 }, { "epoch": 2.0834499427553745, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.758028507232666, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8740066885948181, "num_tokens": 621843129.0, "step": 16378 }, { "epoch": 2.083577153033965, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6152615547180176, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8822598457336426, "num_tokens": 621882450.0, "step": 16379 }, { "epoch": 2.0837043633125556, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7660319805145264, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8807953596115112, "num_tokens": 621915797.0, "step": 16380 }, { "epoch": 2.083831573591146, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.610527276992798, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8729525804519653, "num_tokens": 621953563.0, "step": 16381 }, { "epoch": 2.0839587838697367, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.750687837600708, "learning_rate": 1e-05, "loss": 0.374, "mean_token_accuracy": 0.8953638076782227, "num_tokens": 621989579.0, "step": 16382 }, { "epoch": 2.084085994148327, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.597715377807617, "learning_rate": 1e-05, "loss": 0.3595, "mean_token_accuracy": 0.896973729133606, "num_tokens": 622027446.0, "step": 16383 }, { "epoch": 2.0842132044269177, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.765150547027588, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8716661930084229, "num_tokens": 622064972.0, "step": 16384 }, { "epoch": 2.0843404147055082, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5907602310180664, "learning_rate": 1e-05, "loss": 0.3712, "mean_token_accuracy": 0.8935686945915222, "num_tokens": 622103260.0, "step": 16385 }, { "epoch": 2.0844676249840988, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7803943157196045, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.8874186277389526, "num_tokens": 622140894.0, "step": 16386 }, { "epoch": 2.0845948352626893, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5938076972961426, "learning_rate": 1e-05, "loss": 0.3871, "mean_token_accuracy": 0.889331579208374, "num_tokens": 622177537.0, "step": 16387 }, { "epoch": 2.08472204554128, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.808648109436035, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8860172033309937, "num_tokens": 622211172.0, "step": 16388 }, { "epoch": 2.0848492558198704, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5964243412017822, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8803154826164246, "num_tokens": 622244805.0, "step": 16389 }, { "epoch": 2.084976466098461, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.799058675765991, "learning_rate": 1e-05, "loss": 0.3769, "mean_token_accuracy": 0.8940505981445312, "num_tokens": 622279496.0, "step": 16390 }, { "epoch": 2.0851036763770514, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.584221124649048, "learning_rate": 1e-05, "loss": 0.3807, "mean_token_accuracy": 0.8918308615684509, "num_tokens": 622318818.0, "step": 16391 }, { "epoch": 2.085230886655642, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7933483123779297, "learning_rate": 1e-05, "loss": 0.39, "mean_token_accuracy": 0.8878028988838196, "num_tokens": 622359313.0, "step": 16392 }, { "epoch": 2.0853580969342325, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6120426654815674, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8787471055984497, "num_tokens": 622392840.0, "step": 16393 }, { "epoch": 2.085485307212823, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7926650047302246, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8692666888237, "num_tokens": 622433522.0, "step": 16394 }, { "epoch": 2.0856125174914135, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.613023281097412, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8839994668960571, "num_tokens": 622468805.0, "step": 16395 }, { "epoch": 2.0857397277700036, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7684526443481445, "learning_rate": 1e-05, "loss": 0.3783, "mean_token_accuracy": 0.8926722407341003, "num_tokens": 622511462.0, "step": 16396 }, { "epoch": 2.085866938048594, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.615241765975952, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8849273920059204, "num_tokens": 622546246.0, "step": 16397 }, { "epoch": 2.0859941483271847, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.765799045562744, "learning_rate": 1e-05, "loss": 0.3803, "mean_token_accuracy": 0.8907510042190552, "num_tokens": 622583113.0, "step": 16398 }, { "epoch": 2.086121358605775, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6065664291381836, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8805835247039795, "num_tokens": 622621414.0, "step": 16399 }, { "epoch": 2.0862485688843657, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7728521823883057, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8661708831787109, "num_tokens": 622660079.0, "step": 16400 }, { "epoch": 2.0863757791629562, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6054227352142334, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8808276653289795, "num_tokens": 622700413.0, "step": 16401 }, { "epoch": 2.0865029894415468, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.768483877182007, "learning_rate": 1e-05, "loss": 0.4008, "mean_token_accuracy": 0.8879845142364502, "num_tokens": 622736448.0, "step": 16402 }, { "epoch": 2.0866301997201373, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6013622283935547, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8833508491516113, "num_tokens": 622777014.0, "step": 16403 }, { "epoch": 2.086757409998728, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7568085193634033, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8854647278785706, "num_tokens": 622817608.0, "step": 16404 }, { "epoch": 2.0868846202773184, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6112616062164307, "learning_rate": 1e-05, "loss": 0.3975, "mean_token_accuracy": 0.8823980093002319, "num_tokens": 622853090.0, "step": 16405 }, { "epoch": 2.087011830555909, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7592129707336426, "learning_rate": 1e-05, "loss": 0.3641, "mean_token_accuracy": 0.8974496126174927, "num_tokens": 622891843.0, "step": 16406 }, { "epoch": 2.0871390408344994, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.616915702819824, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8611946702003479, "num_tokens": 622930364.0, "step": 16407 }, { "epoch": 2.08726625111309, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.761915445327759, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.863952100276947, "num_tokens": 622970199.0, "step": 16408 }, { "epoch": 2.0873934613916805, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6213104724884033, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8761986494064331, "num_tokens": 623003669.0, "step": 16409 }, { "epoch": 2.087520671670271, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.752984046936035, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8677862882614136, "num_tokens": 623043110.0, "step": 16410 }, { "epoch": 2.0876478819488615, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6219687461853027, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8654767274856567, "num_tokens": 623082474.0, "step": 16411 }, { "epoch": 2.087775092227452, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7526984214782715, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8672288060188293, "num_tokens": 623119650.0, "step": 16412 }, { "epoch": 2.0879023025060426, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6401100158691406, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8705828189849854, "num_tokens": 623150359.0, "step": 16413 }, { "epoch": 2.088029512784633, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7268502712249756, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.8827283382415771, "num_tokens": 623189251.0, "step": 16414 }, { "epoch": 2.0881567230632236, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6090641021728516, "learning_rate": 1e-05, "loss": 0.3648, "mean_token_accuracy": 0.8952120542526245, "num_tokens": 623231763.0, "step": 16415 }, { "epoch": 2.088283933341814, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.721034526824951, "learning_rate": 1e-05, "loss": 0.3851, "mean_token_accuracy": 0.8897771835327148, "num_tokens": 623269322.0, "step": 16416 }, { "epoch": 2.0884111436204047, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6293790340423584, "learning_rate": 1e-05, "loss": 0.3839, "mean_token_accuracy": 0.8892364501953125, "num_tokens": 623303765.0, "step": 16417 }, { "epoch": 2.0885383538989952, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7104110717773438, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8751261234283447, "num_tokens": 623355484.0, "step": 16418 }, { "epoch": 2.0886655641775858, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.624727964401245, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8785310983657837, "num_tokens": 623399441.0, "step": 16419 }, { "epoch": 2.0887927744561763, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7276370525360107, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8588147163391113, "num_tokens": 623438690.0, "step": 16420 }, { "epoch": 2.0889199847347664, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6298553943634033, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8748822808265686, "num_tokens": 623476291.0, "step": 16421 }, { "epoch": 2.089047195013357, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.708493232727051, "learning_rate": 1e-05, "loss": 0.3834, "mean_token_accuracy": 0.8932635188102722, "num_tokens": 623513830.0, "step": 16422 }, { "epoch": 2.0891744052919474, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6249380111694336, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8760563135147095, "num_tokens": 623554789.0, "step": 16423 }, { "epoch": 2.089301615570538, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 4.86029577255249, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8845045566558838, "num_tokens": 623593646.0, "step": 16424 }, { "epoch": 2.0894288258491285, "ewc_loss": 0.0947265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 6.389617919921875e-05, "grad_norm": 3.173027753829956, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8913321495056152, "num_tokens": 623630646.0, "step": 16425 }, { "epoch": 2.089556036127719, "ewc_loss": 0.0703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.00543212890625e-05, "grad_norm": 2.7058496475219727, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8616850972175598, "num_tokens": 623675488.0, "step": 16426 }, { "epoch": 2.0896832464063095, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.776653528213501, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8713496327400208, "num_tokens": 623719717.0, "step": 16427 }, { "epoch": 2.0898104566849, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 2.8235533237457275, "learning_rate": 1e-05, "loss": 0.3935, "mean_token_accuracy": 0.8877456784248352, "num_tokens": 623753248.0, "step": 16428 }, { "epoch": 2.0899376669634906, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.72853422164917, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8684938549995422, "num_tokens": 623791562.0, "step": 16429 }, { "epoch": 2.090064877242081, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.7835655212402344, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.867776095867157, "num_tokens": 623834187.0, "step": 16430 }, { "epoch": 2.0901920875206716, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7087647914886475, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8682510852813721, "num_tokens": 623869664.0, "step": 16431 }, { "epoch": 2.090319297799262, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.776569128036499, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8797442317008972, "num_tokens": 623907461.0, "step": 16432 }, { "epoch": 2.0904465080778527, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6590514183044434, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8786874413490295, "num_tokens": 623947679.0, "step": 16433 }, { "epoch": 2.0905737183564432, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.751960039138794, "learning_rate": 1e-05, "loss": 0.3921, "mean_token_accuracy": 0.887167751789093, "num_tokens": 623992659.0, "step": 16434 }, { "epoch": 2.0907009286350338, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6518304347991943, "learning_rate": 1e-05, "loss": 0.3745, "mean_token_accuracy": 0.8916391134262085, "num_tokens": 624028305.0, "step": 16435 }, { "epoch": 2.0908281389136243, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7633020877838135, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8831153512001038, "num_tokens": 624062585.0, "step": 16436 }, { "epoch": 2.090955349192215, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.644395112991333, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8853620886802673, "num_tokens": 624100525.0, "step": 16437 }, { "epoch": 2.0910825594708053, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7649309635162354, "learning_rate": 1e-05, "loss": 0.3808, "mean_token_accuracy": 0.8906689882278442, "num_tokens": 624135982.0, "step": 16438 }, { "epoch": 2.091209769749396, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6150641441345215, "learning_rate": 1e-05, "loss": 0.3774, "mean_token_accuracy": 0.890713095664978, "num_tokens": 624172867.0, "step": 16439 }, { "epoch": 2.0913369800279864, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7572503089904785, "learning_rate": 1e-05, "loss": 0.3629, "mean_token_accuracy": 0.895989716053009, "num_tokens": 624208588.0, "step": 16440 }, { "epoch": 2.091464190306577, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.640634059906006, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8770654201507568, "num_tokens": 624241060.0, "step": 16441 }, { "epoch": 2.0915914005851675, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.739149808883667, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8753786683082581, "num_tokens": 624284508.0, "step": 16442 }, { "epoch": 2.091718610863758, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.617133140563965, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8843363523483276, "num_tokens": 624326787.0, "step": 16443 }, { "epoch": 2.0918458211423485, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.752492904663086, "learning_rate": 1e-05, "loss": 0.3825, "mean_token_accuracy": 0.8899262547492981, "num_tokens": 624363375.0, "step": 16444 }, { "epoch": 2.0919730314209386, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6178207397460938, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8752274513244629, "num_tokens": 624405508.0, "step": 16445 }, { "epoch": 2.092100241699529, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7590792179107666, "learning_rate": 1e-05, "loss": 0.3757, "mean_token_accuracy": 0.8913581371307373, "num_tokens": 624438505.0, "step": 16446 }, { "epoch": 2.0922274519781197, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.613887310028076, "learning_rate": 1e-05, "loss": 0.4008, "mean_token_accuracy": 0.8840186595916748, "num_tokens": 624475463.0, "step": 16447 }, { "epoch": 2.09235466225671, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7627408504486084, "learning_rate": 1e-05, "loss": 0.3918, "mean_token_accuracy": 0.8882806301116943, "num_tokens": 624511577.0, "step": 16448 }, { "epoch": 2.0924818725353007, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.602470874786377, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8831955194473267, "num_tokens": 624552213.0, "step": 16449 }, { "epoch": 2.0926090828138912, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.765352725982666, "learning_rate": 1e-05, "loss": 0.3874, "mean_token_accuracy": 0.8877891898155212, "num_tokens": 624589026.0, "step": 16450 }, { "epoch": 2.0927362930924818, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.613440752029419, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.879555344581604, "num_tokens": 624623245.0, "step": 16451 }, { "epoch": 2.0928635033710723, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.76349139213562, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.877982497215271, "num_tokens": 624659713.0, "step": 16452 }, { "epoch": 2.092990713649663, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6108884811401367, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.884835958480835, "num_tokens": 624695539.0, "step": 16453 }, { "epoch": 2.0931179239282534, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7612950801849365, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8758249282836914, "num_tokens": 624735311.0, "step": 16454 }, { "epoch": 2.093245134206844, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.604714870452881, "learning_rate": 1e-05, "loss": 0.4004, "mean_token_accuracy": 0.8864480257034302, "num_tokens": 624771941.0, "step": 16455 }, { "epoch": 2.0933723444854344, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.757136344909668, "learning_rate": 1e-05, "loss": 0.3773, "mean_token_accuracy": 0.8928829431533813, "num_tokens": 624809963.0, "step": 16456 }, { "epoch": 2.093499554764025, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.611133575439453, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8812280297279358, "num_tokens": 624849839.0, "step": 16457 }, { "epoch": 2.0936267650426155, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.764307975769043, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8805137276649475, "num_tokens": 624883920.0, "step": 16458 }, { "epoch": 2.093753975321206, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.609985589981079, "learning_rate": 1e-05, "loss": 0.3858, "mean_token_accuracy": 0.8890865445137024, "num_tokens": 624916024.0, "step": 16459 }, { "epoch": 2.0938811855997965, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7577219009399414, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8847630023956299, "num_tokens": 624950978.0, "step": 16460 }, { "epoch": 2.094008395878387, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5926475524902344, "learning_rate": 1e-05, "loss": 0.3706, "mean_token_accuracy": 0.8902890682220459, "num_tokens": 624992177.0, "step": 16461 }, { "epoch": 2.0941356061569776, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7774009704589844, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.882331907749176, "num_tokens": 625028211.0, "step": 16462 }, { "epoch": 2.094262816435568, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.58123779296875, "learning_rate": 1e-05, "loss": 0.3596, "mean_token_accuracy": 0.8984271287918091, "num_tokens": 625065895.0, "step": 16463 }, { "epoch": 2.0943900267141586, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7893831729888916, "learning_rate": 1e-05, "loss": 0.393, "mean_token_accuracy": 0.8894635438919067, "num_tokens": 625100521.0, "step": 16464 }, { "epoch": 2.094517236992749, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.578718662261963, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8829780220985413, "num_tokens": 625144810.0, "step": 16465 }, { "epoch": 2.0946444472713397, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.819775342941284, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8796288371086121, "num_tokens": 625178001.0, "step": 16466 }, { "epoch": 2.09477165754993, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.595912456512451, "learning_rate": 1e-05, "loss": 0.3946, "mean_token_accuracy": 0.8837041854858398, "num_tokens": 625214857.0, "step": 16467 }, { "epoch": 2.0948988678285207, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8066961765289307, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8849838376045227, "num_tokens": 625253396.0, "step": 16468 }, { "epoch": 2.095026078107111, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.599900722503662, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8677884340286255, "num_tokens": 625291922.0, "step": 16469 }, { "epoch": 2.0951532883857014, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8242666721343994, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8805685043334961, "num_tokens": 625326548.0, "step": 16470 }, { "epoch": 2.095280498664292, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6034746170043945, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.8873294591903687, "num_tokens": 625361354.0, "step": 16471 }, { "epoch": 2.0954077089428824, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8165736198425293, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8717480301856995, "num_tokens": 625400885.0, "step": 16472 }, { "epoch": 2.095534919221473, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.611630916595459, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8847074508666992, "num_tokens": 625438975.0, "step": 16473 }, { "epoch": 2.0956621295000635, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8101956844329834, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8642534613609314, "num_tokens": 625475384.0, "step": 16474 }, { "epoch": 2.095789339778654, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6155807971954346, "learning_rate": 1e-05, "loss": 0.3638, "mean_token_accuracy": 0.8958995342254639, "num_tokens": 625508422.0, "step": 16475 }, { "epoch": 2.0959165500572445, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.789132833480835, "learning_rate": 1e-05, "loss": 0.3798, "mean_token_accuracy": 0.8908651471138, "num_tokens": 625546434.0, "step": 16476 }, { "epoch": 2.096043760335835, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6068973541259766, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.88609778881073, "num_tokens": 625584005.0, "step": 16477 }, { "epoch": 2.0961709706144256, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8203110694885254, "learning_rate": 1e-05, "loss": 0.3652, "mean_token_accuracy": 0.8976121544837952, "num_tokens": 625616713.0, "step": 16478 }, { "epoch": 2.096298180893016, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.600877046585083, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8828077912330627, "num_tokens": 625657738.0, "step": 16479 }, { "epoch": 2.0964253911716066, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8039097785949707, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8785783052444458, "num_tokens": 625693485.0, "step": 16480 }, { "epoch": 2.096552601450197, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.603339433670044, "learning_rate": 1e-05, "loss": 0.3757, "mean_token_accuracy": 0.8896228671073914, "num_tokens": 625727219.0, "step": 16481 }, { "epoch": 2.0966798117287877, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8063573837280273, "learning_rate": 1e-05, "loss": 0.3721, "mean_token_accuracy": 0.8950052261352539, "num_tokens": 625762858.0, "step": 16482 }, { "epoch": 2.0968070220073782, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6087822914123535, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8793221712112427, "num_tokens": 625799752.0, "step": 16483 }, { "epoch": 2.0969342322859688, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7955751419067383, "learning_rate": 1e-05, "loss": 0.3689, "mean_token_accuracy": 0.894086480140686, "num_tokens": 625835453.0, "step": 16484 }, { "epoch": 2.0970614425645593, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6137290000915527, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8680567145347595, "num_tokens": 625875181.0, "step": 16485 }, { "epoch": 2.09718865284315, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.774292469024658, "learning_rate": 1e-05, "loss": 0.3896, "mean_token_accuracy": 0.8936617374420166, "num_tokens": 625918851.0, "step": 16486 }, { "epoch": 2.0973158631217403, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.623944044113159, "learning_rate": 1e-05, "loss": 0.3936, "mean_token_accuracy": 0.8835570812225342, "num_tokens": 625954859.0, "step": 16487 }, { "epoch": 2.097443073400331, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.777761697769165, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8886109590530396, "num_tokens": 625986042.0, "step": 16488 }, { "epoch": 2.0975702836789214, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6308207511901855, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8613019585609436, "num_tokens": 626023604.0, "step": 16489 }, { "epoch": 2.097697493957512, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7537872791290283, "learning_rate": 1e-05, "loss": 0.3591, "mean_token_accuracy": 0.8992915749549866, "num_tokens": 626059193.0, "step": 16490 }, { "epoch": 2.0978247042361025, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6152069568634033, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.876059889793396, "num_tokens": 626099994.0, "step": 16491 }, { "epoch": 2.097951914514693, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.758157253265381, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8821874856948853, "num_tokens": 626133105.0, "step": 16492 }, { "epoch": 2.0980791247932835, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.618325710296631, "learning_rate": 1e-05, "loss": 0.3932, "mean_token_accuracy": 0.8893059492111206, "num_tokens": 626171837.0, "step": 16493 }, { "epoch": 2.0982063350718736, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.744168996810913, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8759210109710693, "num_tokens": 626214057.0, "step": 16494 }, { "epoch": 2.098333545350464, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6150221824645996, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8870766162872314, "num_tokens": 626249046.0, "step": 16495 }, { "epoch": 2.0984607556290547, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7628281116485596, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8712551593780518, "num_tokens": 626283991.0, "step": 16496 }, { "epoch": 2.098587965907645, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.607367753982544, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8790722489356995, "num_tokens": 626321919.0, "step": 16497 }, { "epoch": 2.0987151761862357, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.747655153274536, "learning_rate": 1e-05, "loss": 0.3763, "mean_token_accuracy": 0.8945882320404053, "num_tokens": 626360281.0, "step": 16498 }, { "epoch": 2.0988423864648262, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6061508655548096, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8832968473434448, "num_tokens": 626404314.0, "step": 16499 }, { "epoch": 2.0989695967434168, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.745163679122925, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8676638007164001, "num_tokens": 626452743.0, "step": 16500 }, { "epoch": 2.0990968070220073, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5985238552093506, "learning_rate": 1e-05, "loss": 0.3633, "mean_token_accuracy": 0.8978526592254639, "num_tokens": 626494310.0, "step": 16501 }, { "epoch": 2.099224017300598, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.746990203857422, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8829536437988281, "num_tokens": 626534995.0, "step": 16502 }, { "epoch": 2.0993512275791884, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.601284980773926, "learning_rate": 1e-05, "loss": 0.3779, "mean_token_accuracy": 0.8906601071357727, "num_tokens": 626575084.0, "step": 16503 }, { "epoch": 2.099478437857779, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7580502033233643, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8882786631584167, "num_tokens": 626608922.0, "step": 16504 }, { "epoch": 2.0996056481363694, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5962374210357666, "learning_rate": 1e-05, "loss": 0.3814, "mean_token_accuracy": 0.8897791504859924, "num_tokens": 626645480.0, "step": 16505 }, { "epoch": 2.09973285841496, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7590019702911377, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8750141263008118, "num_tokens": 626687641.0, "step": 16506 }, { "epoch": 2.0998600686935505, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6017284393310547, "learning_rate": 1e-05, "loss": 0.3721, "mean_token_accuracy": 0.8934878706932068, "num_tokens": 626720624.0, "step": 16507 }, { "epoch": 2.099987278972141, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7641727924346924, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8602321743965149, "num_tokens": 626760174.0, "step": 16508 }, { "epoch": 2.1001144892507315, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.594576597213745, "learning_rate": 1e-05, "loss": 0.3491, "mean_token_accuracy": 0.8976655602455139, "num_tokens": 626792970.0, "step": 16509 }, { "epoch": 2.100241699529322, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.779273509979248, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8799759745597839, "num_tokens": 626826793.0, "step": 16510 }, { "epoch": 2.1003689098079126, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5988454818725586, "learning_rate": 1e-05, "loss": 0.3965, "mean_token_accuracy": 0.8846999406814575, "num_tokens": 626862938.0, "step": 16511 }, { "epoch": 2.100496120086503, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7695600986480713, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.882843554019928, "num_tokens": 626900738.0, "step": 16512 }, { "epoch": 2.1006233303650936, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.60231351852417, "learning_rate": 1e-05, "loss": 0.378, "mean_token_accuracy": 0.8914098739624023, "num_tokens": 626932670.0, "step": 16513 }, { "epoch": 2.100750540643684, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.776618003845215, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8794015645980835, "num_tokens": 626971683.0, "step": 16514 }, { "epoch": 2.1008777509222747, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6024296283721924, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8754891753196716, "num_tokens": 627009955.0, "step": 16515 }, { "epoch": 2.101004961200865, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7702603340148926, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8697733879089355, "num_tokens": 627052469.0, "step": 16516 }, { "epoch": 2.1011321714794557, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.611943483352661, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8798867464065552, "num_tokens": 627088314.0, "step": 16517 }, { "epoch": 2.1012593817580463, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7559971809387207, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8824722766876221, "num_tokens": 627132030.0, "step": 16518 }, { "epoch": 2.1013865920366364, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.612107515335083, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8789283037185669, "num_tokens": 627171135.0, "step": 16519 }, { "epoch": 2.101513802315227, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.760880708694458, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.884380042552948, "num_tokens": 627206303.0, "step": 16520 }, { "epoch": 2.1016410125938174, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.621873617172241, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8795976638793945, "num_tokens": 627242859.0, "step": 16521 }, { "epoch": 2.101768222872408, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7515852451324463, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8786085844039917, "num_tokens": 627284659.0, "step": 16522 }, { "epoch": 2.1018954331509985, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.618373394012451, "learning_rate": 1e-05, "loss": 0.3908, "mean_token_accuracy": 0.8885927200317383, "num_tokens": 627320012.0, "step": 16523 }, { "epoch": 2.102022643429589, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.742427110671997, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.8867975473403931, "num_tokens": 627358117.0, "step": 16524 }, { "epoch": 2.1021498537081795, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.637129306793213, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8710796236991882, "num_tokens": 627393331.0, "step": 16525 }, { "epoch": 2.10227706398677, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.728034019470215, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.872451901435852, "num_tokens": 627433905.0, "step": 16526 }, { "epoch": 2.1024042742653606, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6191117763519287, "learning_rate": 1e-05, "loss": 0.3764, "mean_token_accuracy": 0.8943504095077515, "num_tokens": 627469673.0, "step": 16527 }, { "epoch": 2.102531484543951, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.744490623474121, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8781367540359497, "num_tokens": 627502917.0, "step": 16528 }, { "epoch": 2.1026586948225416, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6233744621276855, "learning_rate": 1e-05, "loss": 0.3721, "mean_token_accuracy": 0.8946874141693115, "num_tokens": 627536632.0, "step": 16529 }, { "epoch": 2.102785905101132, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00030517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7288615703582764, "learning_rate": 1e-05, "loss": 0.3858, "mean_token_accuracy": 0.8895019888877869, "num_tokens": 627580820.0, "step": 16530 }, { "epoch": 2.1029131153797227, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6156718730926514, "learning_rate": 1e-05, "loss": 0.3556, "mean_token_accuracy": 0.9001157283782959, "num_tokens": 627616895.0, "step": 16531 }, { "epoch": 2.1030403256583132, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.732360601425171, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8762571811676025, "num_tokens": 627655648.0, "step": 16532 }, { "epoch": 2.1031675359369038, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.620457172393799, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8805012702941895, "num_tokens": 627692259.0, "step": 16533 }, { "epoch": 2.1032947462154943, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7349257469177246, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8860322833061218, "num_tokens": 627731601.0, "step": 16534 }, { "epoch": 2.103421956494085, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.622765064239502, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8814704418182373, "num_tokens": 627768381.0, "step": 16535 }, { "epoch": 2.1035491667726753, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7308058738708496, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.888560950756073, "num_tokens": 627808327.0, "step": 16536 }, { "epoch": 2.103676377051266, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6281118392944336, "learning_rate": 1e-05, "loss": 0.3896, "mean_token_accuracy": 0.8842275142669678, "num_tokens": 627838779.0, "step": 16537 }, { "epoch": 2.1038035873298564, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.741126775741577, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8818385004997253, "num_tokens": 627875634.0, "step": 16538 }, { "epoch": 2.103930797608447, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6095876693725586, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8815644979476929, "num_tokens": 627918350.0, "step": 16539 }, { "epoch": 2.1040580078870375, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7289938926696777, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8790422677993774, "num_tokens": 627955649.0, "step": 16540 }, { "epoch": 2.104185218165628, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.633803606033325, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8820823431015015, "num_tokens": 627989374.0, "step": 16541 }, { "epoch": 2.1043124284442185, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.937619209289551, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8715431690216064, "num_tokens": 628026338.0, "step": 16542 }, { "epoch": 2.1044396387228086, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5799570083618164, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8874837160110474, "num_tokens": 628062550.0, "step": 16543 }, { "epoch": 2.104566849001399, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8221142292022705, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8765187859535217, "num_tokens": 628104868.0, "step": 16544 }, { "epoch": 2.1046940592799896, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.566714286804199, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8771830201148987, "num_tokens": 628142009.0, "step": 16545 }, { "epoch": 2.10482126955858, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8368163108825684, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.878953754901886, "num_tokens": 628180992.0, "step": 16546 }, { "epoch": 2.1049484798371707, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5907392501831055, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8826388120651245, "num_tokens": 628215636.0, "step": 16547 }, { "epoch": 2.1050756901157612, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8276031017303467, "learning_rate": 1e-05, "loss": 0.3848, "mean_token_accuracy": 0.8890863656997681, "num_tokens": 628256993.0, "step": 16548 }, { "epoch": 2.1052029003943518, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5987961292266846, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8789944648742676, "num_tokens": 628293408.0, "step": 16549 }, { "epoch": 2.1053301106729423, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.8344051837921143, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8845027089118958, "num_tokens": 628329353.0, "step": 16550 }, { "epoch": 2.105457320951533, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6206088066101074, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8707476854324341, "num_tokens": 628364312.0, "step": 16551 }, { "epoch": 2.1055845312301233, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8109538555145264, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.8855394124984741, "num_tokens": 628401354.0, "step": 16552 }, { "epoch": 2.105711741508714, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6304962635040283, "learning_rate": 1e-05, "loss": 0.3852, "mean_token_accuracy": 0.8892983198165894, "num_tokens": 628432840.0, "step": 16553 }, { "epoch": 2.1058389517873044, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.790328025817871, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8797751069068909, "num_tokens": 628476300.0, "step": 16554 }, { "epoch": 2.105966162065895, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.619107961654663, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8822887539863586, "num_tokens": 628514248.0, "step": 16555 }, { "epoch": 2.1060933723444855, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7713513374328613, "learning_rate": 1e-05, "loss": 0.3983, "mean_token_accuracy": 0.8856978416442871, "num_tokens": 628557161.0, "step": 16556 }, { "epoch": 2.106220582623076, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6290855407714844, "learning_rate": 1e-05, "loss": 0.3898, "mean_token_accuracy": 0.8869175910949707, "num_tokens": 628591073.0, "step": 16557 }, { "epoch": 2.1063477929016665, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.761120319366455, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8875775933265686, "num_tokens": 628627215.0, "step": 16558 }, { "epoch": 2.106475003180257, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.617931365966797, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8793537020683289, "num_tokens": 628669858.0, "step": 16559 }, { "epoch": 2.1066022134588476, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.76359486579895, "learning_rate": 1e-05, "loss": 0.3613, "mean_token_accuracy": 0.8975059986114502, "num_tokens": 628704504.0, "step": 16560 }, { "epoch": 2.106729423737438, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.626420259475708, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8679847717285156, "num_tokens": 628743444.0, "step": 16561 }, { "epoch": 2.1068566340160286, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7596287727355957, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8690909147262573, "num_tokens": 628782256.0, "step": 16562 }, { "epoch": 2.106983844294619, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.61812424659729, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.8861817121505737, "num_tokens": 628822150.0, "step": 16563 }, { "epoch": 2.1071110545732097, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.742058753967285, "learning_rate": 1e-05, "loss": 0.3795, "mean_token_accuracy": 0.8929288983345032, "num_tokens": 628862445.0, "step": 16564 }, { "epoch": 2.1072382648518, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6306962966918945, "learning_rate": 1e-05, "loss": 0.3956, "mean_token_accuracy": 0.8828995227813721, "num_tokens": 628897166.0, "step": 16565 }, { "epoch": 2.1073654751303907, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.741760015487671, "learning_rate": 1e-05, "loss": 0.3901, "mean_token_accuracy": 0.8867141008377075, "num_tokens": 628933443.0, "step": 16566 }, { "epoch": 2.107492685408981, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.619936466217041, "learning_rate": 1e-05, "loss": 0.3797, "mean_token_accuracy": 0.8900243639945984, "num_tokens": 628969614.0, "step": 16567 }, { "epoch": 2.1076198956875714, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.72550892829895, "learning_rate": 1e-05, "loss": 0.3945, "mean_token_accuracy": 0.886719822883606, "num_tokens": 629010386.0, "step": 16568 }, { "epoch": 2.107747105966162, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6520607471466064, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8754885792732239, "num_tokens": 629041388.0, "step": 16569 }, { "epoch": 2.1078743162447524, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.720261812210083, "learning_rate": 1e-05, "loss": 0.3749, "mean_token_accuracy": 0.893438458442688, "num_tokens": 629077306.0, "step": 16570 }, { "epoch": 2.108001526523343, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6401829719543457, "learning_rate": 1e-05, "loss": 0.3816, "mean_token_accuracy": 0.8907489776611328, "num_tokens": 629110636.0, "step": 16571 }, { "epoch": 2.1081287368019335, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7043302059173584, "learning_rate": 1e-05, "loss": 0.3569, "mean_token_accuracy": 0.9010509252548218, "num_tokens": 629145834.0, "step": 16572 }, { "epoch": 2.108255947080524, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6286427974700928, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8796523809432983, "num_tokens": 629187148.0, "step": 16573 }, { "epoch": 2.1083831573591145, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7047204971313477, "learning_rate": 1e-05, "loss": 0.3767, "mean_token_accuracy": 0.8938567638397217, "num_tokens": 629224391.0, "step": 16574 }, { "epoch": 2.108510367637705, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.643267869949341, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.885119616985321, "num_tokens": 629261074.0, "step": 16575 }, { "epoch": 2.1086375779162956, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7042958736419678, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8828533887863159, "num_tokens": 629296568.0, "step": 16576 }, { "epoch": 2.108764788194886, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6183388233184814, "learning_rate": 1e-05, "loss": 0.3954, "mean_token_accuracy": 0.8859339952468872, "num_tokens": 629342784.0, "step": 16577 }, { "epoch": 2.1088919984734766, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.711883783340454, "learning_rate": 1e-05, "loss": 0.3859, "mean_token_accuracy": 0.8894102573394775, "num_tokens": 629377705.0, "step": 16578 }, { "epoch": 2.109019208752067, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.617072105407715, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.8865255117416382, "num_tokens": 629414823.0, "step": 16579 }, { "epoch": 2.1091464190306577, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.724277973175049, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8725341558456421, "num_tokens": 629452404.0, "step": 16580 }, { "epoch": 2.109273629309248, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.622464895248413, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8869079947471619, "num_tokens": 629486147.0, "step": 16581 }, { "epoch": 2.1094008395878387, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.719545841217041, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8879092931747437, "num_tokens": 629523581.0, "step": 16582 }, { "epoch": 2.1095280498664293, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.606515645980835, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8827111721038818, "num_tokens": 629562125.0, "step": 16583 }, { "epoch": 2.10965526014502, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7161436080932617, "learning_rate": 1e-05, "loss": 0.3835, "mean_token_accuracy": 0.8919756412506104, "num_tokens": 629601786.0, "step": 16584 }, { "epoch": 2.1097824704236103, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.603684902191162, "learning_rate": 1e-05, "loss": 0.3843, "mean_token_accuracy": 0.8906512260437012, "num_tokens": 629645022.0, "step": 16585 }, { "epoch": 2.109909680702201, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.74076771736145, "learning_rate": 1e-05, "loss": 0.3692, "mean_token_accuracy": 0.8948841094970703, "num_tokens": 629675604.0, "step": 16586 }, { "epoch": 2.1100368909807914, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.598024368286133, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8768903613090515, "num_tokens": 629718508.0, "step": 16587 }, { "epoch": 2.110164101259382, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.747645139694214, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8663718700408936, "num_tokens": 629758476.0, "step": 16588 }, { "epoch": 2.1102913115379724, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.614408493041992, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8758081197738647, "num_tokens": 629793692.0, "step": 16589 }, { "epoch": 2.110418521816563, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.756443977355957, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8743886947631836, "num_tokens": 629829750.0, "step": 16590 }, { "epoch": 2.1105457320951535, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.613959789276123, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8645527362823486, "num_tokens": 629865832.0, "step": 16591 }, { "epoch": 2.1106729423737436, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.755399703979492, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8727700114250183, "num_tokens": 629906450.0, "step": 16592 }, { "epoch": 2.110800152652334, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6421594619750977, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8691454529762268, "num_tokens": 629939650.0, "step": 16593 }, { "epoch": 2.1109273629309246, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7298166751861572, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8833099007606506, "num_tokens": 629977692.0, "step": 16594 }, { "epoch": 2.111054573209515, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6177027225494385, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8851110935211182, "num_tokens": 630019111.0, "step": 16595 }, { "epoch": 2.1111817834881057, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7369298934936523, "learning_rate": 1e-05, "loss": 0.3727, "mean_token_accuracy": 0.8910244107246399, "num_tokens": 630054924.0, "step": 16596 }, { "epoch": 2.1113089937666962, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5997426509857178, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8752857446670532, "num_tokens": 630100344.0, "step": 16597 }, { "epoch": 2.1114362040452868, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7466232776641846, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.882208526134491, "num_tokens": 630137296.0, "step": 16598 }, { "epoch": 2.1115634143238773, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6122210025787354, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8861843943595886, "num_tokens": 630177477.0, "step": 16599 }, { "epoch": 2.111690624602468, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003070831298828125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7576606273651123, "learning_rate": 1e-05, "loss": 0.3542, "mean_token_accuracy": 0.8991889357566833, "num_tokens": 630205830.0, "step": 16600 }, { "epoch": 2.1118178348810583, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.605908155441284, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.8815481662750244, "num_tokens": 630247114.0, "step": 16601 }, { "epoch": 2.111945045159649, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.765467643737793, "learning_rate": 1e-05, "loss": 0.3764, "mean_token_accuracy": 0.8952553868293762, "num_tokens": 630279848.0, "step": 16602 }, { "epoch": 2.1120722554382394, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5884265899658203, "learning_rate": 1e-05, "loss": 0.3878, "mean_token_accuracy": 0.8923624157905579, "num_tokens": 630320993.0, "step": 16603 }, { "epoch": 2.11219946571683, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7964582443237305, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8621413111686707, "num_tokens": 630354703.0, "step": 16604 }, { "epoch": 2.1123266759954205, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.611039161682129, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8779269456863403, "num_tokens": 630388902.0, "step": 16605 }, { "epoch": 2.112453886274011, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7746171951293945, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8845465779304504, "num_tokens": 630431446.0, "step": 16606 }, { "epoch": 2.1125810965526015, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6084141731262207, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8770543336868286, "num_tokens": 630468334.0, "step": 16607 }, { "epoch": 2.112708306831192, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7832064628601074, "learning_rate": 1e-05, "loss": 0.3921, "mean_token_accuracy": 0.8867219090461731, "num_tokens": 630498607.0, "step": 16608 }, { "epoch": 2.1128355171097826, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.61460280418396, "learning_rate": 1e-05, "loss": 0.3888, "mean_token_accuracy": 0.8849406242370605, "num_tokens": 630534900.0, "step": 16609 }, { "epoch": 2.112962727388373, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7637689113616943, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8735693097114563, "num_tokens": 630577045.0, "step": 16610 }, { "epoch": 2.1130899376669636, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6106464862823486, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.8853054642677307, "num_tokens": 630613477.0, "step": 16611 }, { "epoch": 2.113217147945554, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.761962890625, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8827418088912964, "num_tokens": 630651868.0, "step": 16612 }, { "epoch": 2.1133443582241447, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6199564933776855, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8853740692138672, "num_tokens": 630691483.0, "step": 16613 }, { "epoch": 2.113471568502735, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.760049343109131, "learning_rate": 1e-05, "loss": 0.4044, "mean_token_accuracy": 0.8861328959465027, "num_tokens": 630730392.0, "step": 16614 }, { "epoch": 2.1135987787813257, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.618898630142212, "learning_rate": 1e-05, "loss": 0.3908, "mean_token_accuracy": 0.8891196250915527, "num_tokens": 630764336.0, "step": 16615 }, { "epoch": 2.1137259890599163, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7491602897644043, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8744103908538818, "num_tokens": 630805912.0, "step": 16616 }, { "epoch": 2.1138531993385064, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6177453994750977, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.874606728553772, "num_tokens": 630844667.0, "step": 16617 }, { "epoch": 2.113980409617097, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7331390380859375, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.8842396140098572, "num_tokens": 630888217.0, "step": 16618 }, { "epoch": 2.1141076198956874, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6189897060394287, "learning_rate": 1e-05, "loss": 0.3672, "mean_token_accuracy": 0.8922020196914673, "num_tokens": 630922415.0, "step": 16619 }, { "epoch": 2.114234830174278, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7503561973571777, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8842349052429199, "num_tokens": 630955469.0, "step": 16620 }, { "epoch": 2.1143620404528685, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.615891456604004, "learning_rate": 1e-05, "loss": 0.3971, "mean_token_accuracy": 0.8862686157226562, "num_tokens": 630994641.0, "step": 16621 }, { "epoch": 2.114489250731459, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.738487720489502, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8718822002410889, "num_tokens": 631034671.0, "step": 16622 }, { "epoch": 2.1146164610100495, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6187169551849365, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8784099817276001, "num_tokens": 631073518.0, "step": 16623 }, { "epoch": 2.11474367128864, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7488174438476562, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8658798336982727, "num_tokens": 631111631.0, "step": 16624 }, { "epoch": 2.1148708815672306, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.608269214630127, "learning_rate": 1e-05, "loss": 0.4015, "mean_token_accuracy": 0.8804648518562317, "num_tokens": 631148081.0, "step": 16625 }, { "epoch": 2.114998091845821, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.798588275909424, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8757481575012207, "num_tokens": 631183176.0, "step": 16626 }, { "epoch": 2.1151253021244116, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.60001540184021, "learning_rate": 1e-05, "loss": 0.3907, "mean_token_accuracy": 0.8844691514968872, "num_tokens": 631213841.0, "step": 16627 }, { "epoch": 2.115252512403002, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7740280628204346, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8698472380638123, "num_tokens": 631249410.0, "step": 16628 }, { "epoch": 2.1153797226815927, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5845842361450195, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8806675672531128, "num_tokens": 631292153.0, "step": 16629 }, { "epoch": 2.115506932960183, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7906293869018555, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8803337216377258, "num_tokens": 631330461.0, "step": 16630 }, { "epoch": 2.1156341432387737, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5966429710388184, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8773334622383118, "num_tokens": 631367384.0, "step": 16631 }, { "epoch": 2.1157613535173643, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7936997413635254, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8754591941833496, "num_tokens": 631402558.0, "step": 16632 }, { "epoch": 2.115888563795955, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5908203125, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8621024489402771, "num_tokens": 631444040.0, "step": 16633 }, { "epoch": 2.1160157740745453, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7867984771728516, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8898503184318542, "num_tokens": 631486350.0, "step": 16634 }, { "epoch": 2.116142984353136, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6051266193389893, "learning_rate": 1e-05, "loss": 0.3784, "mean_token_accuracy": 0.8890707492828369, "num_tokens": 631519145.0, "step": 16635 }, { "epoch": 2.1162701946317264, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8423912525177, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8765981197357178, "num_tokens": 631554637.0, "step": 16636 }, { "epoch": 2.116397404910317, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.586724042892456, "learning_rate": 1e-05, "loss": 0.3814, "mean_token_accuracy": 0.8900552988052368, "num_tokens": 631593580.0, "step": 16637 }, { "epoch": 2.1165246151889074, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.820039749145508, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.880200982093811, "num_tokens": 631630410.0, "step": 16638 }, { "epoch": 2.116651825467498, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6260488033294678, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8840516805648804, "num_tokens": 631661894.0, "step": 16639 }, { "epoch": 2.116779035746088, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.816843032836914, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8815065622329712, "num_tokens": 631694514.0, "step": 16640 }, { "epoch": 2.1169062460246786, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.601811647415161, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.873886227607727, "num_tokens": 631737715.0, "step": 16641 }, { "epoch": 2.117033456303269, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.811307430267334, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.871009111404419, "num_tokens": 631773972.0, "step": 16642 }, { "epoch": 2.1171606665818596, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.611226797103882, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8780350089073181, "num_tokens": 631812292.0, "step": 16643 }, { "epoch": 2.11728787686045, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.796996831893921, "learning_rate": 1e-05, "loss": 0.379, "mean_token_accuracy": 0.8930314183235168, "num_tokens": 631849518.0, "step": 16644 }, { "epoch": 2.1174150871390407, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.609424591064453, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8834879398345947, "num_tokens": 631889821.0, "step": 16645 }, { "epoch": 2.1175422974176312, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7928049564361572, "learning_rate": 1e-05, "loss": 0.3707, "mean_token_accuracy": 0.8941465616226196, "num_tokens": 631925192.0, "step": 16646 }, { "epoch": 2.1176695076962218, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.609328269958496, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8826547861099243, "num_tokens": 631967876.0, "step": 16647 }, { "epoch": 2.1177967179748123, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.78718638420105, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8862608671188354, "num_tokens": 632006026.0, "step": 16648 }, { "epoch": 2.117923928253403, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6013693809509277, "learning_rate": 1e-05, "loss": 0.3678, "mean_token_accuracy": 0.8959442377090454, "num_tokens": 632045886.0, "step": 16649 }, { "epoch": 2.1180511385319933, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.797973394393921, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.8874486684799194, "num_tokens": 632083583.0, "step": 16650 }, { "epoch": 2.118178348810584, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6052019596099854, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8848170042037964, "num_tokens": 632124427.0, "step": 16651 }, { "epoch": 2.1183055590891744, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8120806217193604, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8780139684677124, "num_tokens": 632159787.0, "step": 16652 }, { "epoch": 2.118432769367765, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.601130962371826, "learning_rate": 1e-05, "loss": 0.3777, "mean_token_accuracy": 0.8894015550613403, "num_tokens": 632194737.0, "step": 16653 }, { "epoch": 2.1185599796463555, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8025753498077393, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8870385885238647, "num_tokens": 632230737.0, "step": 16654 }, { "epoch": 2.118687189924946, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000308990478515625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5980663299560547, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.8794649243354797, "num_tokens": 632271374.0, "step": 16655 }, { "epoch": 2.1188144002035365, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.812065601348877, "learning_rate": 1e-05, "loss": 0.3832, "mean_token_accuracy": 0.8894869685173035, "num_tokens": 632305326.0, "step": 16656 }, { "epoch": 2.118941610482127, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.617469310760498, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.881188690662384, "num_tokens": 632338807.0, "step": 16657 }, { "epoch": 2.1190688207607176, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8134591579437256, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8699806332588196, "num_tokens": 632373933.0, "step": 16658 }, { "epoch": 2.119196031039308, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6063523292541504, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8812013864517212, "num_tokens": 632414149.0, "step": 16659 }, { "epoch": 2.1193232413178986, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7972869873046875, "learning_rate": 1e-05, "loss": 0.3836, "mean_token_accuracy": 0.8884496092796326, "num_tokens": 632453622.0, "step": 16660 }, { "epoch": 2.119450451596489, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6277315616607666, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8693262338638306, "num_tokens": 632488398.0, "step": 16661 }, { "epoch": 2.1195776618750797, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.79299259185791, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8776716589927673, "num_tokens": 632524013.0, "step": 16662 }, { "epoch": 2.11970487215367, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6333436965942383, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8774822950363159, "num_tokens": 632562247.0, "step": 16663 }, { "epoch": 2.1198320824322607, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7843401432037354, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8722379207611084, "num_tokens": 632600066.0, "step": 16664 }, { "epoch": 2.119959292710851, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.625866651535034, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8775249719619751, "num_tokens": 632638157.0, "step": 16665 }, { "epoch": 2.1200865029894413, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.76491117477417, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8596360683441162, "num_tokens": 632682534.0, "step": 16666 }, { "epoch": 2.120213713268032, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6210529804229736, "learning_rate": 1e-05, "loss": 0.3845, "mean_token_accuracy": 0.8892064094543457, "num_tokens": 632717574.0, "step": 16667 }, { "epoch": 2.1203409235466224, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7519805431365967, "learning_rate": 1e-05, "loss": 0.3895, "mean_token_accuracy": 0.8850734233856201, "num_tokens": 632754113.0, "step": 16668 }, { "epoch": 2.120468133825213, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6247894763946533, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8691999912261963, "num_tokens": 632797776.0, "step": 16669 }, { "epoch": 2.1205953441038035, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.747748613357544, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.881868302822113, "num_tokens": 632834841.0, "step": 16670 }, { "epoch": 2.120722554382394, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.621255397796631, "learning_rate": 1e-05, "loss": 0.393, "mean_token_accuracy": 0.8891884088516235, "num_tokens": 632872698.0, "step": 16671 }, { "epoch": 2.1208497646609845, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.748775005340576, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8800318241119385, "num_tokens": 632911067.0, "step": 16672 }, { "epoch": 2.120976974939575, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.613569974899292, "learning_rate": 1e-05, "loss": 0.3968, "mean_token_accuracy": 0.886459231376648, "num_tokens": 632952037.0, "step": 16673 }, { "epoch": 2.1211041852181656, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.761253833770752, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.876891016960144, "num_tokens": 632984667.0, "step": 16674 }, { "epoch": 2.121231395496756, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6063525676727295, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8726516962051392, "num_tokens": 633027165.0, "step": 16675 }, { "epoch": 2.1213586057753466, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7456185817718506, "learning_rate": 1e-05, "loss": 0.3803, "mean_token_accuracy": 0.8930085897445679, "num_tokens": 633068636.0, "step": 16676 }, { "epoch": 2.121485816053937, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6098663806915283, "learning_rate": 1e-05, "loss": 0.3852, "mean_token_accuracy": 0.8898940086364746, "num_tokens": 633103546.0, "step": 16677 }, { "epoch": 2.1216130263325277, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7636349201202393, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.872880756855011, "num_tokens": 633141529.0, "step": 16678 }, { "epoch": 2.121740236611118, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.605832815170288, "learning_rate": 1e-05, "loss": 0.3781, "mean_token_accuracy": 0.8922356963157654, "num_tokens": 633177324.0, "step": 16679 }, { "epoch": 2.1218674468897087, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.756765365600586, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8816855549812317, "num_tokens": 633216726.0, "step": 16680 }, { "epoch": 2.1219946571682993, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.604124069213867, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8772079348564148, "num_tokens": 633255583.0, "step": 16681 }, { "epoch": 2.12212186744689, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7601919174194336, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8859872817993164, "num_tokens": 633295167.0, "step": 16682 }, { "epoch": 2.1222490777254803, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6035959720611572, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8686150312423706, "num_tokens": 633336662.0, "step": 16683 }, { "epoch": 2.122376288004071, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7487432956695557, "learning_rate": 1e-05, "loss": 0.3956, "mean_token_accuracy": 0.886451244354248, "num_tokens": 633379826.0, "step": 16684 }, { "epoch": 2.1225034982826614, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6084201335906982, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8716562986373901, "num_tokens": 633421447.0, "step": 16685 }, { "epoch": 2.122630708561252, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7450473308563232, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8903985619544983, "num_tokens": 633462704.0, "step": 16686 }, { "epoch": 2.1227579188398424, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.612267017364502, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8815770149230957, "num_tokens": 633496452.0, "step": 16687 }, { "epoch": 2.122885129118433, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7343568801879883, "learning_rate": 1e-05, "loss": 0.3629, "mean_token_accuracy": 0.8999781608581543, "num_tokens": 633534560.0, "step": 16688 }, { "epoch": 2.1230123393970235, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6130611896514893, "learning_rate": 1e-05, "loss": 0.3926, "mean_token_accuracy": 0.8870332837104797, "num_tokens": 633572429.0, "step": 16689 }, { "epoch": 2.1231395496756136, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7640180587768555, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8739528656005859, "num_tokens": 633603588.0, "step": 16690 }, { "epoch": 2.123266759954204, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6122727394104004, "learning_rate": 1e-05, "loss": 0.3986, "mean_token_accuracy": 0.885402262210846, "num_tokens": 633640975.0, "step": 16691 }, { "epoch": 2.1233939702327946, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7386789321899414, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8872484564781189, "num_tokens": 633684340.0, "step": 16692 }, { "epoch": 2.123521180511385, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.615501880645752, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8843650221824646, "num_tokens": 633720969.0, "step": 16693 }, { "epoch": 2.1236483907899757, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.751485824584961, "learning_rate": 1e-05, "loss": 0.3656, "mean_token_accuracy": 0.894280195236206, "num_tokens": 633751726.0, "step": 16694 }, { "epoch": 2.123775601068566, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.619539260864258, "learning_rate": 1e-05, "loss": 0.3909, "mean_token_accuracy": 0.8885231018066406, "num_tokens": 633784645.0, "step": 16695 }, { "epoch": 2.1239028113471567, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.745851755142212, "learning_rate": 1e-05, "loss": 0.3618, "mean_token_accuracy": 0.894726037979126, "num_tokens": 633816028.0, "step": 16696 }, { "epoch": 2.1240300216257473, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6076340675354004, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8734134435653687, "num_tokens": 633859018.0, "step": 16697 }, { "epoch": 2.124157231904338, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7470335960388184, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8772478699684143, "num_tokens": 633897549.0, "step": 16698 }, { "epoch": 2.1242844421829283, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6963677406311035, "learning_rate": 1e-05, "loss": 0.3728, "mean_token_accuracy": 0.891247034072876, "num_tokens": 633933111.0, "step": 16699 }, { "epoch": 2.124411652461519, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7065839767456055, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.870630145072937, "num_tokens": 633972409.0, "step": 16700 }, { "epoch": 2.1245388627401094, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.65537691116333, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8649822473526001, "num_tokens": 634007112.0, "step": 16701 }, { "epoch": 2.1246660730187, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6873905658721924, "learning_rate": 1e-05, "loss": 0.3664, "mean_token_accuracy": 0.896409273147583, "num_tokens": 634043509.0, "step": 16702 }, { "epoch": 2.1247932832972904, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.652710437774658, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8827040195465088, "num_tokens": 634081883.0, "step": 16703 }, { "epoch": 2.124920493575881, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6884467601776123, "learning_rate": 1e-05, "loss": 0.3857, "mean_token_accuracy": 0.8916494250297546, "num_tokens": 634116203.0, "step": 16704 }, { "epoch": 2.1250477038544715, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6510043144226074, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8844560384750366, "num_tokens": 634152538.0, "step": 16705 }, { "epoch": 2.125174914133062, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6923420429229736, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8844128251075745, "num_tokens": 634188563.0, "step": 16706 }, { "epoch": 2.1253021244116526, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6413097381591797, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8686275482177734, "num_tokens": 634235164.0, "step": 16707 }, { "epoch": 2.125429334690243, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.684239625930786, "learning_rate": 1e-05, "loss": 0.3923, "mean_token_accuracy": 0.8903025388717651, "num_tokens": 634275136.0, "step": 16708 }, { "epoch": 2.1255565449688336, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6374106407165527, "learning_rate": 1e-05, "loss": 0.3926, "mean_token_accuracy": 0.8872122168540955, "num_tokens": 634312056.0, "step": 16709 }, { "epoch": 2.125683755247424, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7214415073394775, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8668797016143799, "num_tokens": 634345737.0, "step": 16710 }, { "epoch": 2.1258109655260147, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.632370948791504, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8805516958236694, "num_tokens": 634381122.0, "step": 16711 }, { "epoch": 2.125938175804605, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.702129364013672, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8813226819038391, "num_tokens": 634420529.0, "step": 16712 }, { "epoch": 2.1260653860831957, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.614320755004883, "learning_rate": 1e-05, "loss": 0.389, "mean_token_accuracy": 0.8866095542907715, "num_tokens": 634460540.0, "step": 16713 }, { "epoch": 2.1261925963617863, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7352559566497803, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.870876669883728, "num_tokens": 634497818.0, "step": 16714 }, { "epoch": 2.1263198066403763, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6339352130889893, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8751158118247986, "num_tokens": 634532262.0, "step": 16715 }, { "epoch": 2.126447016918967, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7186789512634277, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.880888819694519, "num_tokens": 634567316.0, "step": 16716 }, { "epoch": 2.1265742271975574, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.623957395553589, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8778468370437622, "num_tokens": 634608603.0, "step": 16717 }, { "epoch": 2.126701437476148, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7180685997009277, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8805198669433594, "num_tokens": 634646539.0, "step": 16718 }, { "epoch": 2.1268286477547385, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.642491579055786, "learning_rate": 1e-05, "loss": 0.3817, "mean_token_accuracy": 0.8883935809135437, "num_tokens": 634673998.0, "step": 16719 }, { "epoch": 2.126955858033329, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.722001791000366, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8828314542770386, "num_tokens": 634707609.0, "step": 16720 }, { "epoch": 2.1270830683119195, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6217761039733887, "learning_rate": 1e-05, "loss": 0.398, "mean_token_accuracy": 0.8845906257629395, "num_tokens": 634742893.0, "step": 16721 }, { "epoch": 2.12721027859051, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.712318181991577, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8864221572875977, "num_tokens": 634780887.0, "step": 16722 }, { "epoch": 2.1273374888691006, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6208252906799316, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8843657374382019, "num_tokens": 634819512.0, "step": 16723 }, { "epoch": 2.127464699147691, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7294259071350098, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8671178221702576, "num_tokens": 634857213.0, "step": 16724 }, { "epoch": 2.1275919094262816, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.621666669845581, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8748383522033691, "num_tokens": 634899221.0, "step": 16725 }, { "epoch": 2.127719119704872, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.703484296798706, "learning_rate": 1e-05, "loss": 0.3609, "mean_token_accuracy": 0.8964774012565613, "num_tokens": 634943486.0, "step": 16726 }, { "epoch": 2.1278463299834627, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6247849464416504, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8745965957641602, "num_tokens": 634981303.0, "step": 16727 }, { "epoch": 2.127973540262053, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.762404203414917, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8781055808067322, "num_tokens": 635021411.0, "step": 16728 }, { "epoch": 2.1281007505406437, "ewc_loss": 0.0712890625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6012496948242188, "learning_rate": 1e-05, "loss": 0.386, "mean_token_accuracy": 0.8900449872016907, "num_tokens": 635063449.0, "step": 16729 }, { "epoch": 2.1282279608192343, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7362186908721924, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8696463108062744, "num_tokens": 635107052.0, "step": 16730 }, { "epoch": 2.128355171097825, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6129298210144043, "learning_rate": 1e-05, "loss": 0.3789, "mean_token_accuracy": 0.8910938501358032, "num_tokens": 635141036.0, "step": 16731 }, { "epoch": 2.1284823813764153, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.743584156036377, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.873124361038208, "num_tokens": 635181060.0, "step": 16732 }, { "epoch": 2.128609591655006, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.5997414588928223, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.8802915215492249, "num_tokens": 635222679.0, "step": 16733 }, { "epoch": 2.1287368019335964, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.757397413253784, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.8843218088150024, "num_tokens": 635258351.0, "step": 16734 }, { "epoch": 2.128864012212187, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.608989715576172, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8631253838539124, "num_tokens": 635300647.0, "step": 16735 }, { "epoch": 2.1289912224907774, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7611615657806396, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8812987804412842, "num_tokens": 635335239.0, "step": 16736 }, { "epoch": 2.129118432769368, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6071078777313232, "learning_rate": 1e-05, "loss": 0.4017, "mean_token_accuracy": 0.8863840699195862, "num_tokens": 635370786.0, "step": 16737 }, { "epoch": 2.129245643047958, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.763481378555298, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8816105127334595, "num_tokens": 635410033.0, "step": 16738 }, { "epoch": 2.129372853326549, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.60874342918396, "learning_rate": 1e-05, "loss": 0.3775, "mean_token_accuracy": 0.8930203318595886, "num_tokens": 635447157.0, "step": 16739 }, { "epoch": 2.129500063605139, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7592508792877197, "learning_rate": 1e-05, "loss": 0.3971, "mean_token_accuracy": 0.8893234729766846, "num_tokens": 635485580.0, "step": 16740 }, { "epoch": 2.1296272738837296, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.605069875717163, "learning_rate": 1e-05, "loss": 0.3574, "mean_token_accuracy": 0.8956319093704224, "num_tokens": 635520289.0, "step": 16741 }, { "epoch": 2.12975448416232, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7641589641571045, "learning_rate": 1e-05, "loss": 0.3885, "mean_token_accuracy": 0.8873165249824524, "num_tokens": 635552390.0, "step": 16742 }, { "epoch": 2.1298816944409107, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.611119031906128, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8807913064956665, "num_tokens": 635588281.0, "step": 16743 }, { "epoch": 2.130008904719501, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.762537956237793, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.869461178779602, "num_tokens": 635632229.0, "step": 16744 }, { "epoch": 2.1301361149980917, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.596722364425659, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8855075240135193, "num_tokens": 635671315.0, "step": 16745 }, { "epoch": 2.1302633252766823, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.778046131134033, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8769695162773132, "num_tokens": 635711765.0, "step": 16746 }, { "epoch": 2.130390535555273, "ewc_loss": 0.07177734375, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6206982135772705, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8638546466827393, "num_tokens": 635747710.0, "step": 16747 }, { "epoch": 2.1305177458338633, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 3.015291452407837, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8696648478507996, "num_tokens": 635781355.0, "step": 16748 }, { "epoch": 2.130644956112454, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5700244903564453, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8829377889633179, "num_tokens": 635821340.0, "step": 16749 }, { "epoch": 2.1307721663910444, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.893079996109009, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8708672523498535, "num_tokens": 635857943.0, "step": 16750 }, { "epoch": 2.130899376669635, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003108978271484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5719683170318604, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8749906420707703, "num_tokens": 635896808.0, "step": 16751 }, { "epoch": 2.1310265869482254, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.982948303222656e-05, "grad_norm": 2.9146082401275635, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.8868974447250366, "num_tokens": 635934952.0, "step": 16752 }, { "epoch": 2.131153797226816, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.602447032928467, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.882554829120636, "num_tokens": 635972968.0, "step": 16753 }, { "epoch": 2.1312810075054065, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.9591064453125e-05, "grad_norm": 2.913504123687744, "learning_rate": 1e-05, "loss": 0.3678, "mean_token_accuracy": 0.8935068249702454, "num_tokens": 636004761.0, "step": 16754 }, { "epoch": 2.131408217783997, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6242332458496094, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8758671283721924, "num_tokens": 636043003.0, "step": 16755 }, { "epoch": 2.1315354280625876, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.9591064453125e-05, "grad_norm": 2.88736629486084, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8807353973388672, "num_tokens": 636083580.0, "step": 16756 }, { "epoch": 2.131662638341178, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.6363842487335205, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8851120471954346, "num_tokens": 636124248.0, "step": 16757 }, { "epoch": 2.1317898486197686, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.866968870162964, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8732502460479736, "num_tokens": 636164665.0, "step": 16758 }, { "epoch": 2.131917058898359, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6486477851867676, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8784154057502747, "num_tokens": 636201526.0, "step": 16759 }, { "epoch": 2.1320442691769497, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8285696506500244, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8798151016235352, "num_tokens": 636247793.0, "step": 16760 }, { "epoch": 2.13217147945554, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.632132053375244, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8894429802894592, "num_tokens": 636289989.0, "step": 16761 }, { "epoch": 2.1322986897341307, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8258893489837646, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8678625226020813, "num_tokens": 636328729.0, "step": 16762 }, { "epoch": 2.132425900012721, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.656423807144165, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8795921802520752, "num_tokens": 636362593.0, "step": 16763 }, { "epoch": 2.1325531102913113, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.783127784729004, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8899765014648438, "num_tokens": 636402790.0, "step": 16764 }, { "epoch": 2.132680320569902, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.638871908187866, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8795789480209351, "num_tokens": 636443052.0, "step": 16765 }, { "epoch": 2.1328075308484924, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.783376693725586, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8823237419128418, "num_tokens": 636481345.0, "step": 16766 }, { "epoch": 2.132934741127083, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6332693099975586, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8814665079116821, "num_tokens": 636516638.0, "step": 16767 }, { "epoch": 2.1330619514056735, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7623419761657715, "learning_rate": 1e-05, "loss": 0.3652, "mean_token_accuracy": 0.8943663835525513, "num_tokens": 636551606.0, "step": 16768 }, { "epoch": 2.133189161684264, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.637169599533081, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8753706216812134, "num_tokens": 636589298.0, "step": 16769 }, { "epoch": 2.1333163719628545, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.76513671875, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8805790543556213, "num_tokens": 636622010.0, "step": 16770 }, { "epoch": 2.133443582241445, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.617070198059082, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8839702010154724, "num_tokens": 636665355.0, "step": 16771 }, { "epoch": 2.1335707925200356, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.751840353012085, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8883016109466553, "num_tokens": 636706135.0, "step": 16772 }, { "epoch": 2.133698002798626, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6135177612304688, "learning_rate": 1e-05, "loss": 0.368, "mean_token_accuracy": 0.8946375250816345, "num_tokens": 636744587.0, "step": 16773 }, { "epoch": 2.1338252130772166, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.747025728225708, "learning_rate": 1e-05, "loss": 0.3662, "mean_token_accuracy": 0.8928584456443787, "num_tokens": 636784175.0, "step": 16774 }, { "epoch": 2.133952423355807, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.628964900970459, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8712886571884155, "num_tokens": 636821752.0, "step": 16775 }, { "epoch": 2.1340796336343977, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7734975814819336, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8667020797729492, "num_tokens": 636855507.0, "step": 16776 }, { "epoch": 2.134206843912988, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6143040657043457, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8755693435668945, "num_tokens": 636895401.0, "step": 16777 }, { "epoch": 2.1343340541915787, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.751077890396118, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8633618354797363, "num_tokens": 636940225.0, "step": 16778 }, { "epoch": 2.1344612644701693, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.619925022125244, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8758893013000488, "num_tokens": 636976685.0, "step": 16779 }, { "epoch": 2.13458847474876, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7510733604431152, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8815118074417114, "num_tokens": 637018170.0, "step": 16780 }, { "epoch": 2.1347156850273503, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6129162311553955, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8836819529533386, "num_tokens": 637055115.0, "step": 16781 }, { "epoch": 2.134842895305941, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.751438617706299, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8686952590942383, "num_tokens": 637093964.0, "step": 16782 }, { "epoch": 2.1349701055845314, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6003642082214355, "learning_rate": 1e-05, "loss": 0.3539, "mean_token_accuracy": 0.901035487651825, "num_tokens": 637135242.0, "step": 16783 }, { "epoch": 2.135097315863122, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.754288911819458, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8807468414306641, "num_tokens": 637173407.0, "step": 16784 }, { "epoch": 2.1352245261417124, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6206886768341064, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8790197372436523, "num_tokens": 637207613.0, "step": 16785 }, { "epoch": 2.135351736420303, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.74817156791687, "learning_rate": 1e-05, "loss": 0.3641, "mean_token_accuracy": 0.8945978879928589, "num_tokens": 637242010.0, "step": 16786 }, { "epoch": 2.1354789466988935, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.59977126121521, "learning_rate": 1e-05, "loss": 0.3868, "mean_token_accuracy": 0.8860712051391602, "num_tokens": 637284034.0, "step": 16787 }, { "epoch": 2.1356061569774836, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.738734483718872, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.8886587619781494, "num_tokens": 637325657.0, "step": 16788 }, { "epoch": 2.135733367256074, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6120996475219727, "learning_rate": 1e-05, "loss": 0.3773, "mean_token_accuracy": 0.8906201720237732, "num_tokens": 637360946.0, "step": 16789 }, { "epoch": 2.1358605775346646, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.759056806564331, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8783015012741089, "num_tokens": 637397540.0, "step": 16790 }, { "epoch": 2.135987787813255, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5925309658050537, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8872759938240051, "num_tokens": 637443148.0, "step": 16791 }, { "epoch": 2.1361149980918457, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7677457332611084, "learning_rate": 1e-05, "loss": 0.3834, "mean_token_accuracy": 0.8906146287918091, "num_tokens": 637480796.0, "step": 16792 }, { "epoch": 2.136242208370436, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6262245178222656, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8768879771232605, "num_tokens": 637525465.0, "step": 16793 }, { "epoch": 2.1363694186490267, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.752943277359009, "learning_rate": 1e-05, "loss": 0.3983, "mean_token_accuracy": 0.8833270072937012, "num_tokens": 637563235.0, "step": 16794 }, { "epoch": 2.1364966289276173, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6047942638397217, "learning_rate": 1e-05, "loss": 0.3894, "mean_token_accuracy": 0.889259934425354, "num_tokens": 637600061.0, "step": 16795 }, { "epoch": 2.136623839206208, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7454497814178467, "learning_rate": 1e-05, "loss": 0.3686, "mean_token_accuracy": 0.8955127000808716, "num_tokens": 637640877.0, "step": 16796 }, { "epoch": 2.1367510494847983, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.00031280517578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.611898899078369, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8777918815612793, "num_tokens": 637679248.0, "step": 16797 }, { "epoch": 2.136878259763389, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.75650691986084, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8764413595199585, "num_tokens": 637715973.0, "step": 16798 }, { "epoch": 2.1370054700419794, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6059324741363525, "learning_rate": 1e-05, "loss": 0.3961, "mean_token_accuracy": 0.8871070742607117, "num_tokens": 637753358.0, "step": 16799 }, { "epoch": 2.13713268032057, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7433712482452393, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.882118821144104, "num_tokens": 637798608.0, "step": 16800 }, { "epoch": 2.1372598905991604, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6101646423339844, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8835171461105347, "num_tokens": 637837173.0, "step": 16801 }, { "epoch": 2.137387100877751, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.749657392501831, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8897653222084045, "num_tokens": 637874072.0, "step": 16802 }, { "epoch": 2.1375143111563415, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6223719120025635, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.872844934463501, "num_tokens": 637909369.0, "step": 16803 }, { "epoch": 2.137641521434932, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.744595766067505, "learning_rate": 1e-05, "loss": 0.3827, "mean_token_accuracy": 0.8925046920776367, "num_tokens": 637946526.0, "step": 16804 }, { "epoch": 2.1377687317135226, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6189773082733154, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8767832517623901, "num_tokens": 637984476.0, "step": 16805 }, { "epoch": 2.137895941992113, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.744760036468506, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8772943019866943, "num_tokens": 638019098.0, "step": 16806 }, { "epoch": 2.1380231522707036, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6172940731048584, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8824779987335205, "num_tokens": 638059049.0, "step": 16807 }, { "epoch": 2.138150362549294, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.74204421043396, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8828297853469849, "num_tokens": 638095887.0, "step": 16808 }, { "epoch": 2.1382775728278847, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6162822246551514, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.8636173009872437, "num_tokens": 638141855.0, "step": 16809 }, { "epoch": 2.138404783106475, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.739607334136963, "learning_rate": 1e-05, "loss": 0.379, "mean_token_accuracy": 0.8924126625061035, "num_tokens": 638184491.0, "step": 16810 }, { "epoch": 2.1385319933850653, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.625857353210449, "learning_rate": 1e-05, "loss": 0.3692, "mean_token_accuracy": 0.8911454677581787, "num_tokens": 638218341.0, "step": 16811 }, { "epoch": 2.1386592036636562, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.744527578353882, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8756941556930542, "num_tokens": 638257422.0, "step": 16812 }, { "epoch": 2.1387864139422463, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6152215003967285, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8854358196258545, "num_tokens": 638298053.0, "step": 16813 }, { "epoch": 2.138913624220837, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7511990070343018, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.864879846572876, "num_tokens": 638337870.0, "step": 16814 }, { "epoch": 2.1390408344994274, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.619511365890503, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8811647891998291, "num_tokens": 638371587.0, "step": 16815 }, { "epoch": 2.139168044778018, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.739546060562134, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8810044527053833, "num_tokens": 638416879.0, "step": 16816 }, { "epoch": 2.1392952550566084, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6087865829467773, "learning_rate": 1e-05, "loss": 0.3739, "mean_token_accuracy": 0.8910118937492371, "num_tokens": 638452809.0, "step": 16817 }, { "epoch": 2.139422465335199, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.754147529602051, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8755393028259277, "num_tokens": 638487027.0, "step": 16818 }, { "epoch": 2.1395496756137895, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.617673635482788, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8848305940628052, "num_tokens": 638517922.0, "step": 16819 }, { "epoch": 2.13967688589238, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.739908456802368, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8716444969177246, "num_tokens": 638562360.0, "step": 16820 }, { "epoch": 2.1398040961709706, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6228294372558594, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8772746324539185, "num_tokens": 638595649.0, "step": 16821 }, { "epoch": 2.139931306449561, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7519843578338623, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8833708167076111, "num_tokens": 638632588.0, "step": 16822 }, { "epoch": 2.1400585167281516, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.629958152770996, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.883965015411377, "num_tokens": 638663000.0, "step": 16823 }, { "epoch": 2.140185727006742, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7229626178741455, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8791844248771667, "num_tokens": 638703737.0, "step": 16824 }, { "epoch": 2.1403129372853327, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6215364933013916, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8690831661224365, "num_tokens": 638745526.0, "step": 16825 }, { "epoch": 2.140440147563923, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.731780767440796, "learning_rate": 1e-05, "loss": 0.3946, "mean_token_accuracy": 0.886544942855835, "num_tokens": 638780865.0, "step": 16826 }, { "epoch": 2.1405673578425137, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.62292218208313, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8788449764251709, "num_tokens": 638822274.0, "step": 16827 }, { "epoch": 2.1406945681211043, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7399775981903076, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8708353042602539, "num_tokens": 638859077.0, "step": 16828 }, { "epoch": 2.140821778399695, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.611409902572632, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8824132680892944, "num_tokens": 638898055.0, "step": 16829 }, { "epoch": 2.1409489886782853, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.742429733276367, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8693215847015381, "num_tokens": 638939597.0, "step": 16830 }, { "epoch": 2.141076198956876, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.615406036376953, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.876362681388855, "num_tokens": 638978268.0, "step": 16831 }, { "epoch": 2.1412034092354664, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.746047019958496, "learning_rate": 1e-05, "loss": 0.3715, "mean_token_accuracy": 0.8944388628005981, "num_tokens": 639010554.0, "step": 16832 }, { "epoch": 2.141330619514057, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6010239124298096, "learning_rate": 1e-05, "loss": 0.3426, "mean_token_accuracy": 0.9044644236564636, "num_tokens": 639046721.0, "step": 16833 }, { "epoch": 2.1414578297926474, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.775128126144409, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8756610155105591, "num_tokens": 639080078.0, "step": 16834 }, { "epoch": 2.141585040071238, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.604116916656494, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8824114799499512, "num_tokens": 639119469.0, "step": 16835 }, { "epoch": 2.141712250349828, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7596938610076904, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8786941170692444, "num_tokens": 639159954.0, "step": 16836 }, { "epoch": 2.141839460628419, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.597169876098633, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8853097558021545, "num_tokens": 639202455.0, "step": 16837 }, { "epoch": 2.141966670907009, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.761220932006836, "learning_rate": 1e-05, "loss": 0.3783, "mean_token_accuracy": 0.890534520149231, "num_tokens": 639240889.0, "step": 16838 }, { "epoch": 2.1420938811855996, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.610400676727295, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.8824292421340942, "num_tokens": 639275079.0, "step": 16839 }, { "epoch": 2.14222109146419, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7890384197235107, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8734530806541443, "num_tokens": 639308166.0, "step": 16840 }, { "epoch": 2.1423483017427807, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6019835472106934, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8912162780761719, "num_tokens": 639341862.0, "step": 16841 }, { "epoch": 2.142475512021371, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.781712532043457, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8909191489219666, "num_tokens": 639381065.0, "step": 16842 }, { "epoch": 2.1426027222999617, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5896193981170654, "learning_rate": 1e-05, "loss": 0.3832, "mean_token_accuracy": 0.889552116394043, "num_tokens": 639421232.0, "step": 16843 }, { "epoch": 2.1427299325785523, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7927873134613037, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8769674897193909, "num_tokens": 639459462.0, "step": 16844 }, { "epoch": 2.142857142857143, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.60006046295166, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.881702184677124, "num_tokens": 639499759.0, "step": 16845 }, { "epoch": 2.1429843531357333, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8078906536102295, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8841686248779297, "num_tokens": 639534285.0, "step": 16846 }, { "epoch": 2.143111563414324, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.592505931854248, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8774026036262512, "num_tokens": 639576058.0, "step": 16847 }, { "epoch": 2.1432387736929144, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8248255252838135, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8838881850242615, "num_tokens": 639610464.0, "step": 16848 }, { "epoch": 2.143365983971505, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.59016489982605, "learning_rate": 1e-05, "loss": 0.367, "mean_token_accuracy": 0.8943800330162048, "num_tokens": 639645327.0, "step": 16849 }, { "epoch": 2.1434931942500954, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8258461952209473, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8714590072631836, "num_tokens": 639684501.0, "step": 16850 }, { "epoch": 2.143620404528686, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.598825216293335, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8746799230575562, "num_tokens": 639721687.0, "step": 16851 }, { "epoch": 2.1437476148072765, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.820949077606201, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8630330562591553, "num_tokens": 639761503.0, "step": 16852 }, { "epoch": 2.143874825085867, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.596830129623413, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.885011613368988, "num_tokens": 639800938.0, "step": 16853 }, { "epoch": 2.1440020353644575, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8137261867523193, "learning_rate": 1e-05, "loss": 0.3655, "mean_token_accuracy": 0.8957935571670532, "num_tokens": 639835318.0, "step": 16854 }, { "epoch": 2.144129245643048, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.598144292831421, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8859330415725708, "num_tokens": 639878155.0, "step": 16855 }, { "epoch": 2.1442564559216386, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.809361219406128, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8856188654899597, "num_tokens": 639920369.0, "step": 16856 }, { "epoch": 2.144383666200229, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.588690996170044, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8835310339927673, "num_tokens": 639964690.0, "step": 16857 }, { "epoch": 2.1445108764788197, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.859205722808838, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8693200349807739, "num_tokens": 640002969.0, "step": 16858 }, { "epoch": 2.14463808675741, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.588160514831543, "learning_rate": 1e-05, "loss": 0.3738, "mean_token_accuracy": 0.8927195072174072, "num_tokens": 640044061.0, "step": 16859 }, { "epoch": 2.1447652970360007, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8433094024658203, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8863170146942139, "num_tokens": 640082803.0, "step": 16860 }, { "epoch": 2.144892507314591, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6112918853759766, "learning_rate": 1e-05, "loss": 0.3642, "mean_token_accuracy": 0.8991076350212097, "num_tokens": 640123223.0, "step": 16861 }, { "epoch": 2.1450197175931813, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.828953742980957, "learning_rate": 1e-05, "loss": 0.3649, "mean_token_accuracy": 0.8973690867424011, "num_tokens": 640160674.0, "step": 16862 }, { "epoch": 2.145146927871772, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6174824237823486, "learning_rate": 1e-05, "loss": 0.3942, "mean_token_accuracy": 0.8886653184890747, "num_tokens": 640197347.0, "step": 16863 }, { "epoch": 2.1452741381503624, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8359663486480713, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8819194436073303, "num_tokens": 640235599.0, "step": 16864 }, { "epoch": 2.145401348428953, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6110568046569824, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.878240704536438, "num_tokens": 640277753.0, "step": 16865 }, { "epoch": 2.1455285587075434, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8203670978546143, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8864492177963257, "num_tokens": 640317486.0, "step": 16866 }, { "epoch": 2.145655768986134, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6220669746398926, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8668873310089111, "num_tokens": 640357240.0, "step": 16867 }, { "epoch": 2.1457829792647245, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8050968647003174, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8856382369995117, "num_tokens": 640392429.0, "step": 16868 }, { "epoch": 2.145910189543315, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6316370964050293, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.871018648147583, "num_tokens": 640432529.0, "step": 16869 }, { "epoch": 2.1460373998219056, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.798093318939209, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8841163516044617, "num_tokens": 640468443.0, "step": 16870 }, { "epoch": 2.146164610100496, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.628850221633911, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8720027208328247, "num_tokens": 640505647.0, "step": 16871 }, { "epoch": 2.1462918203790866, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.773995876312256, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8698010444641113, "num_tokens": 640551180.0, "step": 16872 }, { "epoch": 2.146419030657677, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.621729612350464, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8873546123504639, "num_tokens": 640590939.0, "step": 16873 }, { "epoch": 2.1465462409362677, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7855682373046875, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.867828369140625, "num_tokens": 640625688.0, "step": 16874 }, { "epoch": 2.146673451214858, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6332526206970215, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.8604552745819092, "num_tokens": 640669142.0, "step": 16875 }, { "epoch": 2.1468006614934487, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.761369466781616, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8739010691642761, "num_tokens": 640713345.0, "step": 16876 }, { "epoch": 2.1469278717720393, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.643507719039917, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8587038516998291, "num_tokens": 640750930.0, "step": 16877 }, { "epoch": 2.14705508205063, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7332873344421387, "learning_rate": 1e-05, "loss": 0.3926, "mean_token_accuracy": 0.8872973918914795, "num_tokens": 640792029.0, "step": 16878 }, { "epoch": 2.1471822923292203, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6431360244750977, "learning_rate": 1e-05, "loss": 0.3987, "mean_token_accuracy": 0.8852440714836121, "num_tokens": 640833952.0, "step": 16879 }, { "epoch": 2.147309502607811, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.724126100540161, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8804342746734619, "num_tokens": 640867980.0, "step": 16880 }, { "epoch": 2.1474367128864014, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6304564476013184, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8673601150512695, "num_tokens": 640913241.0, "step": 16881 }, { "epoch": 2.147563923164992, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7220258712768555, "learning_rate": 1e-05, "loss": 0.3671, "mean_token_accuracy": 0.8977831602096558, "num_tokens": 640946705.0, "step": 16882 }, { "epoch": 2.1476911334435824, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.7202162742614746, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8888667821884155, "num_tokens": 640985658.0, "step": 16883 }, { "epoch": 2.147818343722173, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.670233964920044, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8735869526863098, "num_tokens": 641025963.0, "step": 16884 }, { "epoch": 2.1479455540007635, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.655566453933716, "learning_rate": 1e-05, "loss": 0.3918, "mean_token_accuracy": 0.888377845287323, "num_tokens": 641070853.0, "step": 16885 }, { "epoch": 2.1480727642793536, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6620707511901855, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.8851251602172852, "num_tokens": 641106336.0, "step": 16886 }, { "epoch": 2.148199974557944, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6557674407958984, "learning_rate": 1e-05, "loss": 0.3603, "mean_token_accuracy": 0.8982741236686707, "num_tokens": 641147842.0, "step": 16887 }, { "epoch": 2.1483271848365346, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.649348735809326, "learning_rate": 1e-05, "loss": 0.3741, "mean_token_accuracy": 0.8915450572967529, "num_tokens": 641186329.0, "step": 16888 }, { "epoch": 2.148454395115125, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6713454723358154, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8749372959136963, "num_tokens": 641227892.0, "step": 16889 }, { "epoch": 2.1485816053937157, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6557838916778564, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8869878649711609, "num_tokens": 641265611.0, "step": 16890 }, { "epoch": 2.148708815672306, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.690239191055298, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8638570308685303, "num_tokens": 641299564.0, "step": 16891 }, { "epoch": 2.1488360259508967, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6417179107666016, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.884060800075531, "num_tokens": 641335892.0, "step": 16892 }, { "epoch": 2.1489632362294873, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6980297565460205, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.8801549673080444, "num_tokens": 641375470.0, "step": 16893 }, { "epoch": 2.149090446508078, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6476399898529053, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8820043802261353, "num_tokens": 641408793.0, "step": 16894 }, { "epoch": 2.1492176567866683, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6995885372161865, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8787984848022461, "num_tokens": 641444949.0, "step": 16895 }, { "epoch": 2.149344867065259, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6345181465148926, "learning_rate": 1e-05, "loss": 0.4058, "mean_token_accuracy": 0.8816089630126953, "num_tokens": 641482775.0, "step": 16896 }, { "epoch": 2.1494720773438494, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.698624610900879, "learning_rate": 1e-05, "loss": 0.3906, "mean_token_accuracy": 0.8865827918052673, "num_tokens": 641520385.0, "step": 16897 }, { "epoch": 2.14959928762244, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.632732391357422, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8704244494438171, "num_tokens": 641563885.0, "step": 16898 }, { "epoch": 2.1497264979010304, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.705749750137329, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8813868761062622, "num_tokens": 641600868.0, "step": 16899 }, { "epoch": 2.149853708179621, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.631925344467163, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8792569637298584, "num_tokens": 641643936.0, "step": 16900 }, { "epoch": 2.1499809184582115, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7125027179718018, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8745251297950745, "num_tokens": 641680248.0, "step": 16901 }, { "epoch": 2.150108128736802, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6196722984313965, "learning_rate": 1e-05, "loss": 0.3898, "mean_token_accuracy": 0.8887113332748413, "num_tokens": 641721280.0, "step": 16902 }, { "epoch": 2.1502353390153925, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7121357917785645, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8782248497009277, "num_tokens": 641760168.0, "step": 16903 }, { "epoch": 2.150362549293983, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6354761123657227, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8683633804321289, "num_tokens": 641798517.0, "step": 16904 }, { "epoch": 2.1504897595725736, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.695943832397461, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8756369948387146, "num_tokens": 641841486.0, "step": 16905 }, { "epoch": 2.150616969851164, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6362123489379883, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8735218048095703, "num_tokens": 641880042.0, "step": 16906 }, { "epoch": 2.1507441801297547, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.709786891937256, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8845138549804688, "num_tokens": 641916069.0, "step": 16907 }, { "epoch": 2.150871390408345, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.632120370864868, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8788062334060669, "num_tokens": 641951298.0, "step": 16908 }, { "epoch": 2.1509986006869353, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6977388858795166, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8846255540847778, "num_tokens": 641987577.0, "step": 16909 }, { "epoch": 2.1511258109655262, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.625196695327759, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8651570081710815, "num_tokens": 642031914.0, "step": 16910 }, { "epoch": 2.1512530212441163, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7300736904144287, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8772253394126892, "num_tokens": 642069729.0, "step": 16911 }, { "epoch": 2.151380231522707, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.615030527114868, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8793737888336182, "num_tokens": 642106831.0, "step": 16912 }, { "epoch": 2.1515074418012974, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.722705841064453, "learning_rate": 1e-05, "loss": 0.3845, "mean_token_accuracy": 0.8898217082023621, "num_tokens": 642142408.0, "step": 16913 }, { "epoch": 2.151634652079888, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6126837730407715, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8764735460281372, "num_tokens": 642181860.0, "step": 16914 }, { "epoch": 2.1517618623584784, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7259902954101562, "learning_rate": 1e-05, "loss": 0.375, "mean_token_accuracy": 0.8958282470703125, "num_tokens": 642219073.0, "step": 16915 }, { "epoch": 2.151889072637069, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.621722936630249, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.879440426826477, "num_tokens": 642254872.0, "step": 16916 }, { "epoch": 2.1520162829156595, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7144899368286133, "learning_rate": 1e-05, "loss": 0.3806, "mean_token_accuracy": 0.8917531967163086, "num_tokens": 642295379.0, "step": 16917 }, { "epoch": 2.15214349319425, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6275434494018555, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8641951084136963, "num_tokens": 642333316.0, "step": 16918 }, { "epoch": 2.1522707034728406, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7281148433685303, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8734875917434692, "num_tokens": 642373677.0, "step": 16919 }, { "epoch": 2.152397913751431, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.621000289916992, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.871475100517273, "num_tokens": 642416066.0, "step": 16920 }, { "epoch": 2.1525251240300216, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7272887229919434, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8676308989524841, "num_tokens": 642456818.0, "step": 16921 }, { "epoch": 2.152652334308612, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6072285175323486, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.878396213054657, "num_tokens": 642498021.0, "step": 16922 }, { "epoch": 2.1527795445872027, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7536301612854004, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8726851344108582, "num_tokens": 642533999.0, "step": 16923 }, { "epoch": 2.152906754865793, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.621629238128662, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8822582960128784, "num_tokens": 642571943.0, "step": 16924 }, { "epoch": 2.1530339651443837, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.747490644454956, "learning_rate": 1e-05, "loss": 0.3877, "mean_token_accuracy": 0.8868454098701477, "num_tokens": 642605145.0, "step": 16925 }, { "epoch": 2.1531611754229742, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6074821949005127, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8769307136535645, "num_tokens": 642644686.0, "step": 16926 }, { "epoch": 2.1532883857015648, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.761288642883301, "learning_rate": 1e-05, "loss": 0.3857, "mean_token_accuracy": 0.8878800272941589, "num_tokens": 642677815.0, "step": 16927 }, { "epoch": 2.1534155959801553, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.59047532081604, "learning_rate": 1e-05, "loss": 0.3746, "mean_token_accuracy": 0.8952867984771729, "num_tokens": 642716245.0, "step": 16928 }, { "epoch": 2.153542806258746, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.780179738998413, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8793070912361145, "num_tokens": 642750428.0, "step": 16929 }, { "epoch": 2.1536700165373364, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5993824005126953, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8731182217597961, "num_tokens": 642790349.0, "step": 16930 }, { "epoch": 2.153797226815927, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7670071125030518, "learning_rate": 1e-05, "loss": 0.3983, "mean_token_accuracy": 0.8850536942481995, "num_tokens": 642835063.0, "step": 16931 }, { "epoch": 2.1539244370945174, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.597191572189331, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8778855204582214, "num_tokens": 642875013.0, "step": 16932 }, { "epoch": 2.154051647373108, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.783198833465576, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8722184896469116, "num_tokens": 642914934.0, "step": 16933 }, { "epoch": 2.154178857651698, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6107394695281982, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.8818061351776123, "num_tokens": 642949368.0, "step": 16934 }, { "epoch": 2.1543060679302886, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7832791805267334, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8723334670066833, "num_tokens": 642988083.0, "step": 16935 }, { "epoch": 2.154433278208879, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6042282581329346, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.8636366128921509, "num_tokens": 643028868.0, "step": 16936 }, { "epoch": 2.1545604884874696, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7704861164093018, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8766579627990723, "num_tokens": 643072251.0, "step": 16937 }, { "epoch": 2.15468769876606, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6047818660736084, "learning_rate": 1e-05, "loss": 0.4013, "mean_token_accuracy": 0.8834308385848999, "num_tokens": 643109864.0, "step": 16938 }, { "epoch": 2.1548149090446507, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.771907329559326, "learning_rate": 1e-05, "loss": 0.3825, "mean_token_accuracy": 0.8927003145217896, "num_tokens": 643147868.0, "step": 16939 }, { "epoch": 2.154942119323241, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5990688800811768, "learning_rate": 1e-05, "loss": 0.3364, "mean_token_accuracy": 0.9035594463348389, "num_tokens": 643182173.0, "step": 16940 }, { "epoch": 2.1550693296018317, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7921173572540283, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8756222128868103, "num_tokens": 643217006.0, "step": 16941 }, { "epoch": 2.1551965398804223, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.614561080932617, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8647785186767578, "num_tokens": 643253724.0, "step": 16942 }, { "epoch": 2.155323750159013, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7910375595092773, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8682499527931213, "num_tokens": 643294633.0, "step": 16943 }, { "epoch": 2.1554509604376033, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6033172607421875, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8724779486656189, "num_tokens": 643333269.0, "step": 16944 }, { "epoch": 2.155578170716194, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7790112495422363, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.8888462781906128, "num_tokens": 643374892.0, "step": 16945 }, { "epoch": 2.1557053809947844, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5991480350494385, "learning_rate": 1e-05, "loss": 0.3788, "mean_token_accuracy": 0.8918030858039856, "num_tokens": 643411774.0, "step": 16946 }, { "epoch": 2.155832591273375, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7837936878204346, "learning_rate": 1e-05, "loss": 0.3908, "mean_token_accuracy": 0.8883821964263916, "num_tokens": 643450034.0, "step": 16947 }, { "epoch": 2.1559598015519654, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.630655288696289, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8810454607009888, "num_tokens": 643482901.0, "step": 16948 }, { "epoch": 2.156087011830556, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.768000364303589, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8769130110740662, "num_tokens": 643524918.0, "step": 16949 }, { "epoch": 2.1562142221091465, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6201276779174805, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8697806000709534, "num_tokens": 643563931.0, "step": 16950 }, { "epoch": 2.156341432387737, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7973098754882812, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8807723522186279, "num_tokens": 643602838.0, "step": 16951 }, { "epoch": 2.1564686426663275, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6100871562957764, "learning_rate": 1e-05, "loss": 0.4098, "mean_token_accuracy": 0.8815450668334961, "num_tokens": 643641128.0, "step": 16952 }, { "epoch": 2.156595852944918, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.780668020248413, "learning_rate": 1e-05, "loss": 0.4021, "mean_token_accuracy": 0.8838520050048828, "num_tokens": 643674853.0, "step": 16953 }, { "epoch": 2.1567230632235086, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.618335008621216, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8764636516571045, "num_tokens": 643709924.0, "step": 16954 }, { "epoch": 2.156850273502099, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7743465900421143, "learning_rate": 1e-05, "loss": 0.3921, "mean_token_accuracy": 0.8852552175521851, "num_tokens": 643744899.0, "step": 16955 }, { "epoch": 2.1569774837806897, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6084394454956055, "learning_rate": 1e-05, "loss": 0.3796, "mean_token_accuracy": 0.8931605219841003, "num_tokens": 643781316.0, "step": 16956 }, { "epoch": 2.15710469405928, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7596967220306396, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.883905827999115, "num_tokens": 643824267.0, "step": 16957 }, { "epoch": 2.1572319043378707, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.603900671005249, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8784565925598145, "num_tokens": 643867155.0, "step": 16958 }, { "epoch": 2.157359114616461, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7724552154541016, "learning_rate": 1e-05, "loss": 0.3965, "mean_token_accuracy": 0.8888015747070312, "num_tokens": 643907813.0, "step": 16959 }, { "epoch": 2.1574863248950513, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.624053955078125, "learning_rate": 1e-05, "loss": 0.3902, "mean_token_accuracy": 0.8865879774093628, "num_tokens": 643940387.0, "step": 16960 }, { "epoch": 2.157613535173642, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.758594274520874, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8732970356941223, "num_tokens": 643980387.0, "step": 16961 }, { "epoch": 2.1577407454522324, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6147985458374023, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8822829723358154, "num_tokens": 644020883.0, "step": 16962 }, { "epoch": 2.157867955730823, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7852299213409424, "learning_rate": 1e-05, "loss": 0.4639, "mean_token_accuracy": 0.8643830418586731, "num_tokens": 644054492.0, "step": 16963 }, { "epoch": 2.1579951660094134, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6129701137542725, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.869966983795166, "num_tokens": 644095009.0, "step": 16964 }, { "epoch": 2.158122376288004, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7681851387023926, "learning_rate": 1e-05, "loss": 0.378, "mean_token_accuracy": 0.8911169171333313, "num_tokens": 644128469.0, "step": 16965 }, { "epoch": 2.1582495865665945, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6229045391082764, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8639364242553711, "num_tokens": 644167102.0, "step": 16966 }, { "epoch": 2.158376796845185, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.761143445968628, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.878075361251831, "num_tokens": 644204319.0, "step": 16967 }, { "epoch": 2.1585040071237755, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6167712211608887, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8726235628128052, "num_tokens": 644246950.0, "step": 16968 }, { "epoch": 2.158631217402366, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7536733150482178, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.882636308670044, "num_tokens": 644288744.0, "step": 16969 }, { "epoch": 2.1587584276809566, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6123416423797607, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8811619281768799, "num_tokens": 644326719.0, "step": 16970 }, { "epoch": 2.158885637959547, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7663352489471436, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.876520037651062, "num_tokens": 644362027.0, "step": 16971 }, { "epoch": 2.1590128482381377, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6160085201263428, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8790790438652039, "num_tokens": 644401917.0, "step": 16972 }, { "epoch": 2.159140058516728, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7581570148468018, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8750011920928955, "num_tokens": 644441735.0, "step": 16973 }, { "epoch": 2.1592672687953187, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 4.469873428344727, "learning_rate": 1e-05, "loss": 0.3632, "mean_token_accuracy": 0.8937385082244873, "num_tokens": 644475833.0, "step": 16974 }, { "epoch": 2.1593944790739092, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 5.6743621826171875e-05, "grad_norm": 2.9393582344055176, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8812946081161499, "num_tokens": 644516506.0, "step": 16975 }, { "epoch": 2.1595216893524998, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.798837900161743, "learning_rate": 1e-05, "loss": 0.389, "mean_token_accuracy": 0.8864009976387024, "num_tokens": 644554081.0, "step": 16976 }, { "epoch": 2.1596488996310903, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6688618659973145, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8687710762023926, "num_tokens": 644599514.0, "step": 16977 }, { "epoch": 2.159776109909681, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 5.1021575927734375e-05, "grad_norm": 2.8992300033569336, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8718581199645996, "num_tokens": 644639569.0, "step": 16978 }, { "epoch": 2.1599033201882714, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.619288206100464, "learning_rate": 1e-05, "loss": 0.3774, "mean_token_accuracy": 0.8948843479156494, "num_tokens": 644685721.0, "step": 16979 }, { "epoch": 2.160030530466862, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 5.030632019042969e-05, "grad_norm": 2.9083452224731445, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8660832643508911, "num_tokens": 644729939.0, "step": 16980 }, { "epoch": 2.1601577407454524, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.635483741760254, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8734833002090454, "num_tokens": 644768181.0, "step": 16981 }, { "epoch": 2.160284951024043, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 2.9192755222320557, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8731369376182556, "num_tokens": 644806288.0, "step": 16982 }, { "epoch": 2.1604121613026335, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6147494316101074, "learning_rate": 1e-05, "loss": 0.3893, "mean_token_accuracy": 0.8874027729034424, "num_tokens": 644846116.0, "step": 16983 }, { "epoch": 2.1605393715812236, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 5.1021575927734375e-05, "grad_norm": 2.9248688220977783, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.87522292137146, "num_tokens": 644885330.0, "step": 16984 }, { "epoch": 2.160666581859814, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.6397345066070557, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8706957101821899, "num_tokens": 644923341.0, "step": 16985 }, { "epoch": 2.1607937921384046, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 2.91455340385437, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8730778694152832, "num_tokens": 644964165.0, "step": 16986 }, { "epoch": 2.160921002416995, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6318202018737793, "learning_rate": 1e-05, "loss": 0.3655, "mean_token_accuracy": 0.8937937021255493, "num_tokens": 644999319.0, "step": 16987 }, { "epoch": 2.1610482126955857, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 2.8957269191741943, "learning_rate": 1e-05, "loss": 0.3915, "mean_token_accuracy": 0.8892765045166016, "num_tokens": 645040687.0, "step": 16988 }, { "epoch": 2.161175422974176, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6484782695770264, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8764086961746216, "num_tokens": 645075772.0, "step": 16989 }, { "epoch": 2.1613026332527667, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.935264587402344e-05, "grad_norm": 2.8824853897094727, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8780301213264465, "num_tokens": 645111856.0, "step": 16990 }, { "epoch": 2.1614298435313573, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6389060020446777, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8744162321090698, "num_tokens": 645155995.0, "step": 16991 }, { "epoch": 2.161557053809948, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8626813888549805, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8646713495254517, "num_tokens": 645200101.0, "step": 16992 }, { "epoch": 2.1616842640885383, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6580638885498047, "learning_rate": 1e-05, "loss": 0.3861, "mean_token_accuracy": 0.8909320831298828, "num_tokens": 645235243.0, "step": 16993 }, { "epoch": 2.161811474367129, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8403990268707275, "learning_rate": 1e-05, "loss": 0.3893, "mean_token_accuracy": 0.891878604888916, "num_tokens": 645270369.0, "step": 16994 }, { "epoch": 2.1619386846457194, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6478610038757324, "learning_rate": 1e-05, "loss": 0.3734, "mean_token_accuracy": 0.8929212093353271, "num_tokens": 645304111.0, "step": 16995 }, { "epoch": 2.16206589492431, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8256523609161377, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8680067658424377, "num_tokens": 645339884.0, "step": 16996 }, { "epoch": 2.1621931052029004, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6383659839630127, "learning_rate": 1e-05, "loss": 0.3537, "mean_token_accuracy": 0.8967519998550415, "num_tokens": 645372465.0, "step": 16997 }, { "epoch": 2.162320315481491, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8063764572143555, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8843571543693542, "num_tokens": 645413022.0, "step": 16998 }, { "epoch": 2.1624475257600815, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.637615203857422, "learning_rate": 1e-05, "loss": 0.3753, "mean_token_accuracy": 0.8936331272125244, "num_tokens": 645451344.0, "step": 16999 }, { "epoch": 2.162574736038672, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.780886650085449, "learning_rate": 1e-05, "loss": 0.3562, "mean_token_accuracy": 0.8983438014984131, "num_tokens": 645493463.0, "step": 17000 }, { "epoch": 2.1627019463172625, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6269259452819824, "learning_rate": 1e-05, "loss": 0.3935, "mean_token_accuracy": 0.8884198665618896, "num_tokens": 645533508.0, "step": 17001 }, { "epoch": 2.162829156595853, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.786892890930176, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8851369619369507, "num_tokens": 645573866.0, "step": 17002 }, { "epoch": 2.1629563668744436, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.632986545562744, "learning_rate": 1e-05, "loss": 0.3875, "mean_token_accuracy": 0.8871556520462036, "num_tokens": 645609835.0, "step": 17003 }, { "epoch": 2.163083577153034, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003147125244140625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7899978160858154, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8855639696121216, "num_tokens": 645644228.0, "step": 17004 }, { "epoch": 2.1632107874316246, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.628041982650757, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8831202983856201, "num_tokens": 645682619.0, "step": 17005 }, { "epoch": 2.163337997710215, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7767300605773926, "learning_rate": 1e-05, "loss": 0.3971, "mean_token_accuracy": 0.8848939538002014, "num_tokens": 645719506.0, "step": 17006 }, { "epoch": 2.1634652079888053, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6168899536132812, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8754057288169861, "num_tokens": 645761930.0, "step": 17007 }, { "epoch": 2.1635924182673962, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7791407108306885, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8691351413726807, "num_tokens": 645800322.0, "step": 17008 }, { "epoch": 2.1637196285459863, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.612921714782715, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8795398473739624, "num_tokens": 645842914.0, "step": 17009 }, { "epoch": 2.163846838824577, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7692127227783203, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8736361265182495, "num_tokens": 645885798.0, "step": 17010 }, { "epoch": 2.1639740491031674, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.621500253677368, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8783479332923889, "num_tokens": 645922187.0, "step": 17011 }, { "epoch": 2.164101259381758, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7656960487365723, "learning_rate": 1e-05, "loss": 0.3722, "mean_token_accuracy": 0.893681526184082, "num_tokens": 645955594.0, "step": 17012 }, { "epoch": 2.1642284696603484, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.627080202102661, "learning_rate": 1e-05, "loss": 0.3908, "mean_token_accuracy": 0.8902687430381775, "num_tokens": 645992579.0, "step": 17013 }, { "epoch": 2.164355679938939, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7623648643493652, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8778846263885498, "num_tokens": 646033243.0, "step": 17014 }, { "epoch": 2.1644828902175295, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.624274492263794, "learning_rate": 1e-05, "loss": 0.3829, "mean_token_accuracy": 0.8907177448272705, "num_tokens": 646069309.0, "step": 17015 }, { "epoch": 2.16461010049612, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7665562629699707, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8807588219642639, "num_tokens": 646100720.0, "step": 17016 }, { "epoch": 2.1647373107747105, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.605320692062378, "learning_rate": 1e-05, "loss": 0.3501, "mean_token_accuracy": 0.8997358083724976, "num_tokens": 646135917.0, "step": 17017 }, { "epoch": 2.164864521053301, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7608959674835205, "learning_rate": 1e-05, "loss": 0.3676, "mean_token_accuracy": 0.8950570821762085, "num_tokens": 646173227.0, "step": 17018 }, { "epoch": 2.1649917313318916, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6289167404174805, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8741816878318787, "num_tokens": 646210023.0, "step": 17019 }, { "epoch": 2.165118941610482, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7658677101135254, "learning_rate": 1e-05, "loss": 0.3994, "mean_token_accuracy": 0.8848085403442383, "num_tokens": 646241124.0, "step": 17020 }, { "epoch": 2.1652461518890727, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.639554977416992, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.864854097366333, "num_tokens": 646276892.0, "step": 17021 }, { "epoch": 2.165373362167663, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.743913412094116, "learning_rate": 1e-05, "loss": 0.3709, "mean_token_accuracy": 0.8938020467758179, "num_tokens": 646308959.0, "step": 17022 }, { "epoch": 2.1655005724462537, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000316619873046875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.630765676498413, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.8872237801551819, "num_tokens": 646342614.0, "step": 17023 }, { "epoch": 2.1656277827248442, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7374091148376465, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8797454237937927, "num_tokens": 646380395.0, "step": 17024 }, { "epoch": 2.1657549930034348, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6446211338043213, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8709682822227478, "num_tokens": 646417068.0, "step": 17025 }, { "epoch": 2.1658822032820253, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7263917922973633, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8796322345733643, "num_tokens": 646456655.0, "step": 17026 }, { "epoch": 2.166009413560616, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6261188983917236, "learning_rate": 1e-05, "loss": 0.352, "mean_token_accuracy": 0.9025460481643677, "num_tokens": 646495035.0, "step": 17027 }, { "epoch": 2.1661366238392064, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.723268508911133, "learning_rate": 1e-05, "loss": 0.366, "mean_token_accuracy": 0.8934173583984375, "num_tokens": 646529695.0, "step": 17028 }, { "epoch": 2.166263834117797, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.669285297393799, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8626788258552551, "num_tokens": 646566563.0, "step": 17029 }, { "epoch": 2.1663910443963874, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.706453323364258, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8780255317687988, "num_tokens": 646609515.0, "step": 17030 }, { "epoch": 2.166518254674978, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.650923490524292, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8879072666168213, "num_tokens": 646650024.0, "step": 17031 }, { "epoch": 2.166645464953568, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6879074573516846, "learning_rate": 1e-05, "loss": 0.3895, "mean_token_accuracy": 0.8896888494491577, "num_tokens": 646690547.0, "step": 17032 }, { "epoch": 2.1667726752321586, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.672532558441162, "learning_rate": 1e-05, "loss": 0.384, "mean_token_accuracy": 0.8886274099349976, "num_tokens": 646722060.0, "step": 17033 }, { "epoch": 2.166899885510749, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6858718395233154, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8785585165023804, "num_tokens": 646761972.0, "step": 17034 }, { "epoch": 2.1670270957893396, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6901259422302246, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8711501955986023, "num_tokens": 646794847.0, "step": 17035 }, { "epoch": 2.16715430606793, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6625680923461914, "learning_rate": 1e-05, "loss": 0.3858, "mean_token_accuracy": 0.8910296559333801, "num_tokens": 646833744.0, "step": 17036 }, { "epoch": 2.1672815163465207, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6649186611175537, "learning_rate": 1e-05, "loss": 0.3885, "mean_token_accuracy": 0.8865232467651367, "num_tokens": 646874257.0, "step": 17037 }, { "epoch": 2.167408726625111, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6571691036224365, "learning_rate": 1e-05, "loss": 0.3767, "mean_token_accuracy": 0.8932908773422241, "num_tokens": 646912439.0, "step": 17038 }, { "epoch": 2.1675359369037017, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.677623987197876, "learning_rate": 1e-05, "loss": 0.375, "mean_token_accuracy": 0.8901528120040894, "num_tokens": 646946311.0, "step": 17039 }, { "epoch": 2.1676631471822922, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.661210060119629, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8683796525001526, "num_tokens": 646982426.0, "step": 17040 }, { "epoch": 2.1677903574608828, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.67000675201416, "learning_rate": 1e-05, "loss": 0.3748, "mean_token_accuracy": 0.8955777883529663, "num_tokens": 647023244.0, "step": 17041 }, { "epoch": 2.1679175677394733, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6767001152038574, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8638927936553955, "num_tokens": 647062434.0, "step": 17042 }, { "epoch": 2.168044778018064, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.665142059326172, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8761122226715088, "num_tokens": 647098021.0, "step": 17043 }, { "epoch": 2.1681719882966544, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.672943115234375, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.878124475479126, "num_tokens": 647134709.0, "step": 17044 }, { "epoch": 2.168299198575245, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.666412115097046, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8680329322814941, "num_tokens": 647169267.0, "step": 17045 }, { "epoch": 2.1684264088538354, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6631991863250732, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8772746920585632, "num_tokens": 647207399.0, "step": 17046 }, { "epoch": 2.168553619132426, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6638166904449463, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8786794543266296, "num_tokens": 647244722.0, "step": 17047 }, { "epoch": 2.1686808294110165, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6664979457855225, "learning_rate": 1e-05, "loss": 0.3942, "mean_token_accuracy": 0.8872703313827515, "num_tokens": 647279051.0, "step": 17048 }, { "epoch": 2.168808039689607, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.660405158996582, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8781033754348755, "num_tokens": 647319201.0, "step": 17049 }, { "epoch": 2.1689352499681975, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.679138422012329, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8770849704742432, "num_tokens": 647358306.0, "step": 17050 }, { "epoch": 2.169062460246788, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.682950973510742, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8710020780563354, "num_tokens": 647393879.0, "step": 17051 }, { "epoch": 2.1691896705253786, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.665813684463501, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8736411333084106, "num_tokens": 647434848.0, "step": 17052 }, { "epoch": 2.169316880803969, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.681814670562744, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8878685235977173, "num_tokens": 647469711.0, "step": 17053 }, { "epoch": 2.1694440910825596, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6692347526550293, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8639646768569946, "num_tokens": 647507788.0, "step": 17054 }, { "epoch": 2.16957130136115, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.689403533935547, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.881788969039917, "num_tokens": 647542193.0, "step": 17055 }, { "epoch": 2.1696985116397407, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6462438106536865, "learning_rate": 1e-05, "loss": 0.3575, "mean_token_accuracy": 0.8984091877937317, "num_tokens": 647578692.0, "step": 17056 }, { "epoch": 2.169825721918331, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.684298515319824, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.8866769075393677, "num_tokens": 647618158.0, "step": 17057 }, { "epoch": 2.1699529321969213, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.640350580215454, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8844218850135803, "num_tokens": 647660788.0, "step": 17058 }, { "epoch": 2.170080142475512, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6928303241729736, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8713961839675903, "num_tokens": 647700418.0, "step": 17059 }, { "epoch": 2.1702073527541024, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.63879132270813, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.8886678218841553, "num_tokens": 647738248.0, "step": 17060 }, { "epoch": 2.170334563032693, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.692643404006958, "learning_rate": 1e-05, "loss": 0.3893, "mean_token_accuracy": 0.8885040879249573, "num_tokens": 647777287.0, "step": 17061 }, { "epoch": 2.1704617733112834, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.623957395553589, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8777292966842651, "num_tokens": 647823213.0, "step": 17062 }, { "epoch": 2.170588983589874, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7010021209716797, "learning_rate": 1e-05, "loss": 0.3788, "mean_token_accuracy": 0.8951536417007446, "num_tokens": 647861506.0, "step": 17063 }, { "epoch": 2.1707161938684645, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.613940715789795, "learning_rate": 1e-05, "loss": 0.3738, "mean_token_accuracy": 0.8917467594146729, "num_tokens": 647900039.0, "step": 17064 }, { "epoch": 2.170843404147055, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7115867137908936, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8779733777046204, "num_tokens": 647938368.0, "step": 17065 }, { "epoch": 2.1709706144256455, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6302108764648438, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8694084286689758, "num_tokens": 647972549.0, "step": 17066 }, { "epoch": 2.171097824704236, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6974363327026367, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8772532939910889, "num_tokens": 648016923.0, "step": 17067 }, { "epoch": 2.1712250349828266, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6118175983428955, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8858195543289185, "num_tokens": 648059880.0, "step": 17068 }, { "epoch": 2.171352245261417, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7898950576782227, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.875353217124939, "num_tokens": 648093657.0, "step": 17069 }, { "epoch": 2.1714794555400077, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6043567657470703, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.881811261177063, "num_tokens": 648130723.0, "step": 17070 }, { "epoch": 2.171606665818598, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.740147590637207, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8732621073722839, "num_tokens": 648176657.0, "step": 17071 }, { "epoch": 2.1717338760971887, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.589759588241577, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8847327828407288, "num_tokens": 648214505.0, "step": 17072 }, { "epoch": 2.1718610863757792, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.768585443496704, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8880616426467896, "num_tokens": 648247816.0, "step": 17073 }, { "epoch": 2.1719882966543698, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.614377021789551, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8831419944763184, "num_tokens": 648276418.0, "step": 17074 }, { "epoch": 2.1721155069329603, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.761228084564209, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8699160814285278, "num_tokens": 648314722.0, "step": 17075 }, { "epoch": 2.172242717211551, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6006972789764404, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8730876445770264, "num_tokens": 648354733.0, "step": 17076 }, { "epoch": 2.1723699274901414, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7747621536254883, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8790890574455261, "num_tokens": 648392164.0, "step": 17077 }, { "epoch": 2.172497137768732, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.604827404022217, "learning_rate": 1e-05, "loss": 0.4003, "mean_token_accuracy": 0.8851587772369385, "num_tokens": 648431728.0, "step": 17078 }, { "epoch": 2.1726243480473224, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7749319076538086, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8762936592102051, "num_tokens": 648471225.0, "step": 17079 }, { "epoch": 2.172751558325913, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.604422092437744, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8811298608779907, "num_tokens": 648507742.0, "step": 17080 }, { "epoch": 2.1728787686045035, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7711734771728516, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8774842023849487, "num_tokens": 648548641.0, "step": 17081 }, { "epoch": 2.1730059788830935, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6169683933258057, "learning_rate": 1e-05, "loss": 0.3807, "mean_token_accuracy": 0.8905493021011353, "num_tokens": 648583540.0, "step": 17082 }, { "epoch": 2.173133189161684, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7531278133392334, "learning_rate": 1e-05, "loss": 0.3772, "mean_token_accuracy": 0.8941267132759094, "num_tokens": 648627225.0, "step": 17083 }, { "epoch": 2.1732603994402746, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6037843227386475, "learning_rate": 1e-05, "loss": 0.371, "mean_token_accuracy": 0.8943411707878113, "num_tokens": 648662760.0, "step": 17084 }, { "epoch": 2.173387609718865, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.771002769470215, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8747870922088623, "num_tokens": 648697748.0, "step": 17085 }, { "epoch": 2.1735148199974557, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.618220806121826, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8748961091041565, "num_tokens": 648732981.0, "step": 17086 }, { "epoch": 2.173642030276046, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7732205390930176, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8833728432655334, "num_tokens": 648764356.0, "step": 17087 }, { "epoch": 2.1737692405546367, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.642738103866577, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8779941201210022, "num_tokens": 648809064.0, "step": 17088 }, { "epoch": 2.1738964508332272, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.757568836212158, "learning_rate": 1e-05, "loss": 0.3867, "mean_token_accuracy": 0.8888804912567139, "num_tokens": 648843299.0, "step": 17089 }, { "epoch": 2.1740236611118178, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6153807640075684, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8734485507011414, "num_tokens": 648881251.0, "step": 17090 }, { "epoch": 2.1741508713904083, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.748706340789795, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8767334222793579, "num_tokens": 648915837.0, "step": 17091 }, { "epoch": 2.174278081668999, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6212270259857178, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8705697059631348, "num_tokens": 648953354.0, "step": 17092 }, { "epoch": 2.1744052919475894, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.734254837036133, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8856788873672485, "num_tokens": 648993224.0, "step": 17093 }, { "epoch": 2.17453250222618, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6312804222106934, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8777916431427002, "num_tokens": 649026429.0, "step": 17094 }, { "epoch": 2.1746597125047704, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.732438564300537, "learning_rate": 1e-05, "loss": 0.3971, "mean_token_accuracy": 0.8849962949752808, "num_tokens": 649064672.0, "step": 17095 }, { "epoch": 2.174786922783361, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6323063373565674, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8829230070114136, "num_tokens": 649097926.0, "step": 17096 }, { "epoch": 2.1749141330619515, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.727309226989746, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8780897259712219, "num_tokens": 649138502.0, "step": 17097 }, { "epoch": 2.175041343340542, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6295711994171143, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.873336672782898, "num_tokens": 649175867.0, "step": 17098 }, { "epoch": 2.1751685536191325, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.735370397567749, "learning_rate": 1e-05, "loss": 0.3746, "mean_token_accuracy": 0.8928353786468506, "num_tokens": 649214132.0, "step": 17099 }, { "epoch": 2.175295763897723, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6226580142974854, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8781976103782654, "num_tokens": 649262397.0, "step": 17100 }, { "epoch": 2.1754229741763136, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.753990888595581, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8663389086723328, "num_tokens": 649298165.0, "step": 17101 }, { "epoch": 2.175550184454904, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003185272216796875, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6030917167663574, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.877263605594635, "num_tokens": 649337802.0, "step": 17102 }, { "epoch": 2.1756773947334946, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.744966745376587, "learning_rate": 1e-05, "loss": 0.3852, "mean_token_accuracy": 0.8902338743209839, "num_tokens": 649378922.0, "step": 17103 }, { "epoch": 2.175804605012085, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.602235794067383, "learning_rate": 1e-05, "loss": 0.3764, "mean_token_accuracy": 0.8907489776611328, "num_tokens": 649414332.0, "step": 17104 }, { "epoch": 2.1759318152906753, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7689969539642334, "learning_rate": 1e-05, "loss": 0.3845, "mean_token_accuracy": 0.8879356384277344, "num_tokens": 649448965.0, "step": 17105 }, { "epoch": 2.1760590255692662, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6012299060821533, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.878551185131073, "num_tokens": 649487322.0, "step": 17106 }, { "epoch": 2.1761862358478563, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.782442808151245, "learning_rate": 1e-05, "loss": 0.5045, "mean_token_accuracy": 0.8531071543693542, "num_tokens": 649526524.0, "step": 17107 }, { "epoch": 2.176313446126447, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.608271598815918, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8725818395614624, "num_tokens": 649566169.0, "step": 17108 }, { "epoch": 2.1764406564050374, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.768533945083618, "learning_rate": 1e-05, "loss": 0.3785, "mean_token_accuracy": 0.8906638622283936, "num_tokens": 649603725.0, "step": 17109 }, { "epoch": 2.176567866683628, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.601532459259033, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8759812712669373, "num_tokens": 649644789.0, "step": 17110 }, { "epoch": 2.1766950769622184, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7751336097717285, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8739892244338989, "num_tokens": 649682051.0, "step": 17111 }, { "epoch": 2.176822287240809, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6101183891296387, "learning_rate": 1e-05, "loss": 0.378, "mean_token_accuracy": 0.8937471508979797, "num_tokens": 649715730.0, "step": 17112 }, { "epoch": 2.1769494975193995, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.784027576446533, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8778544664382935, "num_tokens": 649750237.0, "step": 17113 }, { "epoch": 2.17707670779799, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.591860294342041, "learning_rate": 1e-05, "loss": 0.3776, "mean_token_accuracy": 0.8920270204544067, "num_tokens": 649791250.0, "step": 17114 }, { "epoch": 2.1772039180765805, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7790417671203613, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8828765153884888, "num_tokens": 649831750.0, "step": 17115 }, { "epoch": 2.177331128355171, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6047420501708984, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8760216236114502, "num_tokens": 649869245.0, "step": 17116 }, { "epoch": 2.1774583386337616, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7912943363189697, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8780262470245361, "num_tokens": 649903763.0, "step": 17117 }, { "epoch": 2.177585548912352, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5913310050964355, "learning_rate": 1e-05, "loss": 0.3743, "mean_token_accuracy": 0.8924499750137329, "num_tokens": 649946601.0, "step": 17118 }, { "epoch": 2.1777127591909426, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.800975799560547, "learning_rate": 1e-05, "loss": 0.3801, "mean_token_accuracy": 0.891481876373291, "num_tokens": 649979308.0, "step": 17119 }, { "epoch": 2.177839969469533, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.606890916824341, "learning_rate": 1e-05, "loss": 0.3845, "mean_token_accuracy": 0.8915401101112366, "num_tokens": 650018068.0, "step": 17120 }, { "epoch": 2.1779671797481237, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.793158769607544, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8747597932815552, "num_tokens": 650059559.0, "step": 17121 }, { "epoch": 2.1780943900267142, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.587385416030884, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.8838924765586853, "num_tokens": 650103266.0, "step": 17122 }, { "epoch": 2.1782216003053048, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8130297660827637, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8762443661689758, "num_tokens": 650139525.0, "step": 17123 }, { "epoch": 2.1783488105838953, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5890579223632812, "learning_rate": 1e-05, "loss": 0.3704, "mean_token_accuracy": 0.8948038816452026, "num_tokens": 650179249.0, "step": 17124 }, { "epoch": 2.178476020862486, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.82096529006958, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8694615364074707, "num_tokens": 650217258.0, "step": 17125 }, { "epoch": 2.1786032311410763, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.577118158340454, "learning_rate": 1e-05, "loss": 0.353, "mean_token_accuracy": 0.9002852439880371, "num_tokens": 650261923.0, "step": 17126 }, { "epoch": 2.178730441419667, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8363301753997803, "learning_rate": 1e-05, "loss": 0.3954, "mean_token_accuracy": 0.8885247707366943, "num_tokens": 650295562.0, "step": 17127 }, { "epoch": 2.1788576516982574, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6179323196411133, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.878820538520813, "num_tokens": 650331988.0, "step": 17128 }, { "epoch": 2.178984861976848, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8393924236297607, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.88798987865448, "num_tokens": 650370572.0, "step": 17129 }, { "epoch": 2.179112072255438, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5892436504364014, "learning_rate": 1e-05, "loss": 0.3828, "mean_token_accuracy": 0.8900130987167358, "num_tokens": 650412838.0, "step": 17130 }, { "epoch": 2.1792392825340285, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.825502395629883, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8803558945655823, "num_tokens": 650455588.0, "step": 17131 }, { "epoch": 2.179366492812619, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5978167057037354, "learning_rate": 1e-05, "loss": 0.3795, "mean_token_accuracy": 0.8902262449264526, "num_tokens": 650496369.0, "step": 17132 }, { "epoch": 2.1794937030912096, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8376197814941406, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8713347315788269, "num_tokens": 650535366.0, "step": 17133 }, { "epoch": 2.1796209133698, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.607733726501465, "learning_rate": 1e-05, "loss": 0.3821, "mean_token_accuracy": 0.8908641338348389, "num_tokens": 650574171.0, "step": 17134 }, { "epoch": 2.1797481236483907, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8191604614257812, "learning_rate": 1e-05, "loss": 0.3736, "mean_token_accuracy": 0.8953579068183899, "num_tokens": 650612132.0, "step": 17135 }, { "epoch": 2.179875333926981, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6183528900146484, "learning_rate": 1e-05, "loss": 0.378, "mean_token_accuracy": 0.8924264907836914, "num_tokens": 650648813.0, "step": 17136 }, { "epoch": 2.1800025442055717, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8384711742401123, "learning_rate": 1e-05, "loss": 0.4012, "mean_token_accuracy": 0.8867024779319763, "num_tokens": 650678807.0, "step": 17137 }, { "epoch": 2.1801297544841622, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.632810592651367, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8739120364189148, "num_tokens": 650716011.0, "step": 17138 }, { "epoch": 2.1802569647627528, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.80248761177063, "learning_rate": 1e-05, "loss": 0.3631, "mean_token_accuracy": 0.8971114158630371, "num_tokens": 650749763.0, "step": 17139 }, { "epoch": 2.1803841750413433, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6311066150665283, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8800327181816101, "num_tokens": 650786977.0, "step": 17140 }, { "epoch": 2.180511385319934, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7907683849334717, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8704839944839478, "num_tokens": 650823293.0, "step": 17141 }, { "epoch": 2.1806385955985244, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6313350200653076, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8788278102874756, "num_tokens": 650865037.0, "step": 17142 }, { "epoch": 2.180765805877115, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.772035837173462, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8864248991012573, "num_tokens": 650901803.0, "step": 17143 }, { "epoch": 2.1808930161557054, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6192686557769775, "learning_rate": 1e-05, "loss": 0.3904, "mean_token_accuracy": 0.8873648643493652, "num_tokens": 650942163.0, "step": 17144 }, { "epoch": 2.181020226434296, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.772763252258301, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8785319328308105, "num_tokens": 650977989.0, "step": 17145 }, { "epoch": 2.1811474367128865, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6308963298797607, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8844680786132812, "num_tokens": 651016831.0, "step": 17146 }, { "epoch": 2.181274646991477, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7676899433135986, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8755747079849243, "num_tokens": 651052493.0, "step": 17147 }, { "epoch": 2.1814018572700675, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6210734844207764, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8690251111984253, "num_tokens": 651093140.0, "step": 17148 }, { "epoch": 2.181529067548658, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7799274921417236, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.873542308807373, "num_tokens": 651129367.0, "step": 17149 }, { "epoch": 2.1816562778272486, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.611907720565796, "learning_rate": 1e-05, "loss": 0.3986, "mean_token_accuracy": 0.8864828944206238, "num_tokens": 651169573.0, "step": 17150 }, { "epoch": 2.181783488105839, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7599751949310303, "learning_rate": 1e-05, "loss": 0.3832, "mean_token_accuracy": 0.89107745885849, "num_tokens": 651206771.0, "step": 17151 }, { "epoch": 2.1819106983844296, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6084132194519043, "learning_rate": 1e-05, "loss": 0.3827, "mean_token_accuracy": 0.8916148543357849, "num_tokens": 651251322.0, "step": 17152 }, { "epoch": 2.18203790866302, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.779773235321045, "learning_rate": 1e-05, "loss": 0.3898, "mean_token_accuracy": 0.8885985016822815, "num_tokens": 651287495.0, "step": 17153 }, { "epoch": 2.1821651189416107, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6229517459869385, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8690406084060669, "num_tokens": 651326005.0, "step": 17154 }, { "epoch": 2.1822923292202008, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7731380462646484, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8693146109580994, "num_tokens": 651364843.0, "step": 17155 }, { "epoch": 2.1824195394987913, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6009254455566406, "learning_rate": 1e-05, "loss": 0.3828, "mean_token_accuracy": 0.8879196643829346, "num_tokens": 651403981.0, "step": 17156 }, { "epoch": 2.182546749777382, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7766342163085938, "learning_rate": 1e-05, "loss": 0.3861, "mean_token_accuracy": 0.8878901600837708, "num_tokens": 651443751.0, "step": 17157 }, { "epoch": 2.1826739600559724, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6073312759399414, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8766187429428101, "num_tokens": 651487248.0, "step": 17158 }, { "epoch": 2.182801170334563, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.791839122772217, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.870344877243042, "num_tokens": 651523055.0, "step": 17159 }, { "epoch": 2.1829283806131534, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.628546714782715, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8800976276397705, "num_tokens": 651554749.0, "step": 17160 }, { "epoch": 2.183055590891744, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.777377128601074, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8610025644302368, "num_tokens": 651597342.0, "step": 17161 }, { "epoch": 2.1831828011703345, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.619866371154785, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8807401657104492, "num_tokens": 651634578.0, "step": 17162 }, { "epoch": 2.183310011448925, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.754063606262207, "learning_rate": 1e-05, "loss": 0.3851, "mean_token_accuracy": 0.8905321955680847, "num_tokens": 651674561.0, "step": 17163 }, { "epoch": 2.1834372217275155, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.611520767211914, "learning_rate": 1e-05, "loss": 0.382, "mean_token_accuracy": 0.8930642008781433, "num_tokens": 651713426.0, "step": 17164 }, { "epoch": 2.183564432006106, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.755303144454956, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8846776485443115, "num_tokens": 651753267.0, "step": 17165 }, { "epoch": 2.1836916422846966, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.625098943710327, "learning_rate": 1e-05, "loss": 0.3899, "mean_token_accuracy": 0.8874235153198242, "num_tokens": 651797580.0, "step": 17166 }, { "epoch": 2.183818852563287, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.784351348876953, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.875490128993988, "num_tokens": 651828294.0, "step": 17167 }, { "epoch": 2.1839460628418776, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6249935626983643, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8652398586273193, "num_tokens": 651865572.0, "step": 17168 }, { "epoch": 2.184073273120468, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.773853063583374, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8821781873703003, "num_tokens": 651897453.0, "step": 17169 }, { "epoch": 2.1842004833990587, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.615078926086426, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8763495683670044, "num_tokens": 651933096.0, "step": 17170 }, { "epoch": 2.1843276936776492, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7573883533477783, "learning_rate": 1e-05, "loss": 0.3813, "mean_token_accuracy": 0.8885020017623901, "num_tokens": 651970042.0, "step": 17171 }, { "epoch": 2.1844549039562398, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.623837471008301, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8777287602424622, "num_tokens": 652006513.0, "step": 17172 }, { "epoch": 2.1845821142348303, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7556700706481934, "learning_rate": 1e-05, "loss": 0.3839, "mean_token_accuracy": 0.8914850354194641, "num_tokens": 652042706.0, "step": 17173 }, { "epoch": 2.184709324513421, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.602595090866089, "learning_rate": 1e-05, "loss": 0.3832, "mean_token_accuracy": 0.8907644748687744, "num_tokens": 652084219.0, "step": 17174 }, { "epoch": 2.1848365347920113, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7457361221313477, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8901193141937256, "num_tokens": 652130640.0, "step": 17175 }, { "epoch": 2.184963745070602, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.623382568359375, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8759162425994873, "num_tokens": 652169812.0, "step": 17176 }, { "epoch": 2.1850909553491924, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7494492530822754, "learning_rate": 1e-05, "loss": 0.3739, "mean_token_accuracy": 0.8975246548652649, "num_tokens": 652206317.0, "step": 17177 }, { "epoch": 2.185218165627783, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6255762577056885, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8784124255180359, "num_tokens": 652241420.0, "step": 17178 }, { "epoch": 2.1853453759063735, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7372796535491943, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8851104974746704, "num_tokens": 652283951.0, "step": 17179 }, { "epoch": 2.1854725861849635, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.621985673904419, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8834670782089233, "num_tokens": 652320729.0, "step": 17180 }, { "epoch": 2.185599796463554, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.742946147918701, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8761141300201416, "num_tokens": 652362969.0, "step": 17181 }, { "epoch": 2.1857270067421446, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6138851642608643, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8773006200790405, "num_tokens": 652403120.0, "step": 17182 }, { "epoch": 2.185854217020735, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.773454189300537, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8661571145057678, "num_tokens": 652436904.0, "step": 17183 }, { "epoch": 2.1859814272993257, "ewc_loss": 0.072265625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.605700969696045, "learning_rate": 1e-05, "loss": 0.3767, "mean_token_accuracy": 0.8921149969100952, "num_tokens": 652474938.0, "step": 17184 }, { "epoch": 2.186108637577916, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.785787343978882, "learning_rate": 1e-05, "loss": 0.4003, "mean_token_accuracy": 0.8845655918121338, "num_tokens": 652503718.0, "step": 17185 }, { "epoch": 2.1862358478565067, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.599440336227417, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8834014534950256, "num_tokens": 652542235.0, "step": 17186 }, { "epoch": 2.1863630581350972, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7901899814605713, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8655975461006165, "num_tokens": 652584779.0, "step": 17187 }, { "epoch": 2.1864902684136878, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.5845630168914795, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8738275170326233, "num_tokens": 652627132.0, "step": 17188 }, { "epoch": 2.1866174786922783, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.804469347000122, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8744654655456543, "num_tokens": 652670156.0, "step": 17189 }, { "epoch": 2.186744688970869, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.600025177001953, "learning_rate": 1e-05, "loss": 0.3796, "mean_token_accuracy": 0.8910238742828369, "num_tokens": 652705246.0, "step": 17190 }, { "epoch": 2.1868718992494594, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.811605930328369, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8810166120529175, "num_tokens": 652742234.0, "step": 17191 }, { "epoch": 2.18699910952805, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.597911834716797, "learning_rate": 1e-05, "loss": 0.3914, "mean_token_accuracy": 0.8874809741973877, "num_tokens": 652779100.0, "step": 17192 }, { "epoch": 2.1871263198066404, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8086023330688477, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8856371641159058, "num_tokens": 652821174.0, "step": 17193 }, { "epoch": 2.187253530085231, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.605356454849243, "learning_rate": 1e-05, "loss": 0.3734, "mean_token_accuracy": 0.8916972279548645, "num_tokens": 652856271.0, "step": 17194 }, { "epoch": 2.1873807403638215, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8050734996795654, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8790656328201294, "num_tokens": 652898737.0, "step": 17195 }, { "epoch": 2.187507950642412, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6065876483917236, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8841187953948975, "num_tokens": 652938778.0, "step": 17196 }, { "epoch": 2.1876351609210025, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.801361560821533, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8804495334625244, "num_tokens": 652982202.0, "step": 17197 }, { "epoch": 2.187762371199593, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6087534427642822, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8722203373908997, "num_tokens": 653024990.0, "step": 17198 }, { "epoch": 2.1878895814781836, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.810640335083008, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8727503418922424, "num_tokens": 653061913.0, "step": 17199 }, { "epoch": 2.188016791756774, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.598496198654175, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.880872905254364, "num_tokens": 653105944.0, "step": 17200 }, { "epoch": 2.1881440020353646, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.815713882446289, "learning_rate": 1e-05, "loss": 0.3783, "mean_token_accuracy": 0.8939629793167114, "num_tokens": 653139819.0, "step": 17201 }, { "epoch": 2.188271212313955, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5883419513702393, "learning_rate": 1e-05, "loss": 0.3796, "mean_token_accuracy": 0.8925437927246094, "num_tokens": 653184569.0, "step": 17202 }, { "epoch": 2.1883984225925452, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.823641061782837, "learning_rate": 1e-05, "loss": 0.3529, "mean_token_accuracy": 0.9024657011032104, "num_tokens": 653220346.0, "step": 17203 }, { "epoch": 2.188525632871136, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.618845224380493, "learning_rate": 1e-05, "loss": 0.41, "mean_token_accuracy": 0.8799523711204529, "num_tokens": 653257288.0, "step": 17204 }, { "epoch": 2.1886528431497263, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8073251247406006, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.875941812992096, "num_tokens": 653293055.0, "step": 17205 }, { "epoch": 2.188780053428317, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.613780975341797, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8734027147293091, "num_tokens": 653329988.0, "step": 17206 }, { "epoch": 2.1889072637069074, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8024137020111084, "learning_rate": 1e-05, "loss": 0.3802, "mean_token_accuracy": 0.8943349123001099, "num_tokens": 653367704.0, "step": 17207 }, { "epoch": 2.189034473985498, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.619992256164551, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8867265582084656, "num_tokens": 653403915.0, "step": 17208 }, { "epoch": 2.1891616842640884, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.789402723312378, "learning_rate": 1e-05, "loss": 0.3904, "mean_token_accuracy": 0.889559268951416, "num_tokens": 653438045.0, "step": 17209 }, { "epoch": 2.189288894542679, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6372060775756836, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8797545433044434, "num_tokens": 653469552.0, "step": 17210 }, { "epoch": 2.1894161048212695, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7720787525177, "learning_rate": 1e-05, "loss": 0.3971, "mean_token_accuracy": 0.8874112367630005, "num_tokens": 653511809.0, "step": 17211 }, { "epoch": 2.18954331509986, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.62896466255188, "learning_rate": 1e-05, "loss": 0.3702, "mean_token_accuracy": 0.8947175741195679, "num_tokens": 653543812.0, "step": 17212 }, { "epoch": 2.1896705253784505, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.761505126953125, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8722484707832336, "num_tokens": 653586232.0, "step": 17213 }, { "epoch": 2.189797735657041, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6298110485076904, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8781795501708984, "num_tokens": 653626700.0, "step": 17214 }, { "epoch": 2.1899249459356316, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.754769802093506, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8783178925514221, "num_tokens": 653666648.0, "step": 17215 }, { "epoch": 2.190052156214222, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6432175636291504, "learning_rate": 1e-05, "loss": 0.385, "mean_token_accuracy": 0.8893399834632874, "num_tokens": 653698991.0, "step": 17216 }, { "epoch": 2.1901793664928126, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7349252700805664, "learning_rate": 1e-05, "loss": 0.3765, "mean_token_accuracy": 0.8902502655982971, "num_tokens": 653737696.0, "step": 17217 }, { "epoch": 2.190306576771403, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6248698234558105, "learning_rate": 1e-05, "loss": 0.3851, "mean_token_accuracy": 0.8888314366340637, "num_tokens": 653776386.0, "step": 17218 }, { "epoch": 2.1904337870499937, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.741628646850586, "learning_rate": 1e-05, "loss": 0.3774, "mean_token_accuracy": 0.8924403786659241, "num_tokens": 653811346.0, "step": 17219 }, { "epoch": 2.1905609973285842, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6336898803710938, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8746591210365295, "num_tokens": 653847966.0, "step": 17220 }, { "epoch": 2.1906882076071748, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.726670026779175, "learning_rate": 1e-05, "loss": 0.3592, "mean_token_accuracy": 0.8960145115852356, "num_tokens": 653882381.0, "step": 17221 }, { "epoch": 2.1908154178857653, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.625746488571167, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8860864639282227, "num_tokens": 653925026.0, "step": 17222 }, { "epoch": 2.190942628164356, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.724949359893799, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8845788836479187, "num_tokens": 653962210.0, "step": 17223 }, { "epoch": 2.1910698384429463, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6355702877044678, "learning_rate": 1e-05, "loss": 0.3955, "mean_token_accuracy": 0.8836884498596191, "num_tokens": 653998713.0, "step": 17224 }, { "epoch": 2.191197048721537, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.729156017303467, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8765414953231812, "num_tokens": 654033383.0, "step": 17225 }, { "epoch": 2.1913242590001274, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6157400608062744, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8799315690994263, "num_tokens": 654074712.0, "step": 17226 }, { "epoch": 2.191451469278718, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.750699520111084, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8766132593154907, "num_tokens": 654110650.0, "step": 17227 }, { "epoch": 2.191578679557308, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.612581968307495, "learning_rate": 1e-05, "loss": 0.3674, "mean_token_accuracy": 0.893517017364502, "num_tokens": 654146628.0, "step": 17228 }, { "epoch": 2.1917058898358985, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.738701105117798, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8730217218399048, "num_tokens": 654192239.0, "step": 17229 }, { "epoch": 2.191833100114489, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6211869716644287, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8608546257019043, "num_tokens": 654232458.0, "step": 17230 }, { "epoch": 2.1919603103930796, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.750182628631592, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8628664016723633, "num_tokens": 654271266.0, "step": 17231 }, { "epoch": 2.19208752067167, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.607700824737549, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8810285329818726, "num_tokens": 654307801.0, "step": 17232 }, { "epoch": 2.1922147309502606, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7381479740142822, "learning_rate": 1e-05, "loss": 0.3746, "mean_token_accuracy": 0.8954685926437378, "num_tokens": 654348903.0, "step": 17233 }, { "epoch": 2.192341941228851, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6012065410614014, "learning_rate": 1e-05, "loss": 0.3946, "mean_token_accuracy": 0.8863178491592407, "num_tokens": 654388659.0, "step": 17234 }, { "epoch": 2.1924691515074417, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.756934404373169, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.8835803270339966, "num_tokens": 654425231.0, "step": 17235 }, { "epoch": 2.1925963617860322, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6293303966522217, "learning_rate": 1e-05, "loss": 0.3958, "mean_token_accuracy": 0.8867670297622681, "num_tokens": 654457560.0, "step": 17236 }, { "epoch": 2.1927235720646228, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.759477376937866, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8794618844985962, "num_tokens": 654491000.0, "step": 17237 }, { "epoch": 2.1928507823432133, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6214280128479004, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8789665699005127, "num_tokens": 654531142.0, "step": 17238 }, { "epoch": 2.192977992621804, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7429392337799072, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8795072436332703, "num_tokens": 654574133.0, "step": 17239 }, { "epoch": 2.1931052029003943, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6290395259857178, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8749467134475708, "num_tokens": 654612745.0, "step": 17240 }, { "epoch": 2.193232413178985, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.754396915435791, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8846110105514526, "num_tokens": 654646097.0, "step": 17241 }, { "epoch": 2.1933596234575754, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6062679290771484, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8783750534057617, "num_tokens": 654688216.0, "step": 17242 }, { "epoch": 2.193486833736166, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.801483631134033, "learning_rate": 1e-05, "loss": 0.3567, "mean_token_accuracy": 0.9023145437240601, "num_tokens": 654725697.0, "step": 17243 }, { "epoch": 2.1936140440147565, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6077544689178467, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8810933232307434, "num_tokens": 654762440.0, "step": 17244 }, { "epoch": 2.193741254293347, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.790994167327881, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8743155002593994, "num_tokens": 654795843.0, "step": 17245 }, { "epoch": 2.1938684645719375, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.584366798400879, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8850924968719482, "num_tokens": 654835212.0, "step": 17246 }, { "epoch": 2.193995674850528, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.808683395385742, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8735567331314087, "num_tokens": 654869092.0, "step": 17247 }, { "epoch": 2.1941228851291186, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.602072238922119, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8870484828948975, "num_tokens": 654907202.0, "step": 17248 }, { "epoch": 2.194250095407709, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.818667411804199, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8823333978652954, "num_tokens": 654938867.0, "step": 17249 }, { "epoch": 2.1943773056862996, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.59718656539917, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8768636584281921, "num_tokens": 654978838.0, "step": 17250 }, { "epoch": 2.19450451596489, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.833214044570923, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8888990879058838, "num_tokens": 655016987.0, "step": 17251 }, { "epoch": 2.1946317262434807, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.595777988433838, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8824962377548218, "num_tokens": 655056797.0, "step": 17252 }, { "epoch": 2.1947589365220708, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.813830852508545, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8866679072380066, "num_tokens": 655094959.0, "step": 17253 }, { "epoch": 2.1948861468006613, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.602292060852051, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8708086013793945, "num_tokens": 655136490.0, "step": 17254 }, { "epoch": 2.195013357079252, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8204352855682373, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8848965167999268, "num_tokens": 655175174.0, "step": 17255 }, { "epoch": 2.1951405673578424, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6147706508636475, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8747975826263428, "num_tokens": 655214753.0, "step": 17256 }, { "epoch": 2.195267777636433, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8188819885253906, "learning_rate": 1e-05, "loss": 0.3745, "mean_token_accuracy": 0.8943426609039307, "num_tokens": 655249737.0, "step": 17257 }, { "epoch": 2.1953949879150234, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6037566661834717, "learning_rate": 1e-05, "loss": 0.3887, "mean_token_accuracy": 0.8902833461761475, "num_tokens": 655286553.0, "step": 17258 }, { "epoch": 2.195522198193614, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8188536167144775, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8838185667991638, "num_tokens": 655320409.0, "step": 17259 }, { "epoch": 2.1956494084722045, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003204345703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6116721630096436, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8810113668441772, "num_tokens": 655362436.0, "step": 17260 }, { "epoch": 2.195776618750795, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.800060749053955, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8775492310523987, "num_tokens": 655405059.0, "step": 17261 }, { "epoch": 2.1959038290293855, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6125543117523193, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8785708546638489, "num_tokens": 655448688.0, "step": 17262 }, { "epoch": 2.196031039307976, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8019309043884277, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8865420818328857, "num_tokens": 655486361.0, "step": 17263 }, { "epoch": 2.1961582495865666, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.624023199081421, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8831334114074707, "num_tokens": 655518610.0, "step": 17264 }, { "epoch": 2.196285459865157, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8019495010375977, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8783307075500488, "num_tokens": 655557860.0, "step": 17265 }, { "epoch": 2.1964126701437476, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.606144428253174, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8807280659675598, "num_tokens": 655597885.0, "step": 17266 }, { "epoch": 2.196539880422338, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7932803630828857, "learning_rate": 1e-05, "loss": 0.3817, "mean_token_accuracy": 0.8896826505661011, "num_tokens": 655632847.0, "step": 17267 }, { "epoch": 2.1966670907009287, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6268396377563477, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8785777688026428, "num_tokens": 655665374.0, "step": 17268 }, { "epoch": 2.196794300979519, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.769516944885254, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.8862874507904053, "num_tokens": 655705926.0, "step": 17269 }, { "epoch": 2.1969215112581097, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003223419189453125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6140778064727783, "learning_rate": 1e-05, "loss": 0.3698, "mean_token_accuracy": 0.892582893371582, "num_tokens": 655743500.0, "step": 17270 }, { "epoch": 2.1970487215367003, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7591161727905273, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8733296990394592, "num_tokens": 655788063.0, "step": 17271 }, { "epoch": 2.197175931815291, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6187191009521484, "learning_rate": 1e-05, "loss": 0.3791, "mean_token_accuracy": 0.8920432925224304, "num_tokens": 655828393.0, "step": 17272 }, { "epoch": 2.1973031420938813, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7461390495300293, "learning_rate": 1e-05, "loss": 0.3925, "mean_token_accuracy": 0.8873311281204224, "num_tokens": 655871666.0, "step": 17273 }, { "epoch": 2.197430352372472, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.636887788772583, "learning_rate": 1e-05, "loss": 0.3674, "mean_token_accuracy": 0.8965883851051331, "num_tokens": 655905396.0, "step": 17274 }, { "epoch": 2.1975575626510624, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.731727123260498, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8797256350517273, "num_tokens": 655944515.0, "step": 17275 }, { "epoch": 2.197684772929653, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.644636392593384, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.871585488319397, "num_tokens": 655979773.0, "step": 17276 }, { "epoch": 2.1978119832082434, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7336349487304688, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8824732899665833, "num_tokens": 656017786.0, "step": 17277 }, { "epoch": 2.1979391934868335, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6399219036102295, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8591740131378174, "num_tokens": 656058776.0, "step": 17278 }, { "epoch": 2.198066403765424, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7166388034820557, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8871029615402222, "num_tokens": 656100231.0, "step": 17279 }, { "epoch": 2.1981936140440146, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.657273292541504, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8818353414535522, "num_tokens": 656131786.0, "step": 17280 }, { "epoch": 2.198320824322605, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6871497631073, "learning_rate": 1e-05, "loss": 0.3927, "mean_token_accuracy": 0.886159360408783, "num_tokens": 656172041.0, "step": 17281 }, { "epoch": 2.1984480346011956, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.662728786468506, "learning_rate": 1e-05, "loss": 0.3808, "mean_token_accuracy": 0.8895602822303772, "num_tokens": 656203152.0, "step": 17282 }, { "epoch": 2.198575244879786, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.691789150238037, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8755630254745483, "num_tokens": 656241568.0, "step": 17283 }, { "epoch": 2.1987024551583767, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6577308177948, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8676514625549316, "num_tokens": 656275221.0, "step": 17284 }, { "epoch": 2.1988296654369672, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.677715301513672, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8741443157196045, "num_tokens": 656317535.0, "step": 17285 }, { "epoch": 2.1989568757155578, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.670710802078247, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8864536285400391, "num_tokens": 656351440.0, "step": 17286 }, { "epoch": 2.1990840859941483, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.671286106109619, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8787947297096252, "num_tokens": 656390942.0, "step": 17287 }, { "epoch": 2.199211296272739, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6668331623077393, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8880685567855835, "num_tokens": 656427314.0, "step": 17288 }, { "epoch": 2.1993385065513293, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6775715351104736, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8722085952758789, "num_tokens": 656464014.0, "step": 17289 }, { "epoch": 2.19946571682992, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.7030115127563477, "learning_rate": 1e-05, "loss": 0.4012, "mean_token_accuracy": 0.8819590210914612, "num_tokens": 656497783.0, "step": 17290 }, { "epoch": 2.1995929271085104, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.658931016921997, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.878031849861145, "num_tokens": 656538160.0, "step": 17291 }, { "epoch": 2.199720137387101, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6691324710845947, "learning_rate": 1e-05, "loss": 0.3893, "mean_token_accuracy": 0.8888594508171082, "num_tokens": 656578296.0, "step": 17292 }, { "epoch": 2.1998473476656915, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.652406930923462, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8827247619628906, "num_tokens": 656620662.0, "step": 17293 }, { "epoch": 2.199974557944282, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6635048389434814, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8840987682342529, "num_tokens": 656665511.0, "step": 17294 }, { "epoch": 2.2001017682228725, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.662757158279419, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8814870715141296, "num_tokens": 656703079.0, "step": 17295 }, { "epoch": 2.200228978501463, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6748971939086914, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.874549150466919, "num_tokens": 656741430.0, "step": 17296 }, { "epoch": 2.2003561887800536, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6526248455047607, "learning_rate": 1e-05, "loss": 0.3814, "mean_token_accuracy": 0.8901347517967224, "num_tokens": 656778836.0, "step": 17297 }, { "epoch": 2.200483399058644, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6899807453155518, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8842198252677917, "num_tokens": 656811331.0, "step": 17298 }, { "epoch": 2.2006106093372346, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.652446746826172, "learning_rate": 1e-05, "loss": 0.3857, "mean_token_accuracy": 0.8893167972564697, "num_tokens": 656845579.0, "step": 17299 }, { "epoch": 2.200737819615825, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.689011812210083, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8782042860984802, "num_tokens": 656880718.0, "step": 17300 }, { "epoch": 2.2008650298944152, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.639218807220459, "learning_rate": 1e-05, "loss": 0.3986, "mean_token_accuracy": 0.8833675384521484, "num_tokens": 656918350.0, "step": 17301 }, { "epoch": 2.200992240173006, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7027127742767334, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8784192204475403, "num_tokens": 656954661.0, "step": 17302 }, { "epoch": 2.2011194504515963, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6248550415039062, "learning_rate": 1e-05, "loss": 0.3968, "mean_token_accuracy": 0.8847037553787231, "num_tokens": 656993233.0, "step": 17303 }, { "epoch": 2.201246660730187, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.707711935043335, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8775665760040283, "num_tokens": 657032843.0, "step": 17304 }, { "epoch": 2.2013738710087773, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6170730590820312, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8799005746841431, "num_tokens": 657078625.0, "step": 17305 }, { "epoch": 2.201501081287368, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7303545475006104, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8832253217697144, "num_tokens": 657113696.0, "step": 17306 }, { "epoch": 2.2016282915659584, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.626948356628418, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8792498111724854, "num_tokens": 657149224.0, "step": 17307 }, { "epoch": 2.201755501844549, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7474124431610107, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8756704330444336, "num_tokens": 657181946.0, "step": 17308 }, { "epoch": 2.2018827121231395, "ewc_loss": 0.07275390625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.029273986816406e-05, "grad_norm": 2.6526317596435547, "learning_rate": 1e-05, "loss": 0.4811, "mean_token_accuracy": 0.8574333786964417, "num_tokens": 657214372.0, "step": 17309 }, { "epoch": 2.20200992240173, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7178282737731934, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8818376064300537, "num_tokens": 657255572.0, "step": 17310 }, { "epoch": 2.2021371326803205, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6364588737487793, "learning_rate": 1e-05, "loss": 0.3911, "mean_token_accuracy": 0.8819823861122131, "num_tokens": 657288336.0, "step": 17311 }, { "epoch": 2.202264342958911, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7418017387390137, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.877966046333313, "num_tokens": 657327954.0, "step": 17312 }, { "epoch": 2.2023915532375016, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.638167142868042, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8724621534347534, "num_tokens": 657366691.0, "step": 17313 }, { "epoch": 2.202518763516092, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7234034538269043, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8806484937667847, "num_tokens": 657405694.0, "step": 17314 }, { "epoch": 2.2026459737946826, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6080245971679688, "learning_rate": 1e-05, "loss": 0.3615, "mean_token_accuracy": 0.8967787027359009, "num_tokens": 657442461.0, "step": 17315 }, { "epoch": 2.202773184073273, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7404017448425293, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8858072757720947, "num_tokens": 657476384.0, "step": 17316 }, { "epoch": 2.2029003943518637, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6250996589660645, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8725231885910034, "num_tokens": 657511291.0, "step": 17317 }, { "epoch": 2.203027604630454, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.742576837539673, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8738107085227966, "num_tokens": 657546820.0, "step": 17318 }, { "epoch": 2.2031548149090447, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6236379146575928, "learning_rate": 1e-05, "loss": 0.393, "mean_token_accuracy": 0.8852828741073608, "num_tokens": 657582513.0, "step": 17319 }, { "epoch": 2.2032820251876353, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7370736598968506, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8825769424438477, "num_tokens": 657617108.0, "step": 17320 }, { "epoch": 2.203409235466226, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.602013349533081, "learning_rate": 1e-05, "loss": 0.3804, "mean_token_accuracy": 0.8907716274261475, "num_tokens": 657656220.0, "step": 17321 }, { "epoch": 2.2035364457448163, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.747300624847412, "learning_rate": 1e-05, "loss": 0.3846, "mean_token_accuracy": 0.8914390206336975, "num_tokens": 657691386.0, "step": 17322 }, { "epoch": 2.203663656023407, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6223771572113037, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8681268095970154, "num_tokens": 657728118.0, "step": 17323 }, { "epoch": 2.2037908663019974, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.738708257675171, "learning_rate": 1e-05, "loss": 0.3744, "mean_token_accuracy": 0.8957880139350891, "num_tokens": 657768832.0, "step": 17324 }, { "epoch": 2.203918076580588, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6166937351226807, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.8870670795440674, "num_tokens": 657805511.0, "step": 17325 }, { "epoch": 2.204045286859178, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.750885248184204, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8808099031448364, "num_tokens": 657843250.0, "step": 17326 }, { "epoch": 2.2041724971377685, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.613382339477539, "learning_rate": 1e-05, "loss": 0.3751, "mean_token_accuracy": 0.8918624520301819, "num_tokens": 657878882.0, "step": 17327 }, { "epoch": 2.204299707416359, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.741184949874878, "learning_rate": 1e-05, "loss": 0.3551, "mean_token_accuracy": 0.9023301601409912, "num_tokens": 657917007.0, "step": 17328 }, { "epoch": 2.2044269176949496, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6187613010406494, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8705394864082336, "num_tokens": 657956741.0, "step": 17329 }, { "epoch": 2.20455412797354, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.740971326828003, "learning_rate": 1e-05, "loss": 0.3935, "mean_token_accuracy": 0.8852206468582153, "num_tokens": 657995050.0, "step": 17330 }, { "epoch": 2.2046813382521306, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.618183135986328, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8770344853401184, "num_tokens": 658032708.0, "step": 17331 }, { "epoch": 2.204808548530721, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7473130226135254, "learning_rate": 1e-05, "loss": 0.3673, "mean_token_accuracy": 0.8966690301895142, "num_tokens": 658065725.0, "step": 17332 }, { "epoch": 2.2049357588093117, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.608339309692383, "learning_rate": 1e-05, "loss": 0.3822, "mean_token_accuracy": 0.8902400732040405, "num_tokens": 658107290.0, "step": 17333 }, { "epoch": 2.2050629690879022, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.746731996536255, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8849471807479858, "num_tokens": 658147516.0, "step": 17334 }, { "epoch": 2.2051901793664928, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.608504295349121, "learning_rate": 1e-05, "loss": 0.3904, "mean_token_accuracy": 0.8869594931602478, "num_tokens": 658188477.0, "step": 17335 }, { "epoch": 2.2053173896450833, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7715611457824707, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8841668367385864, "num_tokens": 658223517.0, "step": 17336 }, { "epoch": 2.205444599923674, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6198737621307373, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.886860728263855, "num_tokens": 658259167.0, "step": 17337 }, { "epoch": 2.2055718102022643, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.761035680770874, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8847141265869141, "num_tokens": 658294825.0, "step": 17338 }, { "epoch": 2.205699020480855, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.602273464202881, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8746687769889832, "num_tokens": 658336898.0, "step": 17339 }, { "epoch": 2.2058262307594454, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7613749504089355, "learning_rate": 1e-05, "loss": 0.367, "mean_token_accuracy": 0.8971477746963501, "num_tokens": 658373201.0, "step": 17340 }, { "epoch": 2.205953441038036, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.609412431716919, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8752983212471008, "num_tokens": 658413147.0, "step": 17341 }, { "epoch": 2.2060806513166265, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.778902769088745, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8707091808319092, "num_tokens": 658453832.0, "step": 17342 }, { "epoch": 2.206207861595217, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5955049991607666, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8735523819923401, "num_tokens": 658498302.0, "step": 17343 }, { "epoch": 2.2063350718738075, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.787212371826172, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8851577043533325, "num_tokens": 658535017.0, "step": 17344 }, { "epoch": 2.206462282152398, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6032886505126953, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8652306199073792, "num_tokens": 658575495.0, "step": 17345 }, { "epoch": 2.2065894924309886, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7817325592041016, "learning_rate": 1e-05, "loss": 0.3922, "mean_token_accuracy": 0.8890178203582764, "num_tokens": 658619493.0, "step": 17346 }, { "epoch": 2.206716702709579, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.613198757171631, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8811249732971191, "num_tokens": 658654371.0, "step": 17347 }, { "epoch": 2.2068439129881696, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.779764175415039, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8739566802978516, "num_tokens": 658694543.0, "step": 17348 }, { "epoch": 2.20697112326676, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.627809524536133, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8764645457267761, "num_tokens": 658724859.0, "step": 17349 }, { "epoch": 2.2070983335453507, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7667276859283447, "learning_rate": 1e-05, "loss": 0.3782, "mean_token_accuracy": 0.8936856389045715, "num_tokens": 658759941.0, "step": 17350 }, { "epoch": 2.2072255438239408, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6092004776000977, "learning_rate": 1e-05, "loss": 0.3796, "mean_token_accuracy": 0.8898526430130005, "num_tokens": 658801254.0, "step": 17351 }, { "epoch": 2.2073527541025313, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8268332481384277, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8794393539428711, "num_tokens": 658835177.0, "step": 17352 }, { "epoch": 2.207479964381122, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.5945353507995605, "learning_rate": 1e-05, "loss": 0.3945, "mean_token_accuracy": 0.8876839280128479, "num_tokens": 658874636.0, "step": 17353 }, { "epoch": 2.2076071746597123, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8099427223205566, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8762035369873047, "num_tokens": 658911388.0, "step": 17354 }, { "epoch": 2.207734384938303, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5786232948303223, "learning_rate": 1e-05, "loss": 0.3914, "mean_token_accuracy": 0.8874736428260803, "num_tokens": 658954685.0, "step": 17355 }, { "epoch": 2.2078615952168934, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8201937675476074, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8851048350334167, "num_tokens": 658992573.0, "step": 17356 }, { "epoch": 2.207988805495484, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6133060455322266, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8752259612083435, "num_tokens": 659024460.0, "step": 17357 }, { "epoch": 2.2081160157740745, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8226277828216553, "learning_rate": 1e-05, "loss": 0.4785, "mean_token_accuracy": 0.862594485282898, "num_tokens": 659066103.0, "step": 17358 }, { "epoch": 2.208243226052665, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.67175555229187, "learning_rate": 1e-05, "loss": 0.3837, "mean_token_accuracy": 0.8894408941268921, "num_tokens": 659097293.0, "step": 17359 }, { "epoch": 2.2083704363312555, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.763129949569702, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8744879961013794, "num_tokens": 659141221.0, "step": 17360 }, { "epoch": 2.208497646609846, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.626491069793701, "learning_rate": 1e-05, "loss": 0.3914, "mean_token_accuracy": 0.8843887448310852, "num_tokens": 659181026.0, "step": 17361 }, { "epoch": 2.2086248568884366, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7552967071533203, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8762103915214539, "num_tokens": 659218503.0, "step": 17362 }, { "epoch": 2.208752067167027, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.63163685798645, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8867064714431763, "num_tokens": 659255940.0, "step": 17363 }, { "epoch": 2.2088792774456176, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7504000663757324, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.860490083694458, "num_tokens": 659296994.0, "step": 17364 }, { "epoch": 2.209006487724208, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.634695053100586, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8724510669708252, "num_tokens": 659335372.0, "step": 17365 }, { "epoch": 2.2091336980027987, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7574477195739746, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8770239353179932, "num_tokens": 659375954.0, "step": 17366 }, { "epoch": 2.209260908281389, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.629606246948242, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.880583643913269, "num_tokens": 659409995.0, "step": 17367 }, { "epoch": 2.2093881185599797, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.750762462615967, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8881245255470276, "num_tokens": 659445640.0, "step": 17368 }, { "epoch": 2.2095153288385703, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6088716983795166, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8803761005401611, "num_tokens": 659488609.0, "step": 17369 }, { "epoch": 2.209642539117161, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.737879514694214, "learning_rate": 1e-05, "loss": 0.3755, "mean_token_accuracy": 0.8930448889732361, "num_tokens": 659533077.0, "step": 17370 }, { "epoch": 2.2097697493957513, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6284329891204834, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.882398247718811, "num_tokens": 659574080.0, "step": 17371 }, { "epoch": 2.209896959674342, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.746839761734009, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8816274404525757, "num_tokens": 659610508.0, "step": 17372 }, { "epoch": 2.2100241699529324, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.618957281112671, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8799516558647156, "num_tokens": 659648378.0, "step": 17373 }, { "epoch": 2.210151380231523, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7594404220581055, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8724031448364258, "num_tokens": 659681639.0, "step": 17374 }, { "epoch": 2.2102785905101134, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6093063354492188, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8732041716575623, "num_tokens": 659721579.0, "step": 17375 }, { "epoch": 2.2104058007887035, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7520811557769775, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8779135942459106, "num_tokens": 659761577.0, "step": 17376 }, { "epoch": 2.210533011067294, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6086935997009277, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8801429271697998, "num_tokens": 659799393.0, "step": 17377 }, { "epoch": 2.2106602213458846, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.771270990371704, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8628132343292236, "num_tokens": 659839118.0, "step": 17378 }, { "epoch": 2.210787431624475, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6102852821350098, "learning_rate": 1e-05, "loss": 0.3766, "mean_token_accuracy": 0.8933697938919067, "num_tokens": 659874955.0, "step": 17379 }, { "epoch": 2.2109146419030656, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7852771282196045, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8728147149085999, "num_tokens": 659908590.0, "step": 17380 }, { "epoch": 2.211041852181656, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.614182233810425, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.8791446685791016, "num_tokens": 659945753.0, "step": 17381 }, { "epoch": 2.2111690624602467, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7750179767608643, "learning_rate": 1e-05, "loss": 0.3853, "mean_token_accuracy": 0.8900866508483887, "num_tokens": 659981284.0, "step": 17382 }, { "epoch": 2.211296272738837, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.621157169342041, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8794422149658203, "num_tokens": 660016160.0, "step": 17383 }, { "epoch": 2.2114234830174277, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.759586811065674, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8852040767669678, "num_tokens": 660053755.0, "step": 17384 }, { "epoch": 2.2115506932960183, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5892281532287598, "learning_rate": 1e-05, "loss": 0.3468, "mean_token_accuracy": 0.9012178182601929, "num_tokens": 660092808.0, "step": 17385 }, { "epoch": 2.211677903574609, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7863078117370605, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8722164630889893, "num_tokens": 660128996.0, "step": 17386 }, { "epoch": 2.2118051138531993, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6074256896972656, "learning_rate": 1e-05, "loss": 0.3616, "mean_token_accuracy": 0.8975381851196289, "num_tokens": 660163345.0, "step": 17387 }, { "epoch": 2.21193232413179, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7912354469299316, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8849469423294067, "num_tokens": 660196483.0, "step": 17388 }, { "epoch": 2.2120595344103804, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.620800256729126, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8771187663078308, "num_tokens": 660226039.0, "step": 17389 }, { "epoch": 2.212186744688971, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7674801349639893, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8797693252563477, "num_tokens": 660274301.0, "step": 17390 }, { "epoch": 2.2123139549675614, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6146063804626465, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8715885877609253, "num_tokens": 660312893.0, "step": 17391 }, { "epoch": 2.212441165246152, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7728042602539062, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.8884494304656982, "num_tokens": 660349396.0, "step": 17392 }, { "epoch": 2.2125683755247425, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.618525505065918, "learning_rate": 1e-05, "loss": 0.3921, "mean_token_accuracy": 0.8864717483520508, "num_tokens": 660385902.0, "step": 17393 }, { "epoch": 2.212695585803333, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.77823543548584, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8759230375289917, "num_tokens": 660421527.0, "step": 17394 }, { "epoch": 2.2128227960819236, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6098685264587402, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8842580318450928, "num_tokens": 660459687.0, "step": 17395 }, { "epoch": 2.212950006360514, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7808024883270264, "learning_rate": 1e-05, "loss": 0.4678, "mean_token_accuracy": 0.8649129867553711, "num_tokens": 660502415.0, "step": 17396 }, { "epoch": 2.2130772166391046, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6048734188079834, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8830629587173462, "num_tokens": 660542503.0, "step": 17397 }, { "epoch": 2.213204426917695, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7934036254882812, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8750720024108887, "num_tokens": 660576482.0, "step": 17398 }, { "epoch": 2.2133316371962852, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.597707509994507, "learning_rate": 1e-05, "loss": 0.3596, "mean_token_accuracy": 0.8991007804870605, "num_tokens": 660615151.0, "step": 17399 }, { "epoch": 2.213458847474876, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.797719955444336, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8778338432312012, "num_tokens": 660656555.0, "step": 17400 }, { "epoch": 2.2135860577534663, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6097168922424316, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8691362142562866, "num_tokens": 660695991.0, "step": 17401 }, { "epoch": 2.213713268032057, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.813723087310791, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8823589086532593, "num_tokens": 660727840.0, "step": 17402 }, { "epoch": 2.2138404783106473, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.593888282775879, "learning_rate": 1e-05, "loss": 0.3961, "mean_token_accuracy": 0.8862865567207336, "num_tokens": 660770107.0, "step": 17403 }, { "epoch": 2.213967688589238, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8239188194274902, "learning_rate": 1e-05, "loss": 0.4706, "mean_token_accuracy": 0.8637949228286743, "num_tokens": 660809813.0, "step": 17404 }, { "epoch": 2.2140948988678284, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6053240299224854, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8704183101654053, "num_tokens": 660849427.0, "step": 17405 }, { "epoch": 2.214222109146419, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8074281215667725, "learning_rate": 1e-05, "loss": 0.3839, "mean_token_accuracy": 0.8902539014816284, "num_tokens": 660888456.0, "step": 17406 }, { "epoch": 2.2143493194250095, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6001460552215576, "learning_rate": 1e-05, "loss": 0.382, "mean_token_accuracy": 0.8913308382034302, "num_tokens": 660931889.0, "step": 17407 }, { "epoch": 2.2144765297036, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8034801483154297, "learning_rate": 1e-05, "loss": 0.4058, "mean_token_accuracy": 0.8837301731109619, "num_tokens": 660972965.0, "step": 17408 }, { "epoch": 2.2146037399821905, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.617365837097168, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8798748254776001, "num_tokens": 661007352.0, "step": 17409 }, { "epoch": 2.214730950260781, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.799532175064087, "learning_rate": 1e-05, "loss": 0.4021, "mean_token_accuracy": 0.8853923678398132, "num_tokens": 661044202.0, "step": 17410 }, { "epoch": 2.2148581605393716, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6056437492370605, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8808079361915588, "num_tokens": 661088542.0, "step": 17411 }, { "epoch": 2.214985370817962, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7979207038879395, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8718143701553345, "num_tokens": 661130060.0, "step": 17412 }, { "epoch": 2.2151125810965526, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6215403079986572, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.873394250869751, "num_tokens": 661170325.0, "step": 17413 }, { "epoch": 2.215239791375143, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.781385898590088, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8759320378303528, "num_tokens": 661216151.0, "step": 17414 }, { "epoch": 2.2153670016537337, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6069538593292236, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8852900862693787, "num_tokens": 661257857.0, "step": 17415 }, { "epoch": 2.215494211932324, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8006420135498047, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.881949245929718, "num_tokens": 661294113.0, "step": 17416 }, { "epoch": 2.2156214222109147, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6156070232391357, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8770469427108765, "num_tokens": 661332681.0, "step": 17417 }, { "epoch": 2.2157486324895053, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7871389389038086, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8762162923812866, "num_tokens": 661373567.0, "step": 17418 }, { "epoch": 2.215875842768096, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.612732410430908, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8728090524673462, "num_tokens": 661415727.0, "step": 17419 }, { "epoch": 2.2160030530466863, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7792725563049316, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.886357307434082, "num_tokens": 661454934.0, "step": 17420 }, { "epoch": 2.216130263325277, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6224288940429688, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8642113208770752, "num_tokens": 661499070.0, "step": 17421 }, { "epoch": 2.2162574736038674, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.780656099319458, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8847334384918213, "num_tokens": 661538595.0, "step": 17422 }, { "epoch": 2.216384683882458, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.619295358657837, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.871931791305542, "num_tokens": 661578472.0, "step": 17423 }, { "epoch": 2.216511894161048, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.775104284286499, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8747378587722778, "num_tokens": 661615127.0, "step": 17424 }, { "epoch": 2.2166391044396385, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6222803592681885, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8664976358413696, "num_tokens": 661655514.0, "step": 17425 }, { "epoch": 2.216766314718229, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7875218391418457, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8687151074409485, "num_tokens": 661690838.0, "step": 17426 }, { "epoch": 2.2168935249968196, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5966689586639404, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8809221982955933, "num_tokens": 661735697.0, "step": 17427 }, { "epoch": 2.21702073527541, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8165526390075684, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8654400110244751, "num_tokens": 661766875.0, "step": 17428 }, { "epoch": 2.2171479455540006, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.615542411804199, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8803329467773438, "num_tokens": 661798059.0, "step": 17429 }, { "epoch": 2.217275155832591, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7941269874572754, "learning_rate": 1e-05, "loss": 0.3875, "mean_token_accuracy": 0.8895324468612671, "num_tokens": 661833128.0, "step": 17430 }, { "epoch": 2.2174023661111817, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6166229248046875, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8724571466445923, "num_tokens": 661872672.0, "step": 17431 }, { "epoch": 2.217529576389772, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.8005568981170654, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8712460994720459, "num_tokens": 661919611.0, "step": 17432 }, { "epoch": 2.2176567866683627, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.624795436859131, "learning_rate": 1e-05, "loss": 0.4784, "mean_token_accuracy": 0.863594651222229, "num_tokens": 661955349.0, "step": 17433 }, { "epoch": 2.2177839969469533, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7861263751983643, "learning_rate": 1e-05, "loss": 0.3954, "mean_token_accuracy": 0.8861148953437805, "num_tokens": 661990178.0, "step": 17434 }, { "epoch": 2.217911207225544, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.600874423980713, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8837893009185791, "num_tokens": 662035855.0, "step": 17435 }, { "epoch": 2.2180384175041343, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7894697189331055, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8705468773841858, "num_tokens": 662073959.0, "step": 17436 }, { "epoch": 2.218165627782725, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.606935739517212, "learning_rate": 1e-05, "loss": 0.3541, "mean_token_accuracy": 0.8994783163070679, "num_tokens": 662108598.0, "step": 17437 }, { "epoch": 2.2182928380613154, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8239188194274902, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8794548511505127, "num_tokens": 662136316.0, "step": 17438 }, { "epoch": 2.218420048339906, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6074037551879883, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8820985555648804, "num_tokens": 662175771.0, "step": 17439 }, { "epoch": 2.2185472586184964, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7995262145996094, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.876025915145874, "num_tokens": 662219031.0, "step": 17440 }, { "epoch": 2.218674468897087, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.597261667251587, "learning_rate": 1e-05, "loss": 0.3915, "mean_token_accuracy": 0.8861557245254517, "num_tokens": 662257734.0, "step": 17441 }, { "epoch": 2.2188016791756775, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8083364963531494, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8817645311355591, "num_tokens": 662297329.0, "step": 17442 }, { "epoch": 2.218928889454268, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6210384368896484, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.875226616859436, "num_tokens": 662334597.0, "step": 17443 }, { "epoch": 2.2190560997328586, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.800518751144409, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8756195902824402, "num_tokens": 662370369.0, "step": 17444 }, { "epoch": 2.219183310011449, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.628190279006958, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8724161386489868, "num_tokens": 662406193.0, "step": 17445 }, { "epoch": 2.2193105202900396, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8001275062561035, "learning_rate": 1e-05, "loss": 0.3759, "mean_token_accuracy": 0.8931286931037903, "num_tokens": 662439692.0, "step": 17446 }, { "epoch": 2.21943773056863, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.61887264251709, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8852273225784302, "num_tokens": 662479768.0, "step": 17447 }, { "epoch": 2.2195649408472207, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.791923761367798, "learning_rate": 1e-05, "loss": 0.4004, "mean_token_accuracy": 0.8850417137145996, "num_tokens": 662523591.0, "step": 17448 }, { "epoch": 2.2196921511258108, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6300265789031982, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8863140344619751, "num_tokens": 662557693.0, "step": 17449 }, { "epoch": 2.2198193614044013, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7794246673583984, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8802303075790405, "num_tokens": 662595404.0, "step": 17450 }, { "epoch": 2.219946571682992, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6267781257629395, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8798835277557373, "num_tokens": 662637033.0, "step": 17451 }, { "epoch": 2.2200737819615823, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.778541088104248, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8795486688613892, "num_tokens": 662675199.0, "step": 17452 }, { "epoch": 2.220200992240173, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6196606159210205, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8800680637359619, "num_tokens": 662720809.0, "step": 17453 }, { "epoch": 2.2203282025187634, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.768156051635742, "learning_rate": 1e-05, "loss": 0.3757, "mean_token_accuracy": 0.8932790160179138, "num_tokens": 662759894.0, "step": 17454 }, { "epoch": 2.220455412797354, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6190929412841797, "learning_rate": 1e-05, "loss": 0.3812, "mean_token_accuracy": 0.8898152112960815, "num_tokens": 662795373.0, "step": 17455 }, { "epoch": 2.2205826230759445, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7670795917510986, "learning_rate": 1e-05, "loss": 0.3923, "mean_token_accuracy": 0.8887655735015869, "num_tokens": 662832764.0, "step": 17456 }, { "epoch": 2.220709833354535, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.631044626235962, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.862097978591919, "num_tokens": 662872796.0, "step": 17457 }, { "epoch": 2.2208370436331255, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7751283645629883, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8703708052635193, "num_tokens": 662914027.0, "step": 17458 }, { "epoch": 2.220964253911716, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6282360553741455, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8700358271598816, "num_tokens": 662953924.0, "step": 17459 }, { "epoch": 2.2210914641903066, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.792891263961792, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8760067224502563, "num_tokens": 662982792.0, "step": 17460 }, { "epoch": 2.221218674468897, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.608574151992798, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8852533102035522, "num_tokens": 663023674.0, "step": 17461 }, { "epoch": 2.2213458847474876, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7807044982910156, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.871467113494873, "num_tokens": 663064187.0, "step": 17462 }, { "epoch": 2.221473095026078, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.586710214614868, "learning_rate": 1e-05, "loss": 0.3724, "mean_token_accuracy": 0.8942872285842896, "num_tokens": 663109146.0, "step": 17463 }, { "epoch": 2.2216003053046687, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7925913333892822, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.8847919702529907, "num_tokens": 663147988.0, "step": 17464 }, { "epoch": 2.221727515583259, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.60129451751709, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.8853100538253784, "num_tokens": 663190072.0, "step": 17465 }, { "epoch": 2.2218547258618497, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.800701856613159, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8816109895706177, "num_tokens": 663222896.0, "step": 17466 }, { "epoch": 2.2219819361404403, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.606031894683838, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8790423274040222, "num_tokens": 663261846.0, "step": 17467 }, { "epoch": 2.222109146419031, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.799628734588623, "learning_rate": 1e-05, "loss": 0.3614, "mean_token_accuracy": 0.89736407995224, "num_tokens": 663295766.0, "step": 17468 }, { "epoch": 2.2222363566976213, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.600102663040161, "learning_rate": 1e-05, "loss": 0.371, "mean_token_accuracy": 0.8922333717346191, "num_tokens": 663333218.0, "step": 17469 }, { "epoch": 2.222363566976212, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.817384719848633, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.882365882396698, "num_tokens": 663371491.0, "step": 17470 }, { "epoch": 2.2224907772548024, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6016628742218018, "learning_rate": 1e-05, "loss": 0.3778, "mean_token_accuracy": 0.8941577672958374, "num_tokens": 663410824.0, "step": 17471 }, { "epoch": 2.222617987533393, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.825014352798462, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.885035514831543, "num_tokens": 663445624.0, "step": 17472 }, { "epoch": 2.2227451978119834, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.624319076538086, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8707802295684814, "num_tokens": 663483223.0, "step": 17473 }, { "epoch": 2.2228724080905735, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8208839893341064, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8804186582565308, "num_tokens": 663520577.0, "step": 17474 }, { "epoch": 2.222999618369164, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.619936943054199, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8809081315994263, "num_tokens": 663554667.0, "step": 17475 }, { "epoch": 2.2231268286477546, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8035285472869873, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8879390954971313, "num_tokens": 663593475.0, "step": 17476 }, { "epoch": 2.223254038926345, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6026134490966797, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8880094885826111, "num_tokens": 663635957.0, "step": 17477 }, { "epoch": 2.2233812492049356, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.800605058670044, "learning_rate": 1e-05, "loss": 0.3812, "mean_token_accuracy": 0.8932508230209351, "num_tokens": 663678610.0, "step": 17478 }, { "epoch": 2.223508459483526, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6198573112487793, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.8824694156646729, "num_tokens": 663718204.0, "step": 17479 }, { "epoch": 2.2236356697621167, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.792936086654663, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8832110166549683, "num_tokens": 663760501.0, "step": 17480 }, { "epoch": 2.223762880040707, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.615734100341797, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8898928165435791, "num_tokens": 663802515.0, "step": 17481 }, { "epoch": 2.2238900903192977, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8014798164367676, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8738722205162048, "num_tokens": 663839095.0, "step": 17482 }, { "epoch": 2.2240173005978883, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6335666179656982, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8684881925582886, "num_tokens": 663875734.0, "step": 17483 }, { "epoch": 2.224144510876479, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.791321039199829, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.887060821056366, "num_tokens": 663911321.0, "step": 17484 }, { "epoch": 2.2242717211550693, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.639468193054199, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8772836923599243, "num_tokens": 663942334.0, "step": 17485 }, { "epoch": 2.22439893143366, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7785239219665527, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8780468702316284, "num_tokens": 663976306.0, "step": 17486 }, { "epoch": 2.2245261417122504, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.619694471359253, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.8846203684806824, "num_tokens": 664016234.0, "step": 17487 }, { "epoch": 2.224653351990841, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7718098163604736, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8756217360496521, "num_tokens": 664057012.0, "step": 17488 }, { "epoch": 2.2247805622694314, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6242902278900146, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8798546195030212, "num_tokens": 664092911.0, "step": 17489 }, { "epoch": 2.224907772548022, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7763783931732178, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8745787143707275, "num_tokens": 664133069.0, "step": 17490 }, { "epoch": 2.2250349828266125, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6323082447052, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8835158944129944, "num_tokens": 664166059.0, "step": 17491 }, { "epoch": 2.225162193105203, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7812674045562744, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8732327222824097, "num_tokens": 664204918.0, "step": 17492 }, { "epoch": 2.2252894033837936, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.631438970565796, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8829803466796875, "num_tokens": 664240561.0, "step": 17493 }, { "epoch": 2.225416613662384, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.781656503677368, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8742306232452393, "num_tokens": 664274905.0, "step": 17494 }, { "epoch": 2.2255438239409746, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6159842014312744, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8759387731552124, "num_tokens": 664316666.0, "step": 17495 }, { "epoch": 2.225671034219565, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.767671585083008, "learning_rate": 1e-05, "loss": 0.3934, "mean_token_accuracy": 0.8886890411376953, "num_tokens": 664358590.0, "step": 17496 }, { "epoch": 2.225798244498155, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.628063678741455, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8806679248809814, "num_tokens": 664394764.0, "step": 17497 }, { "epoch": 2.225925454776746, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7608706951141357, "learning_rate": 1e-05, "loss": 0.4014, "mean_token_accuracy": 0.8855453729629517, "num_tokens": 664433730.0, "step": 17498 }, { "epoch": 2.2260526650553363, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6296894550323486, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8802478313446045, "num_tokens": 664469926.0, "step": 17499 }, { "epoch": 2.226179875333927, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.762848138809204, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8725705146789551, "num_tokens": 664511204.0, "step": 17500 }, { "epoch": 2.2263070856125173, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6348021030426025, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8632833957672119, "num_tokens": 664547474.0, "step": 17501 }, { "epoch": 2.226434295891108, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.770967721939087, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8795172572135925, "num_tokens": 664579252.0, "step": 17502 }, { "epoch": 2.2265615061696984, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.632474422454834, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8710505962371826, "num_tokens": 664612129.0, "step": 17503 }, { "epoch": 2.226688716448289, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.744912624359131, "learning_rate": 1e-05, "loss": 0.3732, "mean_token_accuracy": 0.8954532742500305, "num_tokens": 664650847.0, "step": 17504 }, { "epoch": 2.2268159267268794, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6239681243896484, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8716749548912048, "num_tokens": 664694056.0, "step": 17505 }, { "epoch": 2.22694313700547, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.763101816177368, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8733068108558655, "num_tokens": 664732868.0, "step": 17506 }, { "epoch": 2.2270703472840605, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.610030174255371, "learning_rate": 1e-05, "loss": 0.3879, "mean_token_accuracy": 0.8883807063102722, "num_tokens": 664772345.0, "step": 17507 }, { "epoch": 2.227197557562651, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.754077911376953, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8745555281639099, "num_tokens": 664811398.0, "step": 17508 }, { "epoch": 2.2273247678412416, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.623835325241089, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8758025169372559, "num_tokens": 664846773.0, "step": 17509 }, { "epoch": 2.227451978119832, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.747910737991333, "learning_rate": 1e-05, "loss": 0.4044, "mean_token_accuracy": 0.8851487040519714, "num_tokens": 664887027.0, "step": 17510 }, { "epoch": 2.2275791883984226, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6188154220581055, "learning_rate": 1e-05, "loss": 0.3729, "mean_token_accuracy": 0.8953638076782227, "num_tokens": 664922798.0, "step": 17511 }, { "epoch": 2.227706398677013, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7410876750946045, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8836194276809692, "num_tokens": 664965862.0, "step": 17512 }, { "epoch": 2.2278336089556037, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.657853603363037, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8629230260848999, "num_tokens": 665006053.0, "step": 17513 }, { "epoch": 2.227960819234194, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.737075090408325, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8808156251907349, "num_tokens": 665040003.0, "step": 17514 }, { "epoch": 2.2280880295127847, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.628723382949829, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.875003457069397, "num_tokens": 665079989.0, "step": 17515 }, { "epoch": 2.2282152397913753, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.742196559906006, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8866778016090393, "num_tokens": 665114009.0, "step": 17516 }, { "epoch": 2.228342450069966, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.622823715209961, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8818776607513428, "num_tokens": 665153259.0, "step": 17517 }, { "epoch": 2.2284696603485563, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.735384702682495, "learning_rate": 1e-05, "loss": 0.3896, "mean_token_accuracy": 0.8906257748603821, "num_tokens": 665195024.0, "step": 17518 }, { "epoch": 2.228596870627147, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6230125427246094, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.881104588508606, "num_tokens": 665236000.0, "step": 17519 }, { "epoch": 2.2287240809057374, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7452259063720703, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.859635591506958, "num_tokens": 665277837.0, "step": 17520 }, { "epoch": 2.228851291184328, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.628129720687866, "learning_rate": 1e-05, "loss": 0.4017, "mean_token_accuracy": 0.8850065469741821, "num_tokens": 665312927.0, "step": 17521 }, { "epoch": 2.228978501462918, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7464637756347656, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8834446668624878, "num_tokens": 665346962.0, "step": 17522 }, { "epoch": 2.2291057117415085, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.611563205718994, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8843632340431213, "num_tokens": 665385684.0, "step": 17523 }, { "epoch": 2.229232922020099, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7470881938934326, "learning_rate": 1e-05, "loss": 0.3524, "mean_token_accuracy": 0.9021296501159668, "num_tokens": 665421194.0, "step": 17524 }, { "epoch": 2.2293601322986896, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.623307704925537, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8785203695297241, "num_tokens": 665459731.0, "step": 17525 }, { "epoch": 2.22948734257728, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7639384269714355, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8829178810119629, "num_tokens": 665490310.0, "step": 17526 }, { "epoch": 2.2296145528558706, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6283137798309326, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8741194009780884, "num_tokens": 665527543.0, "step": 17527 }, { "epoch": 2.229741763134461, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7431435585021973, "learning_rate": 1e-05, "loss": 0.3719, "mean_token_accuracy": 0.890110433101654, "num_tokens": 665562555.0, "step": 17528 }, { "epoch": 2.2298689734130517, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6228549480438232, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8824365139007568, "num_tokens": 665598165.0, "step": 17529 }, { "epoch": 2.229996183691642, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.760655641555786, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8651589751243591, "num_tokens": 665637225.0, "step": 17530 }, { "epoch": 2.2301233939702327, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6196718215942383, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8735029697418213, "num_tokens": 665676126.0, "step": 17531 }, { "epoch": 2.2302506042488233, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7560086250305176, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8849309682846069, "num_tokens": 665713570.0, "step": 17532 }, { "epoch": 2.230377814527414, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6402063369750977, "learning_rate": 1e-05, "loss": 0.3866, "mean_token_accuracy": 0.8888038396835327, "num_tokens": 665743648.0, "step": 17533 }, { "epoch": 2.2305050248060043, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7493155002593994, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.881239652633667, "num_tokens": 665780249.0, "step": 17534 }, { "epoch": 2.230632235084595, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6185266971588135, "learning_rate": 1e-05, "loss": 0.3727, "mean_token_accuracy": 0.8951584100723267, "num_tokens": 665821175.0, "step": 17535 }, { "epoch": 2.2307594453631854, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.748514175415039, "learning_rate": 1e-05, "loss": 0.3724, "mean_token_accuracy": 0.893135666847229, "num_tokens": 665854598.0, "step": 17536 }, { "epoch": 2.230886655641776, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6181492805480957, "learning_rate": 1e-05, "loss": 0.3858, "mean_token_accuracy": 0.8908644914627075, "num_tokens": 665892294.0, "step": 17537 }, { "epoch": 2.2310138659203664, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7566628456115723, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8729733228683472, "num_tokens": 665928575.0, "step": 17538 }, { "epoch": 2.231141076198957, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6210529804229736, "learning_rate": 1e-05, "loss": 0.376, "mean_token_accuracy": 0.8937175273895264, "num_tokens": 665963136.0, "step": 17539 }, { "epoch": 2.2312682864775475, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.737398147583008, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8820542693138123, "num_tokens": 666008556.0, "step": 17540 }, { "epoch": 2.231395496756138, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.621595859527588, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8814005255699158, "num_tokens": 666047066.0, "step": 17541 }, { "epoch": 2.2315227070347285, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7465085983276367, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8844678997993469, "num_tokens": 666084998.0, "step": 17542 }, { "epoch": 2.231649917313319, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6168930530548096, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8768984079360962, "num_tokens": 666124935.0, "step": 17543 }, { "epoch": 2.2317771275919096, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7575199604034424, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8666826486587524, "num_tokens": 666165580.0, "step": 17544 }, { "epoch": 2.2319043378705, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6040778160095215, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8731433153152466, "num_tokens": 666206020.0, "step": 17545 }, { "epoch": 2.2320315481490907, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7783889770507812, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8704797029495239, "num_tokens": 666239757.0, "step": 17546 }, { "epoch": 2.2321587584276807, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.593792200088501, "learning_rate": 1e-05, "loss": 0.3691, "mean_token_accuracy": 0.8958494663238525, "num_tokens": 666278426.0, "step": 17547 }, { "epoch": 2.2322859687062713, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7820370197296143, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8851860761642456, "num_tokens": 666315784.0, "step": 17548 }, { "epoch": 2.232413178984862, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.606032609939575, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.872430145740509, "num_tokens": 666353851.0, "step": 17549 }, { "epoch": 2.2325403892634523, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8013081550598145, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.875219464302063, "num_tokens": 666391208.0, "step": 17550 }, { "epoch": 2.232667599542043, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.590322494506836, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8798898458480835, "num_tokens": 666434477.0, "step": 17551 }, { "epoch": 2.2327948098206334, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.800274133682251, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8789418339729309, "num_tokens": 666472481.0, "step": 17552 }, { "epoch": 2.232922020099224, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.624049186706543, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8792639970779419, "num_tokens": 666502858.0, "step": 17553 }, { "epoch": 2.2330492303778144, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.783761501312256, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8644028902053833, "num_tokens": 666547311.0, "step": 17554 }, { "epoch": 2.233176440656405, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.605173110961914, "learning_rate": 1e-05, "loss": 0.391, "mean_token_accuracy": 0.8891831636428833, "num_tokens": 666587818.0, "step": 17555 }, { "epoch": 2.2333036509349955, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7854599952697754, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8817577958106995, "num_tokens": 666626071.0, "step": 17556 }, { "epoch": 2.233430861213586, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6138432025909424, "learning_rate": 1e-05, "loss": 0.3647, "mean_token_accuracy": 0.8956875801086426, "num_tokens": 666659221.0, "step": 17557 }, { "epoch": 2.2335580714921766, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.793389320373535, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.874629020690918, "num_tokens": 666696189.0, "step": 17558 }, { "epoch": 2.233685281770767, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6093318462371826, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8720237016677856, "num_tokens": 666738495.0, "step": 17559 }, { "epoch": 2.2338124920493576, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8007209300994873, "learning_rate": 1e-05, "loss": 0.3968, "mean_token_accuracy": 0.8851950168609619, "num_tokens": 666772848.0, "step": 17560 }, { "epoch": 2.233939702327948, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5994091033935547, "learning_rate": 1e-05, "loss": 0.3972, "mean_token_accuracy": 0.8828121423721313, "num_tokens": 666813849.0, "step": 17561 }, { "epoch": 2.2340669126065387, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.809136390686035, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8800375461578369, "num_tokens": 666848564.0, "step": 17562 }, { "epoch": 2.234194122885129, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6403536796569824, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8840596079826355, "num_tokens": 666877381.0, "step": 17563 }, { "epoch": 2.2343213331637197, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7778823375701904, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8712630271911621, "num_tokens": 666920452.0, "step": 17564 }, { "epoch": 2.2344485434423103, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6151933670043945, "learning_rate": 1e-05, "loss": 0.3908, "mean_token_accuracy": 0.8862327337265015, "num_tokens": 666958799.0, "step": 17565 }, { "epoch": 2.234575753720901, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7770473957061768, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8838844299316406, "num_tokens": 666993899.0, "step": 17566 }, { "epoch": 2.2347029639994913, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6177315711975098, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8769102096557617, "num_tokens": 667035154.0, "step": 17567 }, { "epoch": 2.234830174278082, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7716567516326904, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8725028038024902, "num_tokens": 667077047.0, "step": 17568 }, { "epoch": 2.2349573845566724, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6112451553344727, "learning_rate": 1e-05, "loss": 0.3968, "mean_token_accuracy": 0.8825422525405884, "num_tokens": 667114298.0, "step": 17569 }, { "epoch": 2.235084594835263, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7650046348571777, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8842728137969971, "num_tokens": 667157942.0, "step": 17570 }, { "epoch": 2.2352118051138534, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6324353218078613, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.871882438659668, "num_tokens": 667195547.0, "step": 17571 }, { "epoch": 2.2353390153924435, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7576258182525635, "learning_rate": 1e-05, "loss": 0.3755, "mean_token_accuracy": 0.8944434523582458, "num_tokens": 667232484.0, "step": 17572 }, { "epoch": 2.235466225671034, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6295578479766846, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8719507455825806, "num_tokens": 667267661.0, "step": 17573 }, { "epoch": 2.2355934359496246, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7680819034576416, "learning_rate": 1e-05, "loss": 0.3762, "mean_token_accuracy": 0.8939325213432312, "num_tokens": 667303963.0, "step": 17574 }, { "epoch": 2.235720646228215, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6176891326904297, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.880630612373352, "num_tokens": 667344103.0, "step": 17575 }, { "epoch": 2.2358478565068056, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.773944616317749, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8818236589431763, "num_tokens": 667377782.0, "step": 17576 }, { "epoch": 2.235975066785396, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6055283546447754, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8824423551559448, "num_tokens": 667419015.0, "step": 17577 }, { "epoch": 2.2361022770639867, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.779844045639038, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8833894729614258, "num_tokens": 667457465.0, "step": 17578 }, { "epoch": 2.236229487342577, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6091842651367188, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.871766209602356, "num_tokens": 667497671.0, "step": 17579 }, { "epoch": 2.2363566976211677, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7898643016815186, "learning_rate": 1e-05, "loss": 0.3796, "mean_token_accuracy": 0.8898958563804626, "num_tokens": 667534258.0, "step": 17580 }, { "epoch": 2.2364839078997583, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6058850288391113, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.8828046321868896, "num_tokens": 667572689.0, "step": 17581 }, { "epoch": 2.236611118178349, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.792876958847046, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8762299418449402, "num_tokens": 667612425.0, "step": 17582 }, { "epoch": 2.2367383284569393, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5978217124938965, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.874243974685669, "num_tokens": 667653309.0, "step": 17583 }, { "epoch": 2.23686553873553, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7846362590789795, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8836255073547363, "num_tokens": 667699174.0, "step": 17584 }, { "epoch": 2.2369927490141204, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6307179927825928, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8828843832015991, "num_tokens": 667738987.0, "step": 17585 }, { "epoch": 2.237119959292711, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.762403726577759, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8841851949691772, "num_tokens": 667780678.0, "step": 17586 }, { "epoch": 2.2372471695713014, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.618762969970703, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8856536149978638, "num_tokens": 667816298.0, "step": 17587 }, { "epoch": 2.237374379849892, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.773181676864624, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8695827722549438, "num_tokens": 667855723.0, "step": 17588 }, { "epoch": 2.2375015901284825, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6505565643310547, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8732519149780273, "num_tokens": 667893895.0, "step": 17589 }, { "epoch": 2.237628800407073, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7417409420013428, "learning_rate": 1e-05, "loss": 0.3694, "mean_token_accuracy": 0.8966555595397949, "num_tokens": 667932061.0, "step": 17590 }, { "epoch": 2.2377560106856635, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6344361305236816, "learning_rate": 1e-05, "loss": 0.3889, "mean_token_accuracy": 0.8876728415489197, "num_tokens": 667965723.0, "step": 17591 }, { "epoch": 2.237883220964254, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7416985034942627, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8769098520278931, "num_tokens": 668002310.0, "step": 17592 }, { "epoch": 2.2380104312428446, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.628305673599243, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8674950003623962, "num_tokens": 668044970.0, "step": 17593 }, { "epoch": 2.238137641521435, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.745945453643799, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8708360195159912, "num_tokens": 668081898.0, "step": 17594 }, { "epoch": 2.238264851800025, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6340487003326416, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8801620006561279, "num_tokens": 668119072.0, "step": 17595 }, { "epoch": 2.238392062078616, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7330212593078613, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8850259184837341, "num_tokens": 668157063.0, "step": 17596 }, { "epoch": 2.2385192723572063, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6470537185668945, "learning_rate": 1e-05, "loss": 0.4758, "mean_token_accuracy": 0.8579509258270264, "num_tokens": 668196985.0, "step": 17597 }, { "epoch": 2.238646482635797, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.719862699508667, "learning_rate": 1e-05, "loss": 0.3923, "mean_token_accuracy": 0.8883577585220337, "num_tokens": 668235954.0, "step": 17598 }, { "epoch": 2.2387736929143873, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6328935623168945, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8788942098617554, "num_tokens": 668275442.0, "step": 17599 }, { "epoch": 2.238900903192978, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.734795093536377, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8825258016586304, "num_tokens": 668311413.0, "step": 17600 }, { "epoch": 2.2390281134715684, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6353797912597656, "learning_rate": 1e-05, "loss": 0.368, "mean_token_accuracy": 0.8933389186859131, "num_tokens": 668346928.0, "step": 17601 }, { "epoch": 2.239155323750159, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.815932035446167, "learning_rate": 1e-05, "loss": 0.3908, "mean_token_accuracy": 0.8901769518852234, "num_tokens": 668388566.0, "step": 17602 }, { "epoch": 2.2392825340287494, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5903234481811523, "learning_rate": 1e-05, "loss": 0.3749, "mean_token_accuracy": 0.8927698135375977, "num_tokens": 668430253.0, "step": 17603 }, { "epoch": 2.23940974430734, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7834465503692627, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8710595965385437, "num_tokens": 668470561.0, "step": 17604 }, { "epoch": 2.2395369545859305, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.586707592010498, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.8802496790885925, "num_tokens": 668509718.0, "step": 17605 }, { "epoch": 2.239664164864521, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8107857704162598, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8765329122543335, "num_tokens": 668548367.0, "step": 17606 }, { "epoch": 2.2397913751431116, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.606631278991699, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8610562682151794, "num_tokens": 668586977.0, "step": 17607 }, { "epoch": 2.239918585421702, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8259518146514893, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8816741704940796, "num_tokens": 668617801.0, "step": 17608 }, { "epoch": 2.2400457957002926, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.605632781982422, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8721204400062561, "num_tokens": 668656074.0, "step": 17609 }, { "epoch": 2.240173005978883, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.7996151447296143, "learning_rate": 1e-05, "loss": 0.39, "mean_token_accuracy": 0.8893963098526001, "num_tokens": 668698354.0, "step": 17610 }, { "epoch": 2.2403002162574737, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.631096839904785, "learning_rate": 1e-05, "loss": 0.3972, "mean_token_accuracy": 0.8865697383880615, "num_tokens": 668734822.0, "step": 17611 }, { "epoch": 2.240427426536064, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7804415225982666, "learning_rate": 1e-05, "loss": 0.35, "mean_token_accuracy": 0.9012615084648132, "num_tokens": 668773136.0, "step": 17612 }, { "epoch": 2.2405546368146547, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6218273639678955, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8800461292266846, "num_tokens": 668811580.0, "step": 17613 }, { "epoch": 2.2406818470932452, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7878127098083496, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.8824042677879333, "num_tokens": 668846743.0, "step": 17614 }, { "epoch": 2.2408090573718358, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6086573600769043, "learning_rate": 1e-05, "loss": 0.3896, "mean_token_accuracy": 0.8871777653694153, "num_tokens": 668885634.0, "step": 17615 }, { "epoch": 2.2409362676504263, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7979490756988525, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8835451006889343, "num_tokens": 668920319.0, "step": 17616 }, { "epoch": 2.241063477929017, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6232423782348633, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8827013969421387, "num_tokens": 668959741.0, "step": 17617 }, { "epoch": 2.2411906882076074, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7980129718780518, "learning_rate": 1e-05, "loss": 0.4904, "mean_token_accuracy": 0.8582244515419006, "num_tokens": 668998469.0, "step": 17618 }, { "epoch": 2.241317898486198, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6336731910705566, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8689585328102112, "num_tokens": 669030018.0, "step": 17619 }, { "epoch": 2.241445108764788, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7752437591552734, "learning_rate": 1e-05, "loss": 0.385, "mean_token_accuracy": 0.890761137008667, "num_tokens": 669066393.0, "step": 17620 }, { "epoch": 2.2415723190433785, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.605902671813965, "learning_rate": 1e-05, "loss": 0.3915, "mean_token_accuracy": 0.8893799781799316, "num_tokens": 669108321.0, "step": 17621 }, { "epoch": 2.241699529321969, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7821810245513916, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8820585012435913, "num_tokens": 669147878.0, "step": 17622 }, { "epoch": 2.2418267396005596, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6143958568573, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8696256875991821, "num_tokens": 669192348.0, "step": 17623 }, { "epoch": 2.24195394987915, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7828099727630615, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8721778988838196, "num_tokens": 669233763.0, "step": 17624 }, { "epoch": 2.2420811601577406, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.618701457977295, "learning_rate": 1e-05, "loss": 0.3885, "mean_token_accuracy": 0.8867464065551758, "num_tokens": 669269609.0, "step": 17625 }, { "epoch": 2.242208370436331, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000324249267578125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7744996547698975, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8784592747688293, "num_tokens": 669308649.0, "step": 17626 }, { "epoch": 2.2423355807149217, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.612006902694702, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8805149793624878, "num_tokens": 669350550.0, "step": 17627 }, { "epoch": 2.242462790993512, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.783167839050293, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8796353936195374, "num_tokens": 669392795.0, "step": 17628 }, { "epoch": 2.2425900012721027, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6344077587127686, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8763604760169983, "num_tokens": 669426928.0, "step": 17629 }, { "epoch": 2.2427172115506933, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7618837356567383, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8873634338378906, "num_tokens": 669467064.0, "step": 17630 }, { "epoch": 2.242844421829284, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6200573444366455, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8722182512283325, "num_tokens": 669506656.0, "step": 17631 }, { "epoch": 2.2429716321078743, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7841546535491943, "learning_rate": 1e-05, "loss": 0.508, "mean_token_accuracy": 0.8515102863311768, "num_tokens": 669545787.0, "step": 17632 }, { "epoch": 2.243098842386465, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6220309734344482, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8791218996047974, "num_tokens": 669581242.0, "step": 17633 }, { "epoch": 2.2432260526650554, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7630293369293213, "learning_rate": 1e-05, "loss": 0.3741, "mean_token_accuracy": 0.8937113881111145, "num_tokens": 669621586.0, "step": 17634 }, { "epoch": 2.243353262943646, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6137540340423584, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.881968080997467, "num_tokens": 669662903.0, "step": 17635 }, { "epoch": 2.2434804732222364, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7767889499664307, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8819856643676758, "num_tokens": 669698409.0, "step": 17636 }, { "epoch": 2.243607683500827, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.616607666015625, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8649300336837769, "num_tokens": 669741344.0, "step": 17637 }, { "epoch": 2.2437348937794175, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7786993980407715, "learning_rate": 1e-05, "loss": 0.3909, "mean_token_accuracy": 0.887923002243042, "num_tokens": 669780042.0, "step": 17638 }, { "epoch": 2.243862104058008, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.61867094039917, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8628383278846741, "num_tokens": 669819340.0, "step": 17639 }, { "epoch": 2.2439893143365985, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7642099857330322, "learning_rate": 1e-05, "loss": 0.3815, "mean_token_accuracy": 0.8909515738487244, "num_tokens": 669859269.0, "step": 17640 }, { "epoch": 2.244116524615189, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6059811115264893, "learning_rate": 1e-05, "loss": 0.4058, "mean_token_accuracy": 0.8837887644767761, "num_tokens": 669903085.0, "step": 17641 }, { "epoch": 2.2442437348937796, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.774721145629883, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8880380392074585, "num_tokens": 669945329.0, "step": 17642 }, { "epoch": 2.24437094517237, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6098196506500244, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8873987197875977, "num_tokens": 669985141.0, "step": 17643 }, { "epoch": 2.2444981554509607, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.786973476409912, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8697075843811035, "num_tokens": 670024034.0, "step": 17644 }, { "epoch": 2.2446253657295507, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.614981174468994, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8837918043136597, "num_tokens": 670062098.0, "step": 17645 }, { "epoch": 2.2447525760081413, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.784599542617798, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8771181702613831, "num_tokens": 670100801.0, "step": 17646 }, { "epoch": 2.244879786286732, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003261566162109375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6073029041290283, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8753042817115784, "num_tokens": 670142221.0, "step": 17647 }, { "epoch": 2.2450069965653223, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.788583517074585, "learning_rate": 1e-05, "loss": 0.398, "mean_token_accuracy": 0.8866533041000366, "num_tokens": 670183131.0, "step": 17648 }, { "epoch": 2.245134206843913, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.597912073135376, "learning_rate": 1e-05, "loss": 0.3577, "mean_token_accuracy": 0.8955438137054443, "num_tokens": 670219207.0, "step": 17649 }, { "epoch": 2.2452614171225034, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8014512062072754, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8809494376182556, "num_tokens": 670254085.0, "step": 17650 }, { "epoch": 2.245388627401094, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6104354858398438, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8697800636291504, "num_tokens": 670299669.0, "step": 17651 }, { "epoch": 2.2455158376796844, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7965304851531982, "learning_rate": 1e-05, "loss": 0.3693, "mean_token_accuracy": 0.895283043384552, "num_tokens": 670332898.0, "step": 17652 }, { "epoch": 2.245643047958275, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.621102809906006, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8810542821884155, "num_tokens": 670371635.0, "step": 17653 }, { "epoch": 2.2457702582368655, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7854442596435547, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8779258728027344, "num_tokens": 670412568.0, "step": 17654 }, { "epoch": 2.245897468515456, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.595162868499756, "learning_rate": 1e-05, "loss": 0.3808, "mean_token_accuracy": 0.891982913017273, "num_tokens": 670454338.0, "step": 17655 }, { "epoch": 2.2460246787940465, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.799530506134033, "learning_rate": 1e-05, "loss": 0.3907, "mean_token_accuracy": 0.8871014714241028, "num_tokens": 670490826.0, "step": 17656 }, { "epoch": 2.246151889072637, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6135661602020264, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8702112436294556, "num_tokens": 670528806.0, "step": 17657 }, { "epoch": 2.2462790993512276, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8631865978240967, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8776475787162781, "num_tokens": 670563349.0, "step": 17658 }, { "epoch": 2.246406309629818, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6004579067230225, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8764175772666931, "num_tokens": 670604682.0, "step": 17659 }, { "epoch": 2.2465335199084087, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.82973313331604, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.870532751083374, "num_tokens": 670644807.0, "step": 17660 }, { "epoch": 2.246660730186999, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5961575508117676, "learning_rate": 1e-05, "loss": 0.4044, "mean_token_accuracy": 0.8835712671279907, "num_tokens": 670683788.0, "step": 17661 }, { "epoch": 2.2467879404655897, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8321290016174316, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.887462317943573, "num_tokens": 670728270.0, "step": 17662 }, { "epoch": 2.2469151507441802, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6198925971984863, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8833000659942627, "num_tokens": 670763203.0, "step": 17663 }, { "epoch": 2.2470423610227708, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8216230869293213, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8731104135513306, "num_tokens": 670804548.0, "step": 17664 }, { "epoch": 2.2471695713013613, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6310105323791504, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8777967691421509, "num_tokens": 670844922.0, "step": 17665 }, { "epoch": 2.247296781579952, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8046481609344482, "learning_rate": 1e-05, "loss": 0.3827, "mean_token_accuracy": 0.8899325132369995, "num_tokens": 670882369.0, "step": 17666 }, { "epoch": 2.2474239918585424, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.620847463607788, "learning_rate": 1e-05, "loss": 0.3818, "mean_token_accuracy": 0.8928079605102539, "num_tokens": 670922716.0, "step": 17667 }, { "epoch": 2.247551202137133, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.800905466079712, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8740310072898865, "num_tokens": 670964618.0, "step": 17668 }, { "epoch": 2.2476784124157234, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6246955394744873, "learning_rate": 1e-05, "loss": 0.3632, "mean_token_accuracy": 0.8993911147117615, "num_tokens": 670997915.0, "step": 17669 }, { "epoch": 2.2478056226943135, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8941588401794434, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8733621835708618, "num_tokens": 671031675.0, "step": 17670 }, { "epoch": 2.247932832972904, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.621896266937256, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8719834089279175, "num_tokens": 671068060.0, "step": 17671 }, { "epoch": 2.2480600432514946, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8239011764526367, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8826426267623901, "num_tokens": 671109014.0, "step": 17672 }, { "epoch": 2.248187253530085, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.627690076828003, "learning_rate": 1e-05, "loss": 0.399, "mean_token_accuracy": 0.8840713500976562, "num_tokens": 671140173.0, "step": 17673 }, { "epoch": 2.2483144638086756, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.812594175338745, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8759539127349854, "num_tokens": 671177644.0, "step": 17674 }, { "epoch": 2.248441674087266, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6188788414001465, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8785386085510254, "num_tokens": 671221860.0, "step": 17675 }, { "epoch": 2.2485688843658567, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.800743818283081, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8806944489479065, "num_tokens": 671262567.0, "step": 17676 }, { "epoch": 2.248696094644447, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6331119537353516, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8750138878822327, "num_tokens": 671301195.0, "step": 17677 }, { "epoch": 2.2488233049230377, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.786308765411377, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8775528073310852, "num_tokens": 671346028.0, "step": 17678 }, { "epoch": 2.2489505152016283, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.621962547302246, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.8864407539367676, "num_tokens": 671386927.0, "step": 17679 }, { "epoch": 2.249077725480219, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7977263927459717, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.875018835067749, "num_tokens": 671422447.0, "step": 17680 }, { "epoch": 2.2492049357588093, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.618178367614746, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.877179741859436, "num_tokens": 671461390.0, "step": 17681 }, { "epoch": 2.2493321460374, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7905445098876953, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8816050291061401, "num_tokens": 671499150.0, "step": 17682 }, { "epoch": 2.2494593563159904, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.618884563446045, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8795084953308105, "num_tokens": 671537529.0, "step": 17683 }, { "epoch": 2.249586566594581, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.788494825363159, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8753048181533813, "num_tokens": 671577661.0, "step": 17684 }, { "epoch": 2.2497137768731714, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.618863344192505, "learning_rate": 1e-05, "loss": 0.39, "mean_token_accuracy": 0.8861589431762695, "num_tokens": 671611375.0, "step": 17685 }, { "epoch": 2.249840987151762, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7781500816345215, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8695241212844849, "num_tokens": 671655430.0, "step": 17686 }, { "epoch": 2.2499681974303525, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.612213373184204, "learning_rate": 1e-05, "loss": 0.3954, "mean_token_accuracy": 0.8849970102310181, "num_tokens": 671696451.0, "step": 17687 }, { "epoch": 2.250095407708943, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7853798866271973, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8736382722854614, "num_tokens": 671736576.0, "step": 17688 }, { "epoch": 2.2502226179875335, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.618042230606079, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8815335035324097, "num_tokens": 671774638.0, "step": 17689 }, { "epoch": 2.250349828266124, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.793438196182251, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8773258924484253, "num_tokens": 671812626.0, "step": 17690 }, { "epoch": 2.2504770385447146, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6345860958099365, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8837325572967529, "num_tokens": 671844752.0, "step": 17691 }, { "epoch": 2.250604248823305, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.785355806350708, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8775084614753723, "num_tokens": 671883161.0, "step": 17692 }, { "epoch": 2.250731459101895, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6082470417022705, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8868257999420166, "num_tokens": 671924386.0, "step": 17693 }, { "epoch": 2.250858669380486, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7968974113464355, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8787726163864136, "num_tokens": 671958452.0, "step": 17694 }, { "epoch": 2.2509858796590763, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.622776985168457, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.875370979309082, "num_tokens": 671995335.0, "step": 17695 }, { "epoch": 2.251113089937667, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.807936906814575, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8755149245262146, "num_tokens": 672025640.0, "step": 17696 }, { "epoch": 2.2512403002162573, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6148641109466553, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8729970455169678, "num_tokens": 672066013.0, "step": 17697 }, { "epoch": 2.251367510494848, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7892205715179443, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.880702018737793, "num_tokens": 672107243.0, "step": 17698 }, { "epoch": 2.2514947207734384, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.609027147293091, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8811720609664917, "num_tokens": 672145753.0, "step": 17699 }, { "epoch": 2.251621931052029, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.794828176498413, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8812474012374878, "num_tokens": 672185489.0, "step": 17700 }, { "epoch": 2.2517491413306194, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6063084602355957, "learning_rate": 1e-05, "loss": 0.3956, "mean_token_accuracy": 0.8840301036834717, "num_tokens": 672225702.0, "step": 17701 }, { "epoch": 2.25187635160921, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7935307025909424, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8758755922317505, "num_tokens": 672265901.0, "step": 17702 }, { "epoch": 2.2520035618878005, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.623133897781372, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8709840774536133, "num_tokens": 672304081.0, "step": 17703 }, { "epoch": 2.252130772166391, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7933716773986816, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8841844201087952, "num_tokens": 672343425.0, "step": 17704 }, { "epoch": 2.2522579824449815, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6286518573760986, "learning_rate": 1e-05, "loss": 0.3972, "mean_token_accuracy": 0.8852885961532593, "num_tokens": 672377745.0, "step": 17705 }, { "epoch": 2.252385192723572, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7905189990997314, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.88960862159729, "num_tokens": 672413811.0, "step": 17706 }, { "epoch": 2.2525124030021626, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.628444194793701, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8769859075546265, "num_tokens": 672455083.0, "step": 17707 }, { "epoch": 2.252639613280753, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7613179683685303, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8875612020492554, "num_tokens": 672502401.0, "step": 17708 }, { "epoch": 2.2527668235593437, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6237640380859375, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8832480311393738, "num_tokens": 672540504.0, "step": 17709 }, { "epoch": 2.252894033837934, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7704429626464844, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8822982907295227, "num_tokens": 672578157.0, "step": 17710 }, { "epoch": 2.2530212441165247, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.643645763397217, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8802828788757324, "num_tokens": 672611859.0, "step": 17711 }, { "epoch": 2.2531484543951152, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7599174976348877, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8766847252845764, "num_tokens": 672651779.0, "step": 17712 }, { "epoch": 2.2532756646737058, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6288304328918457, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8768505454063416, "num_tokens": 672691559.0, "step": 17713 }, { "epoch": 2.2534028749522963, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7550201416015625, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8813210725784302, "num_tokens": 672731353.0, "step": 17714 }, { "epoch": 2.253530085230887, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.621212959289551, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.885414719581604, "num_tokens": 672770427.0, "step": 17715 }, { "epoch": 2.2536572955094774, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7702178955078125, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8787935972213745, "num_tokens": 672808817.0, "step": 17716 }, { "epoch": 2.253784505788068, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.636979103088379, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8818669319152832, "num_tokens": 672848368.0, "step": 17717 }, { "epoch": 2.253911716066658, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7501935958862305, "learning_rate": 1e-05, "loss": 0.3802, "mean_token_accuracy": 0.8924844264984131, "num_tokens": 672885074.0, "step": 17718 }, { "epoch": 2.254038926345249, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.624830722808838, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8822079300880432, "num_tokens": 672924866.0, "step": 17719 }, { "epoch": 2.254166136623839, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.762620449066162, "learning_rate": 1e-05, "loss": 0.3964, "mean_token_accuracy": 0.8857564926147461, "num_tokens": 672962051.0, "step": 17720 }, { "epoch": 2.2542933469024296, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.621544361114502, "learning_rate": 1e-05, "loss": 0.4004, "mean_token_accuracy": 0.884631872177124, "num_tokens": 672999954.0, "step": 17721 }, { "epoch": 2.25442055718102, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.770608901977539, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8662248253822327, "num_tokens": 673037825.0, "step": 17722 }, { "epoch": 2.2545477674596106, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6192328929901123, "learning_rate": 1e-05, "loss": 0.3731, "mean_token_accuracy": 0.8928771018981934, "num_tokens": 673072036.0, "step": 17723 }, { "epoch": 2.254674977738201, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7736239433288574, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8687778115272522, "num_tokens": 673110733.0, "step": 17724 }, { "epoch": 2.2548021880167917, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.608267307281494, "learning_rate": 1e-05, "loss": 0.3838, "mean_token_accuracy": 0.892060399055481, "num_tokens": 673146611.0, "step": 17725 }, { "epoch": 2.254929398295382, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7672789096832275, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8795198202133179, "num_tokens": 673189965.0, "step": 17726 }, { "epoch": 2.2550566085739727, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.628589630126953, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8690740466117859, "num_tokens": 673226990.0, "step": 17727 }, { "epoch": 2.2551838188525632, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.774705648422241, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8766282796859741, "num_tokens": 673263109.0, "step": 17728 }, { "epoch": 2.2553110291311538, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6082513332366943, "learning_rate": 1e-05, "loss": 0.3886, "mean_token_accuracy": 0.8866394758224487, "num_tokens": 673304081.0, "step": 17729 }, { "epoch": 2.2554382394097443, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7679500579833984, "learning_rate": 1e-05, "loss": 0.3943, "mean_token_accuracy": 0.8891547918319702, "num_tokens": 673345510.0, "step": 17730 }, { "epoch": 2.255565449688335, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6354122161865234, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8760654330253601, "num_tokens": 673383607.0, "step": 17731 }, { "epoch": 2.2556926599669254, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.75966739654541, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8773899078369141, "num_tokens": 673422676.0, "step": 17732 }, { "epoch": 2.255819870245516, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6134109497070312, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8807631134986877, "num_tokens": 673466135.0, "step": 17733 }, { "epoch": 2.2559470805241064, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7556815147399902, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8718528747558594, "num_tokens": 673509003.0, "step": 17734 }, { "epoch": 2.256074290802697, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6313164234161377, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8803820610046387, "num_tokens": 673546231.0, "step": 17735 }, { "epoch": 2.2562015010812875, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.755265474319458, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8759120106697083, "num_tokens": 673585812.0, "step": 17736 }, { "epoch": 2.256328711359878, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.634204626083374, "learning_rate": 1e-05, "loss": 0.3838, "mean_token_accuracy": 0.892166018486023, "num_tokens": 673621566.0, "step": 17737 }, { "epoch": 2.2564559216384685, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7295663356781006, "learning_rate": 1e-05, "loss": 0.3584, "mean_token_accuracy": 0.8998589515686035, "num_tokens": 673665659.0, "step": 17738 }, { "epoch": 2.256583131917059, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.623098134994507, "learning_rate": 1e-05, "loss": 0.3886, "mean_token_accuracy": 0.8896907567977905, "num_tokens": 673705936.0, "step": 17739 }, { "epoch": 2.2567103421956496, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.755113363265991, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8744221329689026, "num_tokens": 673740741.0, "step": 17740 }, { "epoch": 2.2568375524742397, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.617011308670044, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8844086527824402, "num_tokens": 673778304.0, "step": 17741 }, { "epoch": 2.2569647627528306, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7497575283050537, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8835124969482422, "num_tokens": 673815136.0, "step": 17742 }, { "epoch": 2.2570919730314207, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6502676010131836, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8838456273078918, "num_tokens": 673847227.0, "step": 17743 }, { "epoch": 2.2572191833100113, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7327353954315186, "learning_rate": 1e-05, "loss": 0.3817, "mean_token_accuracy": 0.8922601938247681, "num_tokens": 673885110.0, "step": 17744 }, { "epoch": 2.257346393588602, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.626171588897705, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8720771074295044, "num_tokens": 673929534.0, "step": 17745 }, { "epoch": 2.2574736038671923, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7382287979125977, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8715751767158508, "num_tokens": 673972353.0, "step": 17746 }, { "epoch": 2.257600814145783, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6235857009887695, "learning_rate": 1e-05, "loss": 0.3726, "mean_token_accuracy": 0.8898171186447144, "num_tokens": 674009441.0, "step": 17747 }, { "epoch": 2.2577280244243734, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7439827919006348, "learning_rate": 1e-05, "loss": 0.3824, "mean_token_accuracy": 0.8907625079154968, "num_tokens": 674045397.0, "step": 17748 }, { "epoch": 2.257855234702964, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6281015872955322, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8736299872398376, "num_tokens": 674084896.0, "step": 17749 }, { "epoch": 2.2579824449815544, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7594263553619385, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8707987666130066, "num_tokens": 674121223.0, "step": 17750 }, { "epoch": 2.258109655260145, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6265366077423096, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8782181143760681, "num_tokens": 674157395.0, "step": 17751 }, { "epoch": 2.2582368655387355, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.771162271499634, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8617401123046875, "num_tokens": 674200157.0, "step": 17752 }, { "epoch": 2.258364075817326, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.625746965408325, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8671507835388184, "num_tokens": 674234696.0, "step": 17753 }, { "epoch": 2.2584912860959165, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7484290599823, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8789153695106506, "num_tokens": 674277456.0, "step": 17754 }, { "epoch": 2.258618496374507, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.617321491241455, "learning_rate": 1e-05, "loss": 0.3792, "mean_token_accuracy": 0.8939179182052612, "num_tokens": 674316913.0, "step": 17755 }, { "epoch": 2.2587457066530976, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.754101037979126, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8831614255905151, "num_tokens": 674355539.0, "step": 17756 }, { "epoch": 2.258872916931688, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6412525177001953, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8730301260948181, "num_tokens": 674386759.0, "step": 17757 }, { "epoch": 2.2590001272102787, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.735746383666992, "learning_rate": 1e-05, "loss": 0.3898, "mean_token_accuracy": 0.8862640857696533, "num_tokens": 674427098.0, "step": 17758 }, { "epoch": 2.259127337488869, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6304616928100586, "learning_rate": 1e-05, "loss": 0.4008, "mean_token_accuracy": 0.8866861462593079, "num_tokens": 674464766.0, "step": 17759 }, { "epoch": 2.2592545477674597, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7268624305725098, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8768380880355835, "num_tokens": 674506260.0, "step": 17760 }, { "epoch": 2.2593817580460502, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.635974168777466, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8832396268844604, "num_tokens": 674544255.0, "step": 17761 }, { "epoch": 2.2595089683246408, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.720721483230591, "learning_rate": 1e-05, "loss": 0.3738, "mean_token_accuracy": 0.8898305296897888, "num_tokens": 674579806.0, "step": 17762 }, { "epoch": 2.2596361786032313, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.647073268890381, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.868794322013855, "num_tokens": 674620734.0, "step": 17763 }, { "epoch": 2.259763388881822, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7086803913116455, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.8861745595932007, "num_tokens": 674661213.0, "step": 17764 }, { "epoch": 2.2598905991604123, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.65590763092041, "learning_rate": 1e-05, "loss": 0.3886, "mean_token_accuracy": 0.8891201019287109, "num_tokens": 674696323.0, "step": 17765 }, { "epoch": 2.2600178094390024, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.726611852645874, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8776900768280029, "num_tokens": 674728041.0, "step": 17766 }, { "epoch": 2.2601450197175934, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.632138729095459, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8865979909896851, "num_tokens": 674767518.0, "step": 17767 }, { "epoch": 2.2602722299961835, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.718513011932373, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.8828538656234741, "num_tokens": 674807510.0, "step": 17768 }, { "epoch": 2.260399440274774, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6274991035461426, "learning_rate": 1e-05, "loss": 0.3987, "mean_token_accuracy": 0.8804179430007935, "num_tokens": 674844851.0, "step": 17769 }, { "epoch": 2.2605266505533645, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7274632453918457, "learning_rate": 1e-05, "loss": 0.3777, "mean_token_accuracy": 0.893649697303772, "num_tokens": 674881839.0, "step": 17770 }, { "epoch": 2.260653860831955, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.642279863357544, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8785975575447083, "num_tokens": 674920450.0, "step": 17771 }, { "epoch": 2.2607810711105456, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.720566987991333, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.8829492330551147, "num_tokens": 674960031.0, "step": 17772 }, { "epoch": 2.260908281389136, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.634491205215454, "learning_rate": 1e-05, "loss": 0.3817, "mean_token_accuracy": 0.8889607191085815, "num_tokens": 674993354.0, "step": 17773 }, { "epoch": 2.2610354916677267, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7278881072998047, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.88435959815979, "num_tokens": 675029186.0, "step": 17774 }, { "epoch": 2.261162701946317, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6313698291778564, "learning_rate": 1e-05, "loss": 0.3763, "mean_token_accuracy": 0.8904147744178772, "num_tokens": 675065170.0, "step": 17775 }, { "epoch": 2.2612899122249077, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.719977617263794, "learning_rate": 1e-05, "loss": 0.3812, "mean_token_accuracy": 0.893000602722168, "num_tokens": 675102774.0, "step": 17776 }, { "epoch": 2.2614171225034982, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.638495922088623, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8752961754798889, "num_tokens": 675136890.0, "step": 17777 }, { "epoch": 2.2615443327820888, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.731795072555542, "learning_rate": 1e-05, "loss": 0.3911, "mean_token_accuracy": 0.887822687625885, "num_tokens": 675171959.0, "step": 17778 }, { "epoch": 2.2616715430606793, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.641815185546875, "learning_rate": 1e-05, "loss": 0.3922, "mean_token_accuracy": 0.8854715824127197, "num_tokens": 675203931.0, "step": 17779 }, { "epoch": 2.26179875333927, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.713385820388794, "learning_rate": 1e-05, "loss": 0.3699, "mean_token_accuracy": 0.8930180072784424, "num_tokens": 675243645.0, "step": 17780 }, { "epoch": 2.2619259636178604, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.642908811569214, "learning_rate": 1e-05, "loss": 0.3942, "mean_token_accuracy": 0.8843364715576172, "num_tokens": 675280318.0, "step": 17781 }, { "epoch": 2.262053173896451, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7099037170410156, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8764814138412476, "num_tokens": 675321442.0, "step": 17782 }, { "epoch": 2.2621803841750414, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.662533760070801, "learning_rate": 1e-05, "loss": 0.3654, "mean_token_accuracy": 0.8955767154693604, "num_tokens": 675351039.0, "step": 17783 }, { "epoch": 2.262307594453632, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7115862369537354, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8842949867248535, "num_tokens": 675385756.0, "step": 17784 }, { "epoch": 2.2624348047322225, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6252098083496094, "learning_rate": 1e-05, "loss": 0.3613, "mean_token_accuracy": 0.8974767327308655, "num_tokens": 675429282.0, "step": 17785 }, { "epoch": 2.262562015010813, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7112438678741455, "learning_rate": 1e-05, "loss": 0.3804, "mean_token_accuracy": 0.890902578830719, "num_tokens": 675466661.0, "step": 17786 }, { "epoch": 2.2626892252894035, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6438536643981934, "learning_rate": 1e-05, "loss": 0.3822, "mean_token_accuracy": 0.8893915414810181, "num_tokens": 675501885.0, "step": 17787 }, { "epoch": 2.262816435567994, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7122466564178467, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8696420192718506, "num_tokens": 675544423.0, "step": 17788 }, { "epoch": 2.2629436458465846, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.63692569732666, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8787581920623779, "num_tokens": 675585301.0, "step": 17789 }, { "epoch": 2.263070856125175, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.714923143386841, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8790162801742554, "num_tokens": 675625244.0, "step": 17790 }, { "epoch": 2.263198066403765, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6423540115356445, "learning_rate": 1e-05, "loss": 0.3733, "mean_token_accuracy": 0.8928452134132385, "num_tokens": 675657310.0, "step": 17791 }, { "epoch": 2.263325276682356, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7290432453155518, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8804089426994324, "num_tokens": 675690702.0, "step": 17792 }, { "epoch": 2.2634524869609463, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6378190517425537, "learning_rate": 1e-05, "loss": 0.3838, "mean_token_accuracy": 0.8904188871383667, "num_tokens": 675725211.0, "step": 17793 }, { "epoch": 2.263579697239537, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.715894937515259, "learning_rate": 1e-05, "loss": 0.3787, "mean_token_accuracy": 0.8927509784698486, "num_tokens": 675762418.0, "step": 17794 }, { "epoch": 2.2637069075181273, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.630077600479126, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8688712120056152, "num_tokens": 675807640.0, "step": 17795 }, { "epoch": 2.263834117796718, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7241008281707764, "learning_rate": 1e-05, "loss": 0.4037, "mean_token_accuracy": 0.882222592830658, "num_tokens": 675843032.0, "step": 17796 }, { "epoch": 2.2639613280753084, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6464381217956543, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8836315870285034, "num_tokens": 675875150.0, "step": 17797 }, { "epoch": 2.264088538353899, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.720710039138794, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8668711185455322, "num_tokens": 675917635.0, "step": 17798 }, { "epoch": 2.2642157486324894, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6305789947509766, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8754475116729736, "num_tokens": 675955577.0, "step": 17799 }, { "epoch": 2.26434295891108, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.725152015686035, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8664171695709229, "num_tokens": 675996326.0, "step": 17800 }, { "epoch": 2.2644701691896705, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.628783702850342, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.866536557674408, "num_tokens": 676040857.0, "step": 17801 }, { "epoch": 2.264597379468261, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7248077392578125, "learning_rate": 1e-05, "loss": 0.3719, "mean_token_accuracy": 0.8921899795532227, "num_tokens": 676077601.0, "step": 17802 }, { "epoch": 2.2647245897468515, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.630763292312622, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8776036500930786, "num_tokens": 676115558.0, "step": 17803 }, { "epoch": 2.264851800025442, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7268857955932617, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8792552351951599, "num_tokens": 676157828.0, "step": 17804 }, { "epoch": 2.2649790103040326, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.0531158447265625e-05, "grad_norm": 2.6368162631988525, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8703357577323914, "num_tokens": 676199147.0, "step": 17805 }, { "epoch": 2.265106220582623, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7128493785858154, "learning_rate": 1e-05, "loss": 0.3762, "mean_token_accuracy": 0.8919141292572021, "num_tokens": 676241271.0, "step": 17806 }, { "epoch": 2.2652334308612136, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.636230707168579, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8810625672340393, "num_tokens": 676276990.0, "step": 17807 }, { "epoch": 2.265360641139804, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7270679473876953, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8803051710128784, "num_tokens": 676313441.0, "step": 17808 }, { "epoch": 2.2654878514183947, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6315882205963135, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8786094188690186, "num_tokens": 676350322.0, "step": 17809 }, { "epoch": 2.2656150616969852, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.723545789718628, "learning_rate": 1e-05, "loss": 0.4021, "mean_token_accuracy": 0.8861066699028015, "num_tokens": 676386113.0, "step": 17810 }, { "epoch": 2.2657422719755758, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6290016174316406, "learning_rate": 1e-05, "loss": 0.363, "mean_token_accuracy": 0.892760694026947, "num_tokens": 676421995.0, "step": 17811 }, { "epoch": 2.2658694822541663, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.735466241836548, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.87928706407547, "num_tokens": 676459761.0, "step": 17812 }, { "epoch": 2.265996692532757, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6187450885772705, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8780287504196167, "num_tokens": 676503870.0, "step": 17813 }, { "epoch": 2.2661239028113473, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.730313777923584, "learning_rate": 1e-05, "loss": 0.3785, "mean_token_accuracy": 0.8908923864364624, "num_tokens": 676544173.0, "step": 17814 }, { "epoch": 2.266251113089938, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6281392574310303, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8830464482307434, "num_tokens": 676580896.0, "step": 17815 }, { "epoch": 2.266378323368528, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7354907989501953, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8864428400993347, "num_tokens": 676620127.0, "step": 17816 }, { "epoch": 2.266505533647119, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.62550687789917, "learning_rate": 1e-05, "loss": 0.3868, "mean_token_accuracy": 0.8903530240058899, "num_tokens": 676655988.0, "step": 17817 }, { "epoch": 2.266632743925709, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.753199338912964, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.879887580871582, "num_tokens": 676689125.0, "step": 17818 }, { "epoch": 2.2667599542042995, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.610447406768799, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8778550624847412, "num_tokens": 676729231.0, "step": 17819 }, { "epoch": 2.26688716448289, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.751882553100586, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8759488463401794, "num_tokens": 676770130.0, "step": 17820 }, { "epoch": 2.2670143747614806, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.611351251602173, "learning_rate": 1e-05, "loss": 0.3955, "mean_token_accuracy": 0.8849995136260986, "num_tokens": 676807847.0, "step": 17821 }, { "epoch": 2.267141585040071, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7722318172454834, "learning_rate": 1e-05, "loss": 0.3908, "mean_token_accuracy": 0.8899168372154236, "num_tokens": 676841050.0, "step": 17822 }, { "epoch": 2.2672687953186617, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.611995220184326, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8768273591995239, "num_tokens": 676877114.0, "step": 17823 }, { "epoch": 2.267396005597252, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.756908893585205, "learning_rate": 1e-05, "loss": 0.3623, "mean_token_accuracy": 0.898286759853363, "num_tokens": 676917913.0, "step": 17824 }, { "epoch": 2.2675232158758427, "ewc_loss": 0.0732421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.60884952545166, "learning_rate": 1e-05, "loss": 0.3671, "mean_token_accuracy": 0.8916776180267334, "num_tokens": 676955974.0, "step": 17825 }, { "epoch": 2.2676504261544332, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.787221908569336, "learning_rate": 1e-05, "loss": 0.3793, "mean_token_accuracy": 0.8923246264457703, "num_tokens": 676991285.0, "step": 17826 }, { "epoch": 2.2677776364330238, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.616377115249634, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8760806322097778, "num_tokens": 677025472.0, "step": 17827 }, { "epoch": 2.2679048467116143, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7744171619415283, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8865010142326355, "num_tokens": 677065047.0, "step": 17828 }, { "epoch": 2.268032056990205, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6110849380493164, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8754655122756958, "num_tokens": 677103027.0, "step": 17829 }, { "epoch": 2.2681592672687954, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7755138874053955, "learning_rate": 1e-05, "loss": 0.3768, "mean_token_accuracy": 0.893919050693512, "num_tokens": 677142302.0, "step": 17830 }, { "epoch": 2.268286477547386, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6022462844848633, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8765190839767456, "num_tokens": 677186963.0, "step": 17831 }, { "epoch": 2.2684136878259764, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.823176145553589, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8609942197799683, "num_tokens": 677232227.0, "step": 17832 }, { "epoch": 2.268540898104567, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.585458517074585, "learning_rate": 1e-05, "loss": 0.3659, "mean_token_accuracy": 0.8980628848075867, "num_tokens": 677274213.0, "step": 17833 }, { "epoch": 2.2686681083831575, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.828730821609497, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8860228657722473, "num_tokens": 677313870.0, "step": 17834 }, { "epoch": 2.268795318661748, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.5877108573913574, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8852488994598389, "num_tokens": 677350478.0, "step": 17835 }, { "epoch": 2.2689225289403385, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8320865631103516, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8766924738883972, "num_tokens": 677392253.0, "step": 17836 }, { "epoch": 2.269049739218929, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5994627475738525, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8883070349693298, "num_tokens": 677429024.0, "step": 17837 }, { "epoch": 2.2691769494975196, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.839357614517212, "learning_rate": 1e-05, "loss": 0.3763, "mean_token_accuracy": 0.8961508274078369, "num_tokens": 677461705.0, "step": 17838 }, { "epoch": 2.2693041597761097, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.593764305114746, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8821097016334534, "num_tokens": 677505586.0, "step": 17839 }, { "epoch": 2.2694313700547006, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.84196400642395, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.88199782371521, "num_tokens": 677544892.0, "step": 17840 }, { "epoch": 2.2695585803332907, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6040048599243164, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8790150880813599, "num_tokens": 677586721.0, "step": 17841 }, { "epoch": 2.2696857906118812, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.853912353515625, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8730401992797852, "num_tokens": 677627657.0, "step": 17842 }, { "epoch": 2.2698130008904718, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6459200382232666, "learning_rate": 1e-05, "loss": 0.3547, "mean_token_accuracy": 0.8985284566879272, "num_tokens": 677667513.0, "step": 17843 }, { "epoch": 2.2699402111690623, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8267507553100586, "learning_rate": 1e-05, "loss": 0.4037, "mean_token_accuracy": 0.8848280310630798, "num_tokens": 677702120.0, "step": 17844 }, { "epoch": 2.270067421447653, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6301329135894775, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8697311282157898, "num_tokens": 677739353.0, "step": 17845 }, { "epoch": 2.2701946317262434, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.803276538848877, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8678074479103088, "num_tokens": 677778464.0, "step": 17846 }, { "epoch": 2.270321842004834, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6389682292938232, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8800456523895264, "num_tokens": 677815169.0, "step": 17847 }, { "epoch": 2.2704490522834244, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.776876211166382, "learning_rate": 1e-05, "loss": 0.3767, "mean_token_accuracy": 0.8924664855003357, "num_tokens": 677852824.0, "step": 17848 }, { "epoch": 2.270576262562015, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6310036182403564, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8720182180404663, "num_tokens": 677895274.0, "step": 17849 }, { "epoch": 2.2707034728406055, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.777547597885132, "learning_rate": 1e-05, "loss": 0.3852, "mean_token_accuracy": 0.8902207612991333, "num_tokens": 677929956.0, "step": 17850 }, { "epoch": 2.270830683119196, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6402697563171387, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8728899359703064, "num_tokens": 677965921.0, "step": 17851 }, { "epoch": 2.2709578933977865, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.75011944770813, "learning_rate": 1e-05, "loss": 0.3789, "mean_token_accuracy": 0.8913888931274414, "num_tokens": 678005701.0, "step": 17852 }, { "epoch": 2.271085103676377, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.650127649307251, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8809130191802979, "num_tokens": 678040275.0, "step": 17853 }, { "epoch": 2.2712123139549676, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.76182222366333, "learning_rate": 1e-05, "loss": 0.4025, "mean_token_accuracy": 0.884243369102478, "num_tokens": 678071963.0, "step": 17854 }, { "epoch": 2.271339524233558, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.644986867904663, "learning_rate": 1e-05, "loss": 0.3986, "mean_token_accuracy": 0.8792549967765808, "num_tokens": 678103422.0, "step": 17855 }, { "epoch": 2.2714667345121486, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.737694263458252, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8821812272071838, "num_tokens": 678144288.0, "step": 17856 }, { "epoch": 2.271593944790739, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6504390239715576, "learning_rate": 1e-05, "loss": 0.3657, "mean_token_accuracy": 0.893531322479248, "num_tokens": 678175575.0, "step": 17857 }, { "epoch": 2.2717211550693297, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7290799617767334, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8682867884635925, "num_tokens": 678218089.0, "step": 17858 }, { "epoch": 2.2718483653479202, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.640897750854492, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8736222982406616, "num_tokens": 678258777.0, "step": 17859 }, { "epoch": 2.2719755756265108, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7184133529663086, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8760244846343994, "num_tokens": 678300354.0, "step": 17860 }, { "epoch": 2.2721027859051013, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.631265640258789, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8785538673400879, "num_tokens": 678343579.0, "step": 17861 }, { "epoch": 2.272229996183692, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.73262619972229, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8829254508018494, "num_tokens": 678377059.0, "step": 17862 }, { "epoch": 2.2723572064622823, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.642192840576172, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8726449012756348, "num_tokens": 678417608.0, "step": 17863 }, { "epoch": 2.2724844167408724, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7273333072662354, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8802964687347412, "num_tokens": 678455473.0, "step": 17864 }, { "epoch": 2.2726116270194634, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.63250994682312, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8694795370101929, "num_tokens": 678500258.0, "step": 17865 }, { "epoch": 2.2727388372980535, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.725454330444336, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8766573071479797, "num_tokens": 678542219.0, "step": 17866 }, { "epoch": 2.272866047576644, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6349945068359375, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8791273236274719, "num_tokens": 678579364.0, "step": 17867 }, { "epoch": 2.2729932578552345, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7360875606536865, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8695977926254272, "num_tokens": 678616549.0, "step": 17868 }, { "epoch": 2.273120468133825, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6309874057769775, "learning_rate": 1e-05, "loss": 0.3669, "mean_token_accuracy": 0.8942942023277283, "num_tokens": 678652324.0, "step": 17869 }, { "epoch": 2.2732476784124156, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.726134777069092, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8786731958389282, "num_tokens": 678693574.0, "step": 17870 }, { "epoch": 2.273374888691006, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6473374366760254, "learning_rate": 1e-05, "loss": 0.4741, "mean_token_accuracy": 0.8655858039855957, "num_tokens": 678730097.0, "step": 17871 }, { "epoch": 2.2735020989695967, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7249457836151123, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8802533149719238, "num_tokens": 678768392.0, "step": 17872 }, { "epoch": 2.273629309248187, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6488053798675537, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8759084343910217, "num_tokens": 678804290.0, "step": 17873 }, { "epoch": 2.2737565195267777, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.711265802383423, "learning_rate": 1e-05, "loss": 0.359, "mean_token_accuracy": 0.9005632400512695, "num_tokens": 678839535.0, "step": 17874 }, { "epoch": 2.2738837298053682, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6390881538391113, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.8830234408378601, "num_tokens": 678879179.0, "step": 17875 }, { "epoch": 2.2740109400839588, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7111496925354004, "learning_rate": 1e-05, "loss": 0.3846, "mean_token_accuracy": 0.890684962272644, "num_tokens": 678916804.0, "step": 17876 }, { "epoch": 2.2741381503625493, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.658069133758545, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8809131383895874, "num_tokens": 678950386.0, "step": 17877 }, { "epoch": 2.27426536064114, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.71400785446167, "learning_rate": 1e-05, "loss": 0.4061, "mean_token_accuracy": 0.8806744813919067, "num_tokens": 678988367.0, "step": 17878 }, { "epoch": 2.2743925709197303, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6485347747802734, "learning_rate": 1e-05, "loss": 0.3778, "mean_token_accuracy": 0.8916301131248474, "num_tokens": 679022577.0, "step": 17879 }, { "epoch": 2.274519781198321, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7075138092041016, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8829430341720581, "num_tokens": 679059613.0, "step": 17880 }, { "epoch": 2.2746469914769114, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6521925926208496, "learning_rate": 1e-05, "loss": 0.3688, "mean_token_accuracy": 0.8937870264053345, "num_tokens": 679092956.0, "step": 17881 }, { "epoch": 2.274774201755502, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7098634243011475, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8825803995132446, "num_tokens": 679131124.0, "step": 17882 }, { "epoch": 2.2749014120340925, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.644031047821045, "learning_rate": 1e-05, "loss": 0.3648, "mean_token_accuracy": 0.8959973454475403, "num_tokens": 679166077.0, "step": 17883 }, { "epoch": 2.275028622312683, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7096731662750244, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8799927830696106, "num_tokens": 679204427.0, "step": 17884 }, { "epoch": 2.2751558325912735, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.00032806396484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.656184434890747, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8829978704452515, "num_tokens": 679236003.0, "step": 17885 }, { "epoch": 2.275283042869864, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.702857494354248, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8826038837432861, "num_tokens": 679276347.0, "step": 17886 }, { "epoch": 2.2754102531484546, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6419894695281982, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8852773904800415, "num_tokens": 679318335.0, "step": 17887 }, { "epoch": 2.275537463427045, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7115964889526367, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8663688898086548, "num_tokens": 679361253.0, "step": 17888 }, { "epoch": 2.275664673705635, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6968023777008057, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8754023909568787, "num_tokens": 679400234.0, "step": 17889 }, { "epoch": 2.275791883984226, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6931636333465576, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8798219561576843, "num_tokens": 679438538.0, "step": 17890 }, { "epoch": 2.2759190942628162, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.661644458770752, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8781858086585999, "num_tokens": 679475425.0, "step": 17891 }, { "epoch": 2.2760463045414068, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.674854040145874, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8860872983932495, "num_tokens": 679517447.0, "step": 17892 }, { "epoch": 2.2761735148199973, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6769368648529053, "learning_rate": 1e-05, "loss": 0.3955, "mean_token_accuracy": 0.8834901452064514, "num_tokens": 679549933.0, "step": 17893 }, { "epoch": 2.276300725098588, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6655097007751465, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8874053359031677, "num_tokens": 679590498.0, "step": 17894 }, { "epoch": 2.2764279353771784, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.660677909851074, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.877576470375061, "num_tokens": 679634582.0, "step": 17895 }, { "epoch": 2.276555145655769, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6641080379486084, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8856557607650757, "num_tokens": 679672324.0, "step": 17896 }, { "epoch": 2.2766823559343594, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6815972328186035, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.8839961290359497, "num_tokens": 679706426.0, "step": 17897 }, { "epoch": 2.27680956621295, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6587162017822266, "learning_rate": 1e-05, "loss": 0.3687, "mean_token_accuracy": 0.8955469131469727, "num_tokens": 679744706.0, "step": 17898 }, { "epoch": 2.2769367764915405, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7023651599884033, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8831173777580261, "num_tokens": 679779939.0, "step": 17899 }, { "epoch": 2.277063986770131, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.650904417037964, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8838111162185669, "num_tokens": 679820286.0, "step": 17900 }, { "epoch": 2.2771911970487215, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.705023765563965, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8757793307304382, "num_tokens": 679857985.0, "step": 17901 }, { "epoch": 2.277318407327312, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.639214038848877, "learning_rate": 1e-05, "loss": 0.3813, "mean_token_accuracy": 0.892125129699707, "num_tokens": 679896379.0, "step": 17902 }, { "epoch": 2.2774456176059026, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.704460859298706, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8725761771202087, "num_tokens": 679933162.0, "step": 17903 }, { "epoch": 2.277572827884493, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6429555416107178, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8658419847488403, "num_tokens": 679975828.0, "step": 17904 }, { "epoch": 2.2777000381630836, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.727267026901245, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8590982556343079, "num_tokens": 680014678.0, "step": 17905 }, { "epoch": 2.277827248441674, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6345467567443848, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8816173076629639, "num_tokens": 680050912.0, "step": 17906 }, { "epoch": 2.2779544587202647, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.73105788230896, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.887557864189148, "num_tokens": 680082591.0, "step": 17907 }, { "epoch": 2.2780816689988552, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6275177001953125, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.8786557912826538, "num_tokens": 680122100.0, "step": 17908 }, { "epoch": 2.2782088792774458, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7314658164978027, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8708903789520264, "num_tokens": 680160926.0, "step": 17909 }, { "epoch": 2.2783360895560363, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6194844245910645, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8771865367889404, "num_tokens": 680202621.0, "step": 17910 }, { "epoch": 2.278463299834627, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.746518135070801, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8795644640922546, "num_tokens": 680241129.0, "step": 17911 }, { "epoch": 2.2785905101132173, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6125426292419434, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8813234567642212, "num_tokens": 680279832.0, "step": 17912 }, { "epoch": 2.278717720391808, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7809622287750244, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8711556792259216, "num_tokens": 680313541.0, "step": 17913 }, { "epoch": 2.278844930670398, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.600092887878418, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8740535378456116, "num_tokens": 680355756.0, "step": 17914 }, { "epoch": 2.278972140948989, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7751193046569824, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8758650422096252, "num_tokens": 680396276.0, "step": 17915 }, { "epoch": 2.279099351227579, "ewc_loss": 0.07373046875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.603848457336426, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8748645186424255, "num_tokens": 680435196.0, "step": 17916 }, { "epoch": 2.2792265615061695, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7809958457946777, "learning_rate": 1e-05, "loss": 0.3829, "mean_token_accuracy": 0.8926641941070557, "num_tokens": 680473088.0, "step": 17917 }, { "epoch": 2.27935377178476, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.630023717880249, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8734610080718994, "num_tokens": 680506269.0, "step": 17918 }, { "epoch": 2.2794809820633506, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.796546220779419, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8662824630737305, "num_tokens": 680543464.0, "step": 17919 }, { "epoch": 2.279608192341941, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.598950147628784, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.875657320022583, "num_tokens": 680582730.0, "step": 17920 }, { "epoch": 2.2797354026205316, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.787092685699463, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8831413984298706, "num_tokens": 680625773.0, "step": 17921 }, { "epoch": 2.279862612899122, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6274161338806152, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.878226101398468, "num_tokens": 680655890.0, "step": 17922 }, { "epoch": 2.2799898231777127, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7891855239868164, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.88190096616745, "num_tokens": 680689921.0, "step": 17923 }, { "epoch": 2.2801170334563032, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6039116382598877, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8702639937400818, "num_tokens": 680735744.0, "step": 17924 }, { "epoch": 2.2802442437348938, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8110697269439697, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8785984516143799, "num_tokens": 680766925.0, "step": 17925 }, { "epoch": 2.2803714540134843, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.605811357498169, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.884890079498291, "num_tokens": 680804324.0, "step": 17926 }, { "epoch": 2.280498664292075, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8001744747161865, "learning_rate": 1e-05, "loss": 0.3936, "mean_token_accuracy": 0.8874297142028809, "num_tokens": 680842005.0, "step": 17927 }, { "epoch": 2.2806258745706653, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6227338314056396, "learning_rate": 1e-05, "loss": 0.3889, "mean_token_accuracy": 0.8866204619407654, "num_tokens": 680873475.0, "step": 17928 }, { "epoch": 2.280753084849256, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.801485776901245, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8866190314292908, "num_tokens": 680911322.0, "step": 17929 }, { "epoch": 2.2808802951278464, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6297600269317627, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.857144832611084, "num_tokens": 680949778.0, "step": 17930 }, { "epoch": 2.281007505406437, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7790770530700684, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8766653537750244, "num_tokens": 680992795.0, "step": 17931 }, { "epoch": 2.2811347156850275, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.640363931655884, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8666555881500244, "num_tokens": 681026430.0, "step": 17932 }, { "epoch": 2.281261925963618, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.752328395843506, "learning_rate": 1e-05, "loss": 0.3116, "mean_token_accuracy": 0.9146533012390137, "num_tokens": 681062998.0, "step": 17933 }, { "epoch": 2.2813891362422085, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6369197368621826, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.876832127571106, "num_tokens": 681100956.0, "step": 17934 }, { "epoch": 2.281516346520799, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7516093254089355, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.885127067565918, "num_tokens": 681136406.0, "step": 17935 }, { "epoch": 2.2816435567993896, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6458487510681152, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8679670095443726, "num_tokens": 681174899.0, "step": 17936 }, { "epoch": 2.2817707670779797, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7583329677581787, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8757585883140564, "num_tokens": 681207662.0, "step": 17937 }, { "epoch": 2.2818979773565706, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6256916522979736, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8726965188980103, "num_tokens": 681248819.0, "step": 17938 }, { "epoch": 2.2820251876351607, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.758737564086914, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8798837661743164, "num_tokens": 681286416.0, "step": 17939 }, { "epoch": 2.2821523979137512, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6247596740722656, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8639213442802429, "num_tokens": 681325629.0, "step": 17940 }, { "epoch": 2.2822796081923418, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7640810012817383, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8662377595901489, "num_tokens": 681362310.0, "step": 17941 }, { "epoch": 2.2824068184709323, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6206417083740234, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8816770315170288, "num_tokens": 681401123.0, "step": 17942 }, { "epoch": 2.282534028749523, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7543327808380127, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8811581134796143, "num_tokens": 681439801.0, "step": 17943 }, { "epoch": 2.2826612390281134, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.62300705909729, "learning_rate": 1e-05, "loss": 0.3667, "mean_token_accuracy": 0.8942067623138428, "num_tokens": 681470032.0, "step": 17944 }, { "epoch": 2.282788449306704, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7486536502838135, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8861297965049744, "num_tokens": 681512421.0, "step": 17945 }, { "epoch": 2.2829156595852944, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.627913236618042, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8787017464637756, "num_tokens": 681550226.0, "step": 17946 }, { "epoch": 2.283042869863885, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.75312876701355, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8868294954299927, "num_tokens": 681587885.0, "step": 17947 }, { "epoch": 2.2831700801424755, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.619805097579956, "learning_rate": 1e-05, "loss": 0.3625, "mean_token_accuracy": 0.8951728343963623, "num_tokens": 681623230.0, "step": 17948 }, { "epoch": 2.283297290421066, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7547168731689453, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8876873254776001, "num_tokens": 681659419.0, "step": 17949 }, { "epoch": 2.2834245006996565, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.621778726577759, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.877501368522644, "num_tokens": 681699845.0, "step": 17950 }, { "epoch": 2.283551710978247, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.769796133041382, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8712974786758423, "num_tokens": 681737579.0, "step": 17951 }, { "epoch": 2.2836789212568376, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6321029663085938, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8724715709686279, "num_tokens": 681771174.0, "step": 17952 }, { "epoch": 2.283806131535428, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.761197328567505, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8745170831680298, "num_tokens": 681810641.0, "step": 17953 }, { "epoch": 2.2839333418140186, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6307473182678223, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8710521459579468, "num_tokens": 681849273.0, "step": 17954 }, { "epoch": 2.284060552092609, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7393743991851807, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.886764645576477, "num_tokens": 681893105.0, "step": 17955 }, { "epoch": 2.2841877623711997, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.623474359512329, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.885637640953064, "num_tokens": 681931364.0, "step": 17956 }, { "epoch": 2.28431497264979, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.751497983932495, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.8841594457626343, "num_tokens": 681970809.0, "step": 17957 }, { "epoch": 2.2844421829283807, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6329548358917236, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8825421333312988, "num_tokens": 682004581.0, "step": 17958 }, { "epoch": 2.2845693932069713, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7548515796661377, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.885614812374115, "num_tokens": 682041809.0, "step": 17959 }, { "epoch": 2.284696603485562, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6293387413024902, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8733841776847839, "num_tokens": 682078058.0, "step": 17960 }, { "epoch": 2.2848238137641523, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.758274555206299, "learning_rate": 1e-05, "loss": 0.3919, "mean_token_accuracy": 0.8899877071380615, "num_tokens": 682111743.0, "step": 17961 }, { "epoch": 2.2849510240427424, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6284196376800537, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8757065534591675, "num_tokens": 682147710.0, "step": 17962 }, { "epoch": 2.2850782343213334, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.755495071411133, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8751327395439148, "num_tokens": 682187139.0, "step": 17963 }, { "epoch": 2.2852054445999235, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6075899600982666, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.887944757938385, "num_tokens": 682226345.0, "step": 17964 }, { "epoch": 2.285332654878514, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7666096687316895, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8840315937995911, "num_tokens": 682262479.0, "step": 17965 }, { "epoch": 2.2854598651571045, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6319186687469482, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8776978850364685, "num_tokens": 682300062.0, "step": 17966 }, { "epoch": 2.285587075435695, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.763404369354248, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8842583298683167, "num_tokens": 682338075.0, "step": 17967 }, { "epoch": 2.2857142857142856, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6260743141174316, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8799577951431274, "num_tokens": 682373036.0, "step": 17968 }, { "epoch": 2.285841495992876, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7645530700683594, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8692395687103271, "num_tokens": 682411131.0, "step": 17969 }, { "epoch": 2.2859687062714666, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6310150623321533, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8603363037109375, "num_tokens": 682449142.0, "step": 17970 }, { "epoch": 2.286095916550057, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.745349168777466, "learning_rate": 1e-05, "loss": 0.3627, "mean_token_accuracy": 0.9002937078475952, "num_tokens": 682486369.0, "step": 17971 }, { "epoch": 2.2862231268286477, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6347551345825195, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8824794292449951, "num_tokens": 682519756.0, "step": 17972 }, { "epoch": 2.2863503371072382, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7511069774627686, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8770731687545776, "num_tokens": 682559930.0, "step": 17973 }, { "epoch": 2.2864775473858288, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.623448371887207, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.87281334400177, "num_tokens": 682601379.0, "step": 17974 }, { "epoch": 2.2866047576644193, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.754822254180908, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8810460567474365, "num_tokens": 682639070.0, "step": 17975 }, { "epoch": 2.28673196794301, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6537089347839355, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8876312375068665, "num_tokens": 682675684.0, "step": 17976 }, { "epoch": 2.2868591782216003, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.746708869934082, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8832851648330688, "num_tokens": 682711026.0, "step": 17977 }, { "epoch": 2.286986388500191, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.636373519897461, "learning_rate": 1e-05, "loss": 0.3609, "mean_token_accuracy": 0.8957811594009399, "num_tokens": 682743074.0, "step": 17978 }, { "epoch": 2.2871135987787814, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7384626865386963, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8826917409896851, "num_tokens": 682778875.0, "step": 17979 }, { "epoch": 2.287240809057372, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6321825981140137, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8840532302856445, "num_tokens": 682817699.0, "step": 17980 }, { "epoch": 2.2873680193359625, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.754591941833496, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8707693815231323, "num_tokens": 682852742.0, "step": 17981 }, { "epoch": 2.287495229614553, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6247315406799316, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8741599321365356, "num_tokens": 682894911.0, "step": 17982 }, { "epoch": 2.2876224398931435, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7556369304656982, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8762447834014893, "num_tokens": 682932387.0, "step": 17983 }, { "epoch": 2.287749650171734, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.626758575439453, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.8866215348243713, "num_tokens": 682965072.0, "step": 17984 }, { "epoch": 2.2878768604503246, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.743698835372925, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8849997520446777, "num_tokens": 683006008.0, "step": 17985 }, { "epoch": 2.288004070728915, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6307592391967773, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.8835338354110718, "num_tokens": 683042043.0, "step": 17986 }, { "epoch": 2.288131281007505, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7543036937713623, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8768997192382812, "num_tokens": 683079429.0, "step": 17987 }, { "epoch": 2.288258491286096, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6272287368774414, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8733668923377991, "num_tokens": 683115686.0, "step": 17988 }, { "epoch": 2.2883857015646862, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.743363618850708, "learning_rate": 1e-05, "loss": 0.3885, "mean_token_accuracy": 0.8903059959411621, "num_tokens": 683156715.0, "step": 17989 }, { "epoch": 2.2885129118432768, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.621791124343872, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8805136680603027, "num_tokens": 683191758.0, "step": 17990 }, { "epoch": 2.2886401221218673, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.772965431213379, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8809882998466492, "num_tokens": 683225146.0, "step": 17991 }, { "epoch": 2.288767332400458, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.625016450881958, "learning_rate": 1e-05, "loss": 0.3859, "mean_token_accuracy": 0.89070725440979, "num_tokens": 683259552.0, "step": 17992 }, { "epoch": 2.2888945426790483, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7531564235687256, "learning_rate": 1e-05, "loss": 0.3905, "mean_token_accuracy": 0.8921647071838379, "num_tokens": 683296925.0, "step": 17993 }, { "epoch": 2.289021752957639, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6538949012756348, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8824400901794434, "num_tokens": 683328836.0, "step": 17994 }, { "epoch": 2.2891489632362294, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7448132038116455, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8872315287590027, "num_tokens": 683359306.0, "step": 17995 }, { "epoch": 2.28927617351482, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6235549449920654, "learning_rate": 1e-05, "loss": 0.359, "mean_token_accuracy": 0.898259162902832, "num_tokens": 683395743.0, "step": 17996 }, { "epoch": 2.2894033837934105, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.728886365890503, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8849481344223022, "num_tokens": 683434528.0, "step": 17997 }, { "epoch": 2.289530594072001, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.687572956085205, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8765192627906799, "num_tokens": 683470540.0, "step": 17998 }, { "epoch": 2.2896578043505915, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7169110774993896, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8773784041404724, "num_tokens": 683502805.0, "step": 17999 }, { "epoch": 2.289785014629182, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.660984516143799, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8811691999435425, "num_tokens": 683540306.0, "step": 18000 }, { "epoch": 2.2899122249077726, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.706922769546509, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.881056547164917, "num_tokens": 683573089.0, "step": 18001 }, { "epoch": 2.290039435186363, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.655728578567505, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8747479915618896, "num_tokens": 683613540.0, "step": 18002 }, { "epoch": 2.2901666454649536, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7005739212036133, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8763669729232788, "num_tokens": 683655016.0, "step": 18003 }, { "epoch": 2.290293855743544, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.651219606399536, "learning_rate": 1e-05, "loss": 0.3751, "mean_token_accuracy": 0.8916434049606323, "num_tokens": 683692607.0, "step": 18004 }, { "epoch": 2.2904210660221347, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.703834056854248, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8835873603820801, "num_tokens": 683726825.0, "step": 18005 }, { "epoch": 2.290548276300725, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.668074131011963, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8662285208702087, "num_tokens": 683763905.0, "step": 18006 }, { "epoch": 2.2906754865793157, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6940062046051025, "learning_rate": 1e-05, "loss": 0.448, "mean_token_accuracy": 0.8699127435684204, "num_tokens": 683806926.0, "step": 18007 }, { "epoch": 2.2908026968579063, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6629905700683594, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8763880729675293, "num_tokens": 683846069.0, "step": 18008 }, { "epoch": 2.290929907136497, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7059600353240967, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.878806471824646, "num_tokens": 683887530.0, "step": 18009 }, { "epoch": 2.2910571174150873, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.659261465072632, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.875318169593811, "num_tokens": 683925721.0, "step": 18010 }, { "epoch": 2.291184327693678, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6852948665618896, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8841606378555298, "num_tokens": 683967867.0, "step": 18011 }, { "epoch": 2.291311537972268, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.655991315841675, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8788324594497681, "num_tokens": 684006360.0, "step": 18012 }, { "epoch": 2.291438748250859, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6829569339752197, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8868004679679871, "num_tokens": 684047559.0, "step": 18013 }, { "epoch": 2.291565958529449, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.64756178855896, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8761991262435913, "num_tokens": 684092028.0, "step": 18014 }, { "epoch": 2.2916931688080395, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.70684814453125, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8709500432014465, "num_tokens": 684131441.0, "step": 18015 }, { "epoch": 2.29182037908663, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.657992124557495, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.8853164315223694, "num_tokens": 684163864.0, "step": 18016 }, { "epoch": 2.2919475893652206, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.690321207046509, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8818599581718445, "num_tokens": 684200962.0, "step": 18017 }, { "epoch": 2.292074799643811, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.646275758743286, "learning_rate": 1e-05, "loss": 0.3832, "mean_token_accuracy": 0.8899509906768799, "num_tokens": 684237060.0, "step": 18018 }, { "epoch": 2.2922020099224016, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.709751844406128, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8846207857131958, "num_tokens": 684271299.0, "step": 18019 }, { "epoch": 2.292329220200992, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.64959716796875, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8732905387878418, "num_tokens": 684310008.0, "step": 18020 }, { "epoch": 2.2924564304795827, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.702352523803711, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8799811005592346, "num_tokens": 684348477.0, "step": 18021 }, { "epoch": 2.2925836407581732, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6566734313964844, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.8715658187866211, "num_tokens": 684383090.0, "step": 18022 }, { "epoch": 2.2927108510367638, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7303614616394043, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.8633261919021606, "num_tokens": 684415142.0, "step": 18023 }, { "epoch": 2.2928380613153543, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6297574043273926, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8816649913787842, "num_tokens": 684455878.0, "step": 18024 }, { "epoch": 2.292965271593945, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.713046073913574, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8743547201156616, "num_tokens": 684501140.0, "step": 18025 }, { "epoch": 2.2930924818725353, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6352570056915283, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8763829469680786, "num_tokens": 684542682.0, "step": 18026 }, { "epoch": 2.293219692151126, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.730475902557373, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8718749284744263, "num_tokens": 684581301.0, "step": 18027 }, { "epoch": 2.2933469024297164, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6246249675750732, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8885107040405273, "num_tokens": 684620124.0, "step": 18028 }, { "epoch": 2.293474112708307, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7362568378448486, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8737185001373291, "num_tokens": 684657237.0, "step": 18029 }, { "epoch": 2.2936013229868975, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6177964210510254, "learning_rate": 1e-05, "loss": 0.3926, "mean_token_accuracy": 0.8853954076766968, "num_tokens": 684696717.0, "step": 18030 }, { "epoch": 2.293728533265488, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7423412799835205, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8791407346725464, "num_tokens": 684734823.0, "step": 18031 }, { "epoch": 2.2938557435440785, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.626404047012329, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8785573244094849, "num_tokens": 684773036.0, "step": 18032 }, { "epoch": 2.293982953822669, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7440614700317383, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.884475588798523, "num_tokens": 684811210.0, "step": 18033 }, { "epoch": 2.2941101641012596, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.618258476257324, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8816524744033813, "num_tokens": 684850656.0, "step": 18034 }, { "epoch": 2.2942373743798496, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7465100288391113, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8784821033477783, "num_tokens": 684890113.0, "step": 18035 }, { "epoch": 2.2943645846584406, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.608808755874634, "learning_rate": 1e-05, "loss": 0.3935, "mean_token_accuracy": 0.8865823745727539, "num_tokens": 684930418.0, "step": 18036 }, { "epoch": 2.2944917949370307, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7632319927215576, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8793862462043762, "num_tokens": 684967068.0, "step": 18037 }, { "epoch": 2.2946190052156212, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.620800733566284, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8584072589874268, "num_tokens": 685007602.0, "step": 18038 }, { "epoch": 2.2947462154942118, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.766435384750366, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8830999135971069, "num_tokens": 685043332.0, "step": 18039 }, { "epoch": 2.2948734257728023, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.604220151901245, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.8839974403381348, "num_tokens": 685078327.0, "step": 18040 }, { "epoch": 2.295000636051393, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7615599632263184, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8814414739608765, "num_tokens": 685119065.0, "step": 18041 }, { "epoch": 2.2951278463299833, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6453680992126465, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8560267090797424, "num_tokens": 685154988.0, "step": 18042 }, { "epoch": 2.295255056608574, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.757239580154419, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8768860101699829, "num_tokens": 685194739.0, "step": 18043 }, { "epoch": 2.2953822668871644, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.615760087966919, "learning_rate": 1e-05, "loss": 0.4685, "mean_token_accuracy": 0.8676708340644836, "num_tokens": 685235619.0, "step": 18044 }, { "epoch": 2.295509477165755, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7512171268463135, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8759371638298035, "num_tokens": 685275189.0, "step": 18045 }, { "epoch": 2.2956366874443455, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.61439847946167, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8730092644691467, "num_tokens": 685319928.0, "step": 18046 }, { "epoch": 2.295763897722936, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7521657943725586, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8874586820602417, "num_tokens": 685355698.0, "step": 18047 }, { "epoch": 2.2958911080015265, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6180779933929443, "learning_rate": 1e-05, "loss": 0.3942, "mean_token_accuracy": 0.8825188875198364, "num_tokens": 685394873.0, "step": 18048 }, { "epoch": 2.296018318280117, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7566587924957275, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8848785758018494, "num_tokens": 685433896.0, "step": 18049 }, { "epoch": 2.2961455285587076, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6131067276000977, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.876641035079956, "num_tokens": 685476036.0, "step": 18050 }, { "epoch": 2.296272738837298, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.760958194732666, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.882803201675415, "num_tokens": 685514020.0, "step": 18051 }, { "epoch": 2.2963999491158886, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.618039131164551, "learning_rate": 1e-05, "loss": 0.3904, "mean_token_accuracy": 0.8891268372535706, "num_tokens": 685547644.0, "step": 18052 }, { "epoch": 2.296527159394479, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7737061977386475, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8739562630653381, "num_tokens": 685584147.0, "step": 18053 }, { "epoch": 2.2966543696730697, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6163113117218018, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8808095455169678, "num_tokens": 685623093.0, "step": 18054 }, { "epoch": 2.29678157995166, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.780043601989746, "learning_rate": 1e-05, "loss": 0.4098, "mean_token_accuracy": 0.8865329027175903, "num_tokens": 685657837.0, "step": 18055 }, { "epoch": 2.2969087902302507, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6264395713806152, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8787626028060913, "num_tokens": 685691077.0, "step": 18056 }, { "epoch": 2.2970360005088413, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.785507917404175, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8765403628349304, "num_tokens": 685725011.0, "step": 18057 }, { "epoch": 2.297163210787432, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6206371784210205, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.865463137626648, "num_tokens": 685764212.0, "step": 18058 }, { "epoch": 2.2972904210660223, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.79799747467041, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.860541582107544, "num_tokens": 685800780.0, "step": 18059 }, { "epoch": 2.2974176313446124, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.615849494934082, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8784345388412476, "num_tokens": 685837728.0, "step": 18060 }, { "epoch": 2.2975448416232034, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8236727714538574, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8628507852554321, "num_tokens": 685869349.0, "step": 18061 }, { "epoch": 2.2976720519017935, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.598634719848633, "learning_rate": 1e-05, "loss": 0.3919, "mean_token_accuracy": 0.8872897624969482, "num_tokens": 685908114.0, "step": 18062 }, { "epoch": 2.297799262180384, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8271281719207764, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8707681894302368, "num_tokens": 685944980.0, "step": 18063 }, { "epoch": 2.2979264724589745, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6045265197753906, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8848994970321655, "num_tokens": 685982729.0, "step": 18064 }, { "epoch": 2.298053682737565, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.825880289077759, "learning_rate": 1e-05, "loss": 0.37, "mean_token_accuracy": 0.8954015970230103, "num_tokens": 686014090.0, "step": 18065 }, { "epoch": 2.2981808930161556, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.616386890411377, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8850011825561523, "num_tokens": 686047345.0, "step": 18066 }, { "epoch": 2.298308103294746, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003299713134765625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.823417901992798, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8722778558731079, "num_tokens": 686088662.0, "step": 18067 }, { "epoch": 2.2984353135733366, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.628509521484375, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8700231313705444, "num_tokens": 686124210.0, "step": 18068 }, { "epoch": 2.298562523851927, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8205575942993164, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8800801038742065, "num_tokens": 686157759.0, "step": 18069 }, { "epoch": 2.2986897341305177, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.616546630859375, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8738269805908203, "num_tokens": 686198486.0, "step": 18070 }, { "epoch": 2.298816944409108, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.823852777481079, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8598049283027649, "num_tokens": 686235859.0, "step": 18071 }, { "epoch": 2.2989441546876987, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6164257526397705, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8852788805961609, "num_tokens": 686274185.0, "step": 18072 }, { "epoch": 2.2990713649662893, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.817131996154785, "learning_rate": 1e-05, "loss": 0.3768, "mean_token_accuracy": 0.8940889835357666, "num_tokens": 686308706.0, "step": 18073 }, { "epoch": 2.29919857524488, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6131629943847656, "learning_rate": 1e-05, "loss": 0.3876, "mean_token_accuracy": 0.8895511627197266, "num_tokens": 686347871.0, "step": 18074 }, { "epoch": 2.2993257855234703, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.806401014328003, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8687734007835388, "num_tokens": 686395284.0, "step": 18075 }, { "epoch": 2.299452995802061, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.624506950378418, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8718669414520264, "num_tokens": 686436566.0, "step": 18076 }, { "epoch": 2.2995802060806514, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8031084537506104, "learning_rate": 1e-05, "loss": 0.398, "mean_token_accuracy": 0.8868379592895508, "num_tokens": 686477332.0, "step": 18077 }, { "epoch": 2.299707416359242, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.738351345062256, "learning_rate": 1e-05, "loss": 0.3834, "mean_token_accuracy": 0.893463134765625, "num_tokens": 686512000.0, "step": 18078 }, { "epoch": 2.2998346266378324, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7528743743896484, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8821561336517334, "num_tokens": 686544732.0, "step": 18079 }, { "epoch": 2.299961836916423, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6605241298675537, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8792761564254761, "num_tokens": 686583100.0, "step": 18080 }, { "epoch": 2.3000890471950135, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.725713014602661, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8778434991836548, "num_tokens": 686620824.0, "step": 18081 }, { "epoch": 2.300216257473604, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.653749465942383, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8862448930740356, "num_tokens": 686660410.0, "step": 18082 }, { "epoch": 2.3003434677521946, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7275454998016357, "learning_rate": 1e-05, "loss": 0.3809, "mean_token_accuracy": 0.8905971050262451, "num_tokens": 686694385.0, "step": 18083 }, { "epoch": 2.300470678030785, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6435301303863525, "learning_rate": 1e-05, "loss": 0.4017, "mean_token_accuracy": 0.8834962248802185, "num_tokens": 686736844.0, "step": 18084 }, { "epoch": 2.300597888309375, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7215323448181152, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.874264121055603, "num_tokens": 686781104.0, "step": 18085 }, { "epoch": 2.300725098587966, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.65014910697937, "learning_rate": 1e-05, "loss": 0.3783, "mean_token_accuracy": 0.8938963413238525, "num_tokens": 686815244.0, "step": 18086 }, { "epoch": 2.3008523088665562, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.720196485519409, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8791121244430542, "num_tokens": 686858133.0, "step": 18087 }, { "epoch": 2.3009795191451468, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.634005546569824, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8750623464584351, "num_tokens": 686901250.0, "step": 18088 }, { "epoch": 2.3011067294237373, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7291600704193115, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8875088691711426, "num_tokens": 686936408.0, "step": 18089 }, { "epoch": 2.301233939702328, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.63382625579834, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8801161050796509, "num_tokens": 686981924.0, "step": 18090 }, { "epoch": 2.3013611499809183, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7484610080718994, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.875669002532959, "num_tokens": 687016052.0, "step": 18091 }, { "epoch": 2.301488360259509, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.607050895690918, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8862681984901428, "num_tokens": 687059107.0, "step": 18092 }, { "epoch": 2.3016155705380994, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.736456871032715, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8844332695007324, "num_tokens": 687104015.0, "step": 18093 }, { "epoch": 2.30174278081669, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.626312732696533, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8749433755874634, "num_tokens": 687143719.0, "step": 18094 }, { "epoch": 2.3018699910952805, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.759852170944214, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8816298842430115, "num_tokens": 687179732.0, "step": 18095 }, { "epoch": 2.301997201373871, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.607844829559326, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8751853108406067, "num_tokens": 687222309.0, "step": 18096 }, { "epoch": 2.3021244116524615, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7624895572662354, "learning_rate": 1e-05, "loss": 0.3925, "mean_token_accuracy": 0.8887330889701843, "num_tokens": 687259114.0, "step": 18097 }, { "epoch": 2.302251621931052, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6149964332580566, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8850804567337036, "num_tokens": 687293052.0, "step": 18098 }, { "epoch": 2.3023788322096426, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.753070116043091, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.8867658376693726, "num_tokens": 687333245.0, "step": 18099 }, { "epoch": 2.302506042488233, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6216440200805664, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8772480487823486, "num_tokens": 687371740.0, "step": 18100 }, { "epoch": 2.3026332527668236, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.777036666870117, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8581011891365051, "num_tokens": 687407298.0, "step": 18101 }, { "epoch": 2.302760463045414, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6041531562805176, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8873792886734009, "num_tokens": 687447425.0, "step": 18102 }, { "epoch": 2.3028876733240047, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.768265962600708, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8809880614280701, "num_tokens": 687488528.0, "step": 18103 }, { "epoch": 2.303014883602595, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.625847339630127, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8791283369064331, "num_tokens": 687521086.0, "step": 18104 }, { "epoch": 2.3031420938811857, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.77655029296875, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8737475872039795, "num_tokens": 687558882.0, "step": 18105 }, { "epoch": 2.3032693041597763, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.615241527557373, "learning_rate": 1e-05, "loss": 0.3771, "mean_token_accuracy": 0.8917901515960693, "num_tokens": 687592058.0, "step": 18106 }, { "epoch": 2.303396514438367, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7725026607513428, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8783434629440308, "num_tokens": 687629715.0, "step": 18107 }, { "epoch": 2.3035237247169573, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.623436450958252, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8779424428939819, "num_tokens": 687666201.0, "step": 18108 }, { "epoch": 2.303650934995548, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7732605934143066, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.884624183177948, "num_tokens": 687706897.0, "step": 18109 }, { "epoch": 2.303778145274138, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.632143020629883, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8813028335571289, "num_tokens": 687741449.0, "step": 18110 }, { "epoch": 2.303905355552729, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7635505199432373, "learning_rate": 1e-05, "loss": 0.391, "mean_token_accuracy": 0.891625165939331, "num_tokens": 687776758.0, "step": 18111 }, { "epoch": 2.304032565831319, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6368014812469482, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8659330010414124, "num_tokens": 687814636.0, "step": 18112 }, { "epoch": 2.3041597761099095, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7818374633789062, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8617979288101196, "num_tokens": 687843685.0, "step": 18113 }, { "epoch": 2.3042869863885, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6401116847991943, "learning_rate": 1e-05, "loss": 0.3927, "mean_token_accuracy": 0.8882344365119934, "num_tokens": 687875040.0, "step": 18114 }, { "epoch": 2.3044141966670906, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.760427474975586, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8751139640808105, "num_tokens": 687912952.0, "step": 18115 }, { "epoch": 2.304541406945681, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6253280639648438, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8817787170410156, "num_tokens": 687951366.0, "step": 18116 }, { "epoch": 2.3046686172242716, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.750248432159424, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8797950744628906, "num_tokens": 687993593.0, "step": 18117 }, { "epoch": 2.304795827502862, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6324543952941895, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8770905137062073, "num_tokens": 688033449.0, "step": 18118 }, { "epoch": 2.3049230377814527, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7715656757354736, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8779163360595703, "num_tokens": 688073239.0, "step": 18119 }, { "epoch": 2.305050248060043, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6505110263824463, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8754940032958984, "num_tokens": 688112760.0, "step": 18120 }, { "epoch": 2.3051774583386337, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7614312171936035, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8710025548934937, "num_tokens": 688150553.0, "step": 18121 }, { "epoch": 2.3053046686172243, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6280641555786133, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8750060200691223, "num_tokens": 688190582.0, "step": 18122 }, { "epoch": 2.305431878895815, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7769651412963867, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8788020014762878, "num_tokens": 688226512.0, "step": 18123 }, { "epoch": 2.3055590891744053, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.896566867828369, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.880779504776001, "num_tokens": 688264891.0, "step": 18124 }, { "epoch": 2.305686299452996, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6611080169677734, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8787791132926941, "num_tokens": 688301280.0, "step": 18125 }, { "epoch": 2.3058135097315864, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.728817939758301, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.881524384021759, "num_tokens": 688337178.0, "step": 18126 }, { "epoch": 2.305940720010177, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.616701602935791, "learning_rate": 1e-05, "loss": 0.3934, "mean_token_accuracy": 0.8874726295471191, "num_tokens": 688375839.0, "step": 18127 }, { "epoch": 2.3060679302887674, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.744701862335205, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.877409815788269, "num_tokens": 688416395.0, "step": 18128 }, { "epoch": 2.306195140567358, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6199450492858887, "learning_rate": 1e-05, "loss": 0.3934, "mean_token_accuracy": 0.8898305892944336, "num_tokens": 688454801.0, "step": 18129 }, { "epoch": 2.3063223508459485, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.764061450958252, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.87962806224823, "num_tokens": 688489265.0, "step": 18130 }, { "epoch": 2.306449561124539, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.623802423477173, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8746492266654968, "num_tokens": 688525392.0, "step": 18131 }, { "epoch": 2.3065767714031296, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.775818109512329, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8632513880729675, "num_tokens": 688562389.0, "step": 18132 }, { "epoch": 2.3067039816817196, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6196515560150146, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8813093304634094, "num_tokens": 688597728.0, "step": 18133 }, { "epoch": 2.3068311919603106, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7804410457611084, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8782769441604614, "num_tokens": 688633065.0, "step": 18134 }, { "epoch": 2.3069584022389007, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6239778995513916, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.880341112613678, "num_tokens": 688677095.0, "step": 18135 }, { "epoch": 2.3070856125174912, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7794060707092285, "learning_rate": 1e-05, "loss": 0.3983, "mean_token_accuracy": 0.8841910362243652, "num_tokens": 688710421.0, "step": 18136 }, { "epoch": 2.3072128227960818, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6136891841888428, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8799622058868408, "num_tokens": 688747300.0, "step": 18137 }, { "epoch": 2.3073400330746723, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7873587608337402, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.8879757523536682, "num_tokens": 688781867.0, "step": 18138 }, { "epoch": 2.307467243353263, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.606072425842285, "learning_rate": 1e-05, "loss": 0.4015, "mean_token_accuracy": 0.8818009495735168, "num_tokens": 688822777.0, "step": 18139 }, { "epoch": 2.3075944536318533, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7938199043273926, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8711795806884766, "num_tokens": 688862295.0, "step": 18140 }, { "epoch": 2.307721663910444, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6025307178497314, "learning_rate": 1e-05, "loss": 0.3731, "mean_token_accuracy": 0.8947780132293701, "num_tokens": 688904814.0, "step": 18141 }, { "epoch": 2.3078488741890344, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.808105707168579, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8704483509063721, "num_tokens": 688940403.0, "step": 18142 }, { "epoch": 2.307976084467625, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6231186389923096, "learning_rate": 1e-05, "loss": 0.4728, "mean_token_accuracy": 0.8621963858604431, "num_tokens": 688980256.0, "step": 18143 }, { "epoch": 2.3081032947462155, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.844625234603882, "learning_rate": 1e-05, "loss": 0.4793, "mean_token_accuracy": 0.8621617555618286, "num_tokens": 689020629.0, "step": 18144 }, { "epoch": 2.308230505024806, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6399474143981934, "learning_rate": 1e-05, "loss": 0.4642, "mean_token_accuracy": 0.8647670149803162, "num_tokens": 689051257.0, "step": 18145 }, { "epoch": 2.3083577153033965, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.816131114959717, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8822418451309204, "num_tokens": 689083723.0, "step": 18146 }, { "epoch": 2.308484925581987, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.614246129989624, "learning_rate": 1e-05, "loss": 0.474, "mean_token_accuracy": 0.8618755340576172, "num_tokens": 689123785.0, "step": 18147 }, { "epoch": 2.3086121358605776, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.806067705154419, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8765602707862854, "num_tokens": 689164053.0, "step": 18148 }, { "epoch": 2.308739346139168, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6257903575897217, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8827674388885498, "num_tokens": 689198221.0, "step": 18149 }, { "epoch": 2.3088665564177586, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8130085468292236, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8769211173057556, "num_tokens": 689232330.0, "step": 18150 }, { "epoch": 2.308993766696349, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6263411045074463, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8709877729415894, "num_tokens": 689277663.0, "step": 18151 }, { "epoch": 2.3091209769749397, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.815354585647583, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8873395919799805, "num_tokens": 689308563.0, "step": 18152 }, { "epoch": 2.30924818725353, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6244723796844482, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.878484308719635, "num_tokens": 689346897.0, "step": 18153 }, { "epoch": 2.3093753975321207, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.817953109741211, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8686724305152893, "num_tokens": 689380689.0, "step": 18154 }, { "epoch": 2.3095026078107113, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6147146224975586, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8806170225143433, "num_tokens": 689418587.0, "step": 18155 }, { "epoch": 2.309629818089302, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8177359104156494, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8831096291542053, "num_tokens": 689455742.0, "step": 18156 }, { "epoch": 2.3097570283678923, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6170589923858643, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8790814876556396, "num_tokens": 689495618.0, "step": 18157 }, { "epoch": 2.3098842386464824, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8392064571380615, "learning_rate": 1e-05, "loss": 0.4731, "mean_token_accuracy": 0.8614703416824341, "num_tokens": 689529856.0, "step": 18158 }, { "epoch": 2.3100114489250734, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6310067176818848, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8756128549575806, "num_tokens": 689566785.0, "step": 18159 }, { "epoch": 2.3101386592036635, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8194830417633057, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8750537633895874, "num_tokens": 689604332.0, "step": 18160 }, { "epoch": 2.310265869482254, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6289007663726807, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8870185613632202, "num_tokens": 689639159.0, "step": 18161 }, { "epoch": 2.3103930797608445, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.812112808227539, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.871711015701294, "num_tokens": 689681660.0, "step": 18162 }, { "epoch": 2.310520290039435, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.630981922149658, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8884885311126709, "num_tokens": 689717070.0, "step": 18163 }, { "epoch": 2.3106475003180256, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8036608695983887, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8818417191505432, "num_tokens": 689758885.0, "step": 18164 }, { "epoch": 2.310774710596616, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.634115219116211, "learning_rate": 1e-05, "loss": 0.3681, "mean_token_accuracy": 0.8951780796051025, "num_tokens": 689794943.0, "step": 18165 }, { "epoch": 2.3109019208752066, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.78705096244812, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8816956281661987, "num_tokens": 689836943.0, "step": 18166 }, { "epoch": 2.311029131153797, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6432700157165527, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8827589750289917, "num_tokens": 689870787.0, "step": 18167 }, { "epoch": 2.3111563414323877, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.781590223312378, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8768826723098755, "num_tokens": 689908302.0, "step": 18168 }, { "epoch": 2.311283551710978, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6374566555023193, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8723602294921875, "num_tokens": 689950031.0, "step": 18169 }, { "epoch": 2.3114107619895687, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7632224559783936, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8837560415267944, "num_tokens": 689994674.0, "step": 18170 }, { "epoch": 2.3115379722681593, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.626025676727295, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.885582447052002, "num_tokens": 690038697.0, "step": 18171 }, { "epoch": 2.31166518254675, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.778374195098877, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.8742464780807495, "num_tokens": 690077449.0, "step": 18172 }, { "epoch": 2.3117923928253403, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6382482051849365, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8744171857833862, "num_tokens": 690113659.0, "step": 18173 }, { "epoch": 2.311919603103931, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7674849033355713, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8746537566184998, "num_tokens": 690153928.0, "step": 18174 }, { "epoch": 2.3120468133825214, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.629319667816162, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8813734650611877, "num_tokens": 690189526.0, "step": 18175 }, { "epoch": 2.312174023661112, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.772951126098633, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8755460977554321, "num_tokens": 690230220.0, "step": 18176 }, { "epoch": 2.3123012339397024, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.658522129058838, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8770965337753296, "num_tokens": 690264998.0, "step": 18177 }, { "epoch": 2.312428444218293, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7367751598358154, "learning_rate": 1e-05, "loss": 0.3521, "mean_token_accuracy": 0.9053541421890259, "num_tokens": 690304323.0, "step": 18178 }, { "epoch": 2.3125556544968835, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6369102001190186, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8795300722122192, "num_tokens": 690346832.0, "step": 18179 }, { "epoch": 2.312682864775474, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7380623817443848, "learning_rate": 1e-05, "loss": 0.4033, "mean_token_accuracy": 0.8832662105560303, "num_tokens": 690385829.0, "step": 18180 }, { "epoch": 2.3128100750540646, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6335294246673584, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8821089267730713, "num_tokens": 690429949.0, "step": 18181 }, { "epoch": 2.312937285332655, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7281570434570312, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8838649988174438, "num_tokens": 690473252.0, "step": 18182 }, { "epoch": 2.313064495611245, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6316068172454834, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8865875005722046, "num_tokens": 690508765.0, "step": 18183 }, { "epoch": 2.313191705889836, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7530953884124756, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8801476955413818, "num_tokens": 690546749.0, "step": 18184 }, { "epoch": 2.313318916168426, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.642636775970459, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8791650533676147, "num_tokens": 690579808.0, "step": 18185 }, { "epoch": 2.3134461264470167, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7578554153442383, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8787013292312622, "num_tokens": 690611667.0, "step": 18186 }, { "epoch": 2.3135733367256073, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.642883062362671, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8709934949874878, "num_tokens": 690654991.0, "step": 18187 }, { "epoch": 2.313700547004198, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7500531673431396, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8661623001098633, "num_tokens": 690692344.0, "step": 18188 }, { "epoch": 2.3138277572827883, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.622440814971924, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.8881840109825134, "num_tokens": 690726812.0, "step": 18189 }, { "epoch": 2.313954967561379, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.754082679748535, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8758492469787598, "num_tokens": 690768229.0, "step": 18190 }, { "epoch": 2.3140821778399694, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6153404712677, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8819344639778137, "num_tokens": 690810614.0, "step": 18191 }, { "epoch": 2.31420938811856, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.759223699569702, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8673340082168579, "num_tokens": 690855422.0, "step": 18192 }, { "epoch": 2.3143365983971504, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.611356019973755, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8869337439537048, "num_tokens": 690893751.0, "step": 18193 }, { "epoch": 2.314463808675741, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.753441572189331, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8805041909217834, "num_tokens": 690936734.0, "step": 18194 }, { "epoch": 2.3145910189543315, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6152007579803467, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.885310173034668, "num_tokens": 690975005.0, "step": 18195 }, { "epoch": 2.314718229232922, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.771238088607788, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8828598260879517, "num_tokens": 691011607.0, "step": 18196 }, { "epoch": 2.3148454395115126, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6190786361694336, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8845455646514893, "num_tokens": 691047635.0, "step": 18197 }, { "epoch": 2.314972649790103, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.758390188217163, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8746770620346069, "num_tokens": 691095290.0, "step": 18198 }, { "epoch": 2.3150998600686936, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.605992555618286, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8796027898788452, "num_tokens": 691140641.0, "step": 18199 }, { "epoch": 2.315227070347284, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.778336524963379, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8796961307525635, "num_tokens": 691181744.0, "step": 18200 }, { "epoch": 2.3153542806258747, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.609123945236206, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8803365230560303, "num_tokens": 691221279.0, "step": 18201 }, { "epoch": 2.315481490904465, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.783308982849121, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8849102258682251, "num_tokens": 691258519.0, "step": 18202 }, { "epoch": 2.3156087011830557, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6143312454223633, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8764509558677673, "num_tokens": 691297063.0, "step": 18203 }, { "epoch": 2.3157359114616463, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7858388423919678, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8806059956550598, "num_tokens": 691337146.0, "step": 18204 }, { "epoch": 2.315863121740237, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6183853149414062, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8867788314819336, "num_tokens": 691373475.0, "step": 18205 }, { "epoch": 2.3159903320188273, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.778127908706665, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8767547011375427, "num_tokens": 691413286.0, "step": 18206 }, { "epoch": 2.316117542297418, "ewc_loss": 0.07421875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6166956424713135, "learning_rate": 1e-05, "loss": 0.3769, "mean_token_accuracy": 0.8927756547927856, "num_tokens": 691444950.0, "step": 18207 }, { "epoch": 2.316244752576008, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7811901569366455, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8773420453071594, "num_tokens": 691486767.0, "step": 18208 }, { "epoch": 2.316371962854599, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6095306873321533, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8760418891906738, "num_tokens": 691527149.0, "step": 18209 }, { "epoch": 2.316499173133189, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.795287609100342, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8715581893920898, "num_tokens": 691563963.0, "step": 18210 }, { "epoch": 2.3166263834117795, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6234822273254395, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8730466961860657, "num_tokens": 691596423.0, "step": 18211 }, { "epoch": 2.31675359369037, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7796132564544678, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8770935535430908, "num_tokens": 691636928.0, "step": 18212 }, { "epoch": 2.3168808039689606, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6342790126800537, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8781659007072449, "num_tokens": 691673427.0, "step": 18213 }, { "epoch": 2.317008014247551, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.776151180267334, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8752479553222656, "num_tokens": 691710427.0, "step": 18214 }, { "epoch": 2.3171352245261416, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6363842487335205, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8706916570663452, "num_tokens": 691746809.0, "step": 18215 }, { "epoch": 2.317262434804732, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.761486768722534, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8846874237060547, "num_tokens": 691786037.0, "step": 18216 }, { "epoch": 2.3173896450833227, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6295559406280518, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8758062124252319, "num_tokens": 691822270.0, "step": 18217 }, { "epoch": 2.317516855361913, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7475948333740234, "learning_rate": 1e-05, "loss": 0.3624, "mean_token_accuracy": 0.899324357509613, "num_tokens": 691858538.0, "step": 18218 }, { "epoch": 2.3176440656405037, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6285839080810547, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8841104507446289, "num_tokens": 691898810.0, "step": 18219 }, { "epoch": 2.3177712759190943, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7588531970977783, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.881431519985199, "num_tokens": 691933708.0, "step": 18220 }, { "epoch": 2.317898486197685, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6300034523010254, "learning_rate": 1e-05, "loss": 0.3864, "mean_token_accuracy": 0.8925849199295044, "num_tokens": 691968293.0, "step": 18221 }, { "epoch": 2.3180256964762753, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7548117637634277, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8694515824317932, "num_tokens": 692004532.0, "step": 18222 }, { "epoch": 2.318152906754866, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.630152463912964, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8815635442733765, "num_tokens": 692042323.0, "step": 18223 }, { "epoch": 2.3182801170334564, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.750232458114624, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8780510425567627, "num_tokens": 692080970.0, "step": 18224 }, { "epoch": 2.318407327312047, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.638159990310669, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8855180740356445, "num_tokens": 692115027.0, "step": 18225 }, { "epoch": 2.3185345375906374, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000331878662109375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7704708576202393, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8777858018875122, "num_tokens": 692146602.0, "step": 18226 }, { "epoch": 2.318661747869228, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6229610443115234, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.864063024520874, "num_tokens": 692189721.0, "step": 18227 }, { "epoch": 2.3187889581478185, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.755274534225464, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8712953329086304, "num_tokens": 692233032.0, "step": 18228 }, { "epoch": 2.318916168426409, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6400327682495117, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8711372017860413, "num_tokens": 692269033.0, "step": 18229 }, { "epoch": 2.3190433787049995, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.752840757369995, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8770637512207031, "num_tokens": 692307415.0, "step": 18230 }, { "epoch": 2.3191705889835896, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6238701343536377, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8862869739532471, "num_tokens": 692347003.0, "step": 18231 }, { "epoch": 2.3192977992621806, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.758078098297119, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8790782690048218, "num_tokens": 692385141.0, "step": 18232 }, { "epoch": 2.3194250095407707, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6179280281066895, "learning_rate": 1e-05, "loss": 0.3721, "mean_token_accuracy": 0.8927493095397949, "num_tokens": 692421731.0, "step": 18233 }, { "epoch": 2.319552219819361, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7595036029815674, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8783078789710999, "num_tokens": 692466400.0, "step": 18234 }, { "epoch": 2.3196794300979517, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.617570161819458, "learning_rate": 1e-05, "loss": 0.3669, "mean_token_accuracy": 0.8955551981925964, "num_tokens": 692501201.0, "step": 18235 }, { "epoch": 2.3198066403765423, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7685546875, "learning_rate": 1e-05, "loss": 0.3851, "mean_token_accuracy": 0.8923237919807434, "num_tokens": 692541549.0, "step": 18236 }, { "epoch": 2.319933850655133, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6147234439849854, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8844538927078247, "num_tokens": 692582156.0, "step": 18237 }, { "epoch": 2.3200610609337233, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7772138118743896, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8669624328613281, "num_tokens": 692620522.0, "step": 18238 }, { "epoch": 2.320188271212314, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6214756965637207, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8739532232284546, "num_tokens": 692658372.0, "step": 18239 }, { "epoch": 2.3203154814909044, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.790724992752075, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.8822919130325317, "num_tokens": 692687338.0, "step": 18240 }, { "epoch": 2.320442691769495, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.613481283187866, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8766616582870483, "num_tokens": 692724417.0, "step": 18241 }, { "epoch": 2.3205699020480854, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.779144525527954, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8785718679428101, "num_tokens": 692767881.0, "step": 18242 }, { "epoch": 2.320697112326676, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6220734119415283, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8749099373817444, "num_tokens": 692803154.0, "step": 18243 }, { "epoch": 2.3208243226052665, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.78208589553833, "learning_rate": 1e-05, "loss": 0.3753, "mean_token_accuracy": 0.8959048986434937, "num_tokens": 692842435.0, "step": 18244 }, { "epoch": 2.320951532883857, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6076853275299072, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8720712661743164, "num_tokens": 692884491.0, "step": 18245 }, { "epoch": 2.3210787431624476, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.786501884460449, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.875091552734375, "num_tokens": 692920825.0, "step": 18246 }, { "epoch": 2.321205953441038, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6221206188201904, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8843222260475159, "num_tokens": 692956413.0, "step": 18247 }, { "epoch": 2.3213331637196286, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7943902015686035, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8715786933898926, "num_tokens": 692995860.0, "step": 18248 }, { "epoch": 2.321460373998219, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.621640682220459, "learning_rate": 1e-05, "loss": 0.3874, "mean_token_accuracy": 0.8973304033279419, "num_tokens": 693031077.0, "step": 18249 }, { "epoch": 2.3215875842768097, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8102643489837646, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8673315048217773, "num_tokens": 693067181.0, "step": 18250 }, { "epoch": 2.3217147945554, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6110129356384277, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8830804824829102, "num_tokens": 693103919.0, "step": 18251 }, { "epoch": 2.3218420048339907, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8220434188842773, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8642503023147583, "num_tokens": 693138471.0, "step": 18252 }, { "epoch": 2.3219692151125813, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.630721092224121, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8709865212440491, "num_tokens": 693174220.0, "step": 18253 }, { "epoch": 2.322096425391172, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7895970344543457, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8896004557609558, "num_tokens": 693217222.0, "step": 18254 }, { "epoch": 2.3222236356697623, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.636549949645996, "learning_rate": 1e-05, "loss": 0.3818, "mean_token_accuracy": 0.8942149877548218, "num_tokens": 693260052.0, "step": 18255 }, { "epoch": 2.3223508459483524, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.792849540710449, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8682920932769775, "num_tokens": 693293775.0, "step": 18256 }, { "epoch": 2.3224780562269434, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.621408462524414, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8852401971817017, "num_tokens": 693335876.0, "step": 18257 }, { "epoch": 2.3226052665055335, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.778921365737915, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8700418472290039, "num_tokens": 693376886.0, "step": 18258 }, { "epoch": 2.322732476784124, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6244640350341797, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8700166940689087, "num_tokens": 693416386.0, "step": 18259 }, { "epoch": 2.3228596870627145, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7704787254333496, "learning_rate": 1e-05, "loss": 0.3872, "mean_token_accuracy": 0.8908843994140625, "num_tokens": 693454926.0, "step": 18260 }, { "epoch": 2.322986897341305, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.634291172027588, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8773853778839111, "num_tokens": 693492926.0, "step": 18261 }, { "epoch": 2.3231141076198956, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.752856731414795, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8815447092056274, "num_tokens": 693534684.0, "step": 18262 }, { "epoch": 2.323241317898486, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.644240617752075, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8672625422477722, "num_tokens": 693570856.0, "step": 18263 }, { "epoch": 2.3233685281770766, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.749896287918091, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8692382574081421, "num_tokens": 693613286.0, "step": 18264 }, { "epoch": 2.323495738455667, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.631126880645752, "learning_rate": 1e-05, "loss": 0.3899, "mean_token_accuracy": 0.8869789838790894, "num_tokens": 693648748.0, "step": 18265 }, { "epoch": 2.3236229487342577, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7447352409362793, "learning_rate": 1e-05, "loss": 0.3768, "mean_token_accuracy": 0.8952165246009827, "num_tokens": 693688997.0, "step": 18266 }, { "epoch": 2.323750159012848, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.634441375732422, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8763239979743958, "num_tokens": 693727653.0, "step": 18267 }, { "epoch": 2.3238773692914387, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.777237892150879, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8777016401290894, "num_tokens": 693771146.0, "step": 18268 }, { "epoch": 2.3240045795700293, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6426682472229004, "learning_rate": 1e-05, "loss": 0.4993, "mean_token_accuracy": 0.855004072189331, "num_tokens": 693809135.0, "step": 18269 }, { "epoch": 2.32413178984862, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7439825534820557, "learning_rate": 1e-05, "loss": 0.3948, "mean_token_accuracy": 0.8879970908164978, "num_tokens": 693847127.0, "step": 18270 }, { "epoch": 2.3242590001272103, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.624619483947754, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8824511766433716, "num_tokens": 693887186.0, "step": 18271 }, { "epoch": 2.324386210405801, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7490272521972656, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8878358602523804, "num_tokens": 693925943.0, "step": 18272 }, { "epoch": 2.3245134206843914, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6538443565368652, "learning_rate": 1e-05, "loss": 0.3656, "mean_token_accuracy": 0.8948291540145874, "num_tokens": 693958825.0, "step": 18273 }, { "epoch": 2.324640630962982, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7287375926971436, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8672993183135986, "num_tokens": 694003787.0, "step": 18274 }, { "epoch": 2.3247678412415724, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.637960910797119, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.879494845867157, "num_tokens": 694047253.0, "step": 18275 }, { "epoch": 2.324895051520163, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.733682870864868, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8743464350700378, "num_tokens": 694085913.0, "step": 18276 }, { "epoch": 2.3250222617987535, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6579604148864746, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8653711080551147, "num_tokens": 694122848.0, "step": 18277 }, { "epoch": 2.325149472077344, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000335693359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.726424217224121, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8746292591094971, "num_tokens": 694160545.0, "step": 18278 }, { "epoch": 2.3252766823559345, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6514060497283936, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8807120323181152, "num_tokens": 694194272.0, "step": 18279 }, { "epoch": 2.325403892634525, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.727329730987549, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8823143243789673, "num_tokens": 694227164.0, "step": 18280 }, { "epoch": 2.325531102913115, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6317970752716064, "learning_rate": 1e-05, "loss": 0.3805, "mean_token_accuracy": 0.8925125002861023, "num_tokens": 694266589.0, "step": 18281 }, { "epoch": 2.325658313191706, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7390100955963135, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8714860677719116, "num_tokens": 694301558.0, "step": 18282 }, { "epoch": 2.325785523470296, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6312918663024902, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8672710657119751, "num_tokens": 694343332.0, "step": 18283 }, { "epoch": 2.3259127337488867, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.731285810470581, "learning_rate": 1e-05, "loss": 0.3833, "mean_token_accuracy": 0.8908299803733826, "num_tokens": 694383602.0, "step": 18284 }, { "epoch": 2.3260399440274773, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6531760692596436, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8696771860122681, "num_tokens": 694417664.0, "step": 18285 }, { "epoch": 2.326167154306068, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7288501262664795, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8637083172798157, "num_tokens": 694456176.0, "step": 18286 }, { "epoch": 2.3262943645846583, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6411378383636475, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8745898604393005, "num_tokens": 694497114.0, "step": 18287 }, { "epoch": 2.326421574863249, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7047653198242188, "learning_rate": 1e-05, "loss": 0.4014, "mean_token_accuracy": 0.8865007162094116, "num_tokens": 694540832.0, "step": 18288 }, { "epoch": 2.3265487851418394, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6457479000091553, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8854538202285767, "num_tokens": 694578673.0, "step": 18289 }, { "epoch": 2.32667599542043, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7226340770721436, "learning_rate": 1e-05, "loss": 0.4981, "mean_token_accuracy": 0.8618992567062378, "num_tokens": 694624736.0, "step": 18290 }, { "epoch": 2.3268032056990204, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6462132930755615, "learning_rate": 1e-05, "loss": 0.3689, "mean_token_accuracy": 0.8954473733901978, "num_tokens": 694659123.0, "step": 18291 }, { "epoch": 2.326930415977611, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.712977409362793, "learning_rate": 1e-05, "loss": 0.3926, "mean_token_accuracy": 0.8899893164634705, "num_tokens": 694695310.0, "step": 18292 }, { "epoch": 2.3270576262562015, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.651132345199585, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8737406730651855, "num_tokens": 694732470.0, "step": 18293 }, { "epoch": 2.327184836534792, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.728938341140747, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8764129281044006, "num_tokens": 694763693.0, "step": 18294 }, { "epoch": 2.3273120468133826, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.656695604324341, "learning_rate": 1e-05, "loss": 0.3916, "mean_token_accuracy": 0.889154314994812, "num_tokens": 694798127.0, "step": 18295 }, { "epoch": 2.327439257091973, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.717434883117676, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8820419907569885, "num_tokens": 694833462.0, "step": 18296 }, { "epoch": 2.3275664673705636, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.9847655296325684, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8756088018417358, "num_tokens": 694870257.0, "step": 18297 }, { "epoch": 2.327693677649154, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.602738618850708, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8720005750656128, "num_tokens": 694910455.0, "step": 18298 }, { "epoch": 2.3278208879277447, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8063805103302, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.882053017616272, "num_tokens": 694947040.0, "step": 18299 }, { "epoch": 2.327948098206335, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.5720715522766113, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.874805212020874, "num_tokens": 694989043.0, "step": 18300 }, { "epoch": 2.3280753084849257, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.83768892288208, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8934696912765503, "num_tokens": 695027094.0, "step": 18301 }, { "epoch": 2.3282025187635162, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.618959665298462, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8801578879356384, "num_tokens": 695056419.0, "step": 18302 }, { "epoch": 2.3283297290421068, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.8296356201171875, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8757157325744629, "num_tokens": 695101430.0, "step": 18303 }, { "epoch": 2.3284569393206973, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6106936931610107, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8667146563529968, "num_tokens": 695141446.0, "step": 18304 }, { "epoch": 2.328584149599288, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8390629291534424, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8799054622650146, "num_tokens": 695179761.0, "step": 18305 }, { "epoch": 2.328711359877878, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.600127935409546, "learning_rate": 1e-05, "loss": 0.3694, "mean_token_accuracy": 0.8983227610588074, "num_tokens": 695217224.0, "step": 18306 }, { "epoch": 2.328838570156469, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8430025577545166, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8770995140075684, "num_tokens": 695253481.0, "step": 18307 }, { "epoch": 2.328965780435059, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6364119052886963, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8801887631416321, "num_tokens": 695287002.0, "step": 18308 }, { "epoch": 2.3290929907136495, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.813459873199463, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8856991529464722, "num_tokens": 695328296.0, "step": 18309 }, { "epoch": 2.32922020099224, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6337640285491943, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.881209671497345, "num_tokens": 695364291.0, "step": 18310 }, { "epoch": 2.3293474112708306, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.818021297454834, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8615403175354004, "num_tokens": 695407263.0, "step": 18311 }, { "epoch": 2.329474621549421, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.648827314376831, "learning_rate": 1e-05, "loss": 0.4707, "mean_token_accuracy": 0.8627943992614746, "num_tokens": 695444243.0, "step": 18312 }, { "epoch": 2.3296018318280116, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8108348846435547, "learning_rate": 1e-05, "loss": 0.3873, "mean_token_accuracy": 0.8887588977813721, "num_tokens": 695475207.0, "step": 18313 }, { "epoch": 2.329729042106602, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.630844831466675, "learning_rate": 1e-05, "loss": 0.3983, "mean_token_accuracy": 0.8870707154273987, "num_tokens": 695515915.0, "step": 18314 }, { "epoch": 2.3298562523851927, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8026108741760254, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8810416460037231, "num_tokens": 695555168.0, "step": 18315 }, { "epoch": 2.329983462663783, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6598808765411377, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8766430020332336, "num_tokens": 695591829.0, "step": 18316 }, { "epoch": 2.3301106729423737, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.78527569770813, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8752795457839966, "num_tokens": 695628518.0, "step": 18317 }, { "epoch": 2.3302378832209643, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.65242600440979, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8825226426124573, "num_tokens": 695661769.0, "step": 18318 }, { "epoch": 2.330365093499555, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7785234451293945, "learning_rate": 1e-05, "loss": 0.4788, "mean_token_accuracy": 0.8604585528373718, "num_tokens": 695699830.0, "step": 18319 }, { "epoch": 2.3304923037781453, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6379544734954834, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8845018148422241, "num_tokens": 695738871.0, "step": 18320 }, { "epoch": 2.330619514056736, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7818708419799805, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8709967136383057, "num_tokens": 695771868.0, "step": 18321 }, { "epoch": 2.3307467243353264, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6378183364868164, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8704890012741089, "num_tokens": 695808145.0, "step": 18322 }, { "epoch": 2.330873934613917, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7845730781555176, "learning_rate": 1e-05, "loss": 0.3815, "mean_token_accuracy": 0.8973826169967651, "num_tokens": 695837611.0, "step": 18323 }, { "epoch": 2.3310011448925074, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6324758529663086, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.8917759656906128, "num_tokens": 695874640.0, "step": 18324 }, { "epoch": 2.331128355171098, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.773505210876465, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8693331480026245, "num_tokens": 695912580.0, "step": 18325 }, { "epoch": 2.3312555654496885, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6474249362945557, "learning_rate": 1e-05, "loss": 0.3925, "mean_token_accuracy": 0.8881444334983826, "num_tokens": 695952061.0, "step": 18326 }, { "epoch": 2.331382775728279, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.761634349822998, "learning_rate": 1e-05, "loss": 0.3869, "mean_token_accuracy": 0.8901437520980835, "num_tokens": 695986389.0, "step": 18327 }, { "epoch": 2.3315099860068695, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6260082721710205, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8856933116912842, "num_tokens": 696026153.0, "step": 18328 }, { "epoch": 2.3316371962854596, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7697389125823975, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.864998459815979, "num_tokens": 696062693.0, "step": 18329 }, { "epoch": 2.3317644065640506, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.616731643676758, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8844161629676819, "num_tokens": 696101933.0, "step": 18330 }, { "epoch": 2.3318916168426407, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.77729868888855, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8740575313568115, "num_tokens": 696137607.0, "step": 18331 }, { "epoch": 2.332018827121231, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6085803508758545, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8790435194969177, "num_tokens": 696181277.0, "step": 18332 }, { "epoch": 2.3321460373998217, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.792553424835205, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8787209987640381, "num_tokens": 696216592.0, "step": 18333 }, { "epoch": 2.3322732476784123, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6160871982574463, "learning_rate": 1e-05, "loss": 0.3694, "mean_token_accuracy": 0.8960142135620117, "num_tokens": 696250762.0, "step": 18334 }, { "epoch": 2.332400457957003, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.78471302986145, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.877357006072998, "num_tokens": 696289819.0, "step": 18335 }, { "epoch": 2.3325276682355933, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6160337924957275, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8526851534843445, "num_tokens": 696335551.0, "step": 18336 }, { "epoch": 2.332654878514184, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.798074722290039, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8832663893699646, "num_tokens": 696369973.0, "step": 18337 }, { "epoch": 2.3327820887927744, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6329915523529053, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8738102912902832, "num_tokens": 696403718.0, "step": 18338 }, { "epoch": 2.332909299071365, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7847611904144287, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8761011362075806, "num_tokens": 696440769.0, "step": 18339 }, { "epoch": 2.3330365093499554, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.873640537261963, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8790018558502197, "num_tokens": 696482570.0, "step": 18340 }, { "epoch": 2.333163719628546, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6660397052764893, "learning_rate": 1e-05, "loss": 0.3821, "mean_token_accuracy": 0.8920027017593384, "num_tokens": 696519199.0, "step": 18341 }, { "epoch": 2.3332909299071365, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7276995182037354, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8643957376480103, "num_tokens": 696558460.0, "step": 18342 }, { "epoch": 2.333418140185727, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003376007080078125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.647958278656006, "learning_rate": 1e-05, "loss": 0.5035, "mean_token_accuracy": 0.853948712348938, "num_tokens": 696600395.0, "step": 18343 }, { "epoch": 2.3335453504643175, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7443480491638184, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8761349320411682, "num_tokens": 696631478.0, "step": 18344 }, { "epoch": 2.333672560742908, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6266138553619385, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8844159841537476, "num_tokens": 696669634.0, "step": 18345 }, { "epoch": 2.3337997710214986, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7404122352600098, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8728269934654236, "num_tokens": 696707642.0, "step": 18346 }, { "epoch": 2.333926981300089, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.617325782775879, "learning_rate": 1e-05, "loss": 0.4004, "mean_token_accuracy": 0.8850264549255371, "num_tokens": 696747426.0, "step": 18347 }, { "epoch": 2.3340541915786797, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.763026714324951, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8659765720367432, "num_tokens": 696784114.0, "step": 18348 }, { "epoch": 2.33418140185727, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6139075756073, "learning_rate": 1e-05, "loss": 0.3937, "mean_token_accuracy": 0.8873394131660461, "num_tokens": 696822303.0, "step": 18349 }, { "epoch": 2.3343086121358607, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7906155586242676, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8775143623352051, "num_tokens": 696855460.0, "step": 18350 }, { "epoch": 2.3344358224144512, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6030941009521484, "learning_rate": 1e-05, "loss": 0.3487, "mean_token_accuracy": 0.9030868411064148, "num_tokens": 696891070.0, "step": 18351 }, { "epoch": 2.3345630326930418, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8012545108795166, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.880229115486145, "num_tokens": 696924679.0, "step": 18352 }, { "epoch": 2.3346902429716323, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6035380363464355, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.8837819695472717, "num_tokens": 696958894.0, "step": 18353 }, { "epoch": 2.3348174532502224, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.812685251235962, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.872700572013855, "num_tokens": 696995371.0, "step": 18354 }, { "epoch": 2.3349446635288134, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.598444700241089, "learning_rate": 1e-05, "loss": 0.3814, "mean_token_accuracy": 0.8913504481315613, "num_tokens": 697034284.0, "step": 18355 }, { "epoch": 2.3350718738074034, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8160953521728516, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.8916690349578857, "num_tokens": 697069805.0, "step": 18356 }, { "epoch": 2.335199084085994, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.616302728652954, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8812291026115417, "num_tokens": 697105326.0, "step": 18357 }, { "epoch": 2.3353262943645845, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8109452724456787, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8820115923881531, "num_tokens": 697144157.0, "step": 18358 }, { "epoch": 2.335453504643175, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6140124797821045, "learning_rate": 1e-05, "loss": 0.3926, "mean_token_accuracy": 0.8870975375175476, "num_tokens": 697178261.0, "step": 18359 }, { "epoch": 2.3355807149217656, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8144371509552, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8821097612380981, "num_tokens": 697224044.0, "step": 18360 }, { "epoch": 2.335707925200356, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6085503101348877, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8836297392845154, "num_tokens": 697262007.0, "step": 18361 }, { "epoch": 2.3358351354789466, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8106181621551514, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8878190517425537, "num_tokens": 697301689.0, "step": 18362 }, { "epoch": 2.335962345757537, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6156532764434814, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8831713795661926, "num_tokens": 697343509.0, "step": 18363 }, { "epoch": 2.3360895560361277, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.813685178756714, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8821828365325928, "num_tokens": 697380620.0, "step": 18364 }, { "epoch": 2.336216766314718, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6110808849334717, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8737180233001709, "num_tokens": 697422811.0, "step": 18365 }, { "epoch": 2.3363439765933087, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.823319435119629, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8752306699752808, "num_tokens": 697460527.0, "step": 18366 }, { "epoch": 2.3364711868718993, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.624995231628418, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8836942911148071, "num_tokens": 697495417.0, "step": 18367 }, { "epoch": 2.33659839715049, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8170297145843506, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8858252167701721, "num_tokens": 697534745.0, "step": 18368 }, { "epoch": 2.3367256074290803, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.618234634399414, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8803724050521851, "num_tokens": 697572074.0, "step": 18369 }, { "epoch": 2.336852817707671, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8235373497009277, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8766584396362305, "num_tokens": 697606943.0, "step": 18370 }, { "epoch": 2.3369800279862614, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6183245182037354, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8790225982666016, "num_tokens": 697643634.0, "step": 18371 }, { "epoch": 2.337107238264852, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8118319511413574, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8769825100898743, "num_tokens": 697682429.0, "step": 18372 }, { "epoch": 2.3372344485434424, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6259047985076904, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8791317939758301, "num_tokens": 697718039.0, "step": 18373 }, { "epoch": 2.337361658822033, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.804086685180664, "learning_rate": 1e-05, "loss": 0.41, "mean_token_accuracy": 0.886489987373352, "num_tokens": 697756179.0, "step": 18374 }, { "epoch": 2.3374888691006235, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6176631450653076, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8830722570419312, "num_tokens": 697795327.0, "step": 18375 }, { "epoch": 2.337616079379214, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.809295415878296, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8776045441627502, "num_tokens": 697830423.0, "step": 18376 }, { "epoch": 2.3377432896578045, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.637547731399536, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8577048778533936, "num_tokens": 697870890.0, "step": 18377 }, { "epoch": 2.337870499936395, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.794267416000366, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8704214692115784, "num_tokens": 697910880.0, "step": 18378 }, { "epoch": 2.337997710214985, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6317508220672607, "learning_rate": 1e-05, "loss": 0.3928, "mean_token_accuracy": 0.8865447044372559, "num_tokens": 697945509.0, "step": 18379 }, { "epoch": 2.338124920493576, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7829744815826416, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8600845336914062, "num_tokens": 697987374.0, "step": 18380 }, { "epoch": 2.338252130772166, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.621034860610962, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8781647682189941, "num_tokens": 698025130.0, "step": 18381 }, { "epoch": 2.3383793410507567, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.792924165725708, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8798943161964417, "num_tokens": 698057785.0, "step": 18382 }, { "epoch": 2.3385065513293473, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.605182409286499, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8870615363121033, "num_tokens": 698098952.0, "step": 18383 }, { "epoch": 2.338633761607938, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7851638793945312, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.8877476453781128, "num_tokens": 698142280.0, "step": 18384 }, { "epoch": 2.3387609718865283, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.625722646713257, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8741722106933594, "num_tokens": 698183083.0, "step": 18385 }, { "epoch": 2.338888182165119, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8039026260375977, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8784074187278748, "num_tokens": 698215446.0, "step": 18386 }, { "epoch": 2.3390153924437094, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.615187883377075, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.874557375907898, "num_tokens": 698257353.0, "step": 18387 }, { "epoch": 2.3391426027223, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8141608238220215, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8773244619369507, "num_tokens": 698292753.0, "step": 18388 }, { "epoch": 2.3392698130008904, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6245570182800293, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8826152086257935, "num_tokens": 698327282.0, "step": 18389 }, { "epoch": 2.339397023279481, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.800050973892212, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8778060674667358, "num_tokens": 698368006.0, "step": 18390 }, { "epoch": 2.3395242335580715, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6360325813293457, "learning_rate": 1e-05, "loss": 0.4966, "mean_token_accuracy": 0.8499600887298584, "num_tokens": 698406898.0, "step": 18391 }, { "epoch": 2.339651443836662, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.785151243209839, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8739318251609802, "num_tokens": 698448159.0, "step": 18392 }, { "epoch": 2.3397786541152525, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.673020601272583, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8869824409484863, "num_tokens": 698489075.0, "step": 18393 }, { "epoch": 2.339905864393843, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.758416175842285, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8769501447677612, "num_tokens": 698527619.0, "step": 18394 }, { "epoch": 2.3400330746724336, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.649881601333618, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8722336292266846, "num_tokens": 698565250.0, "step": 18395 }, { "epoch": 2.340160284951024, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7448134422302246, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.881257176399231, "num_tokens": 698603207.0, "step": 18396 }, { "epoch": 2.3402874952296147, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6466825008392334, "learning_rate": 1e-05, "loss": 0.3818, "mean_token_accuracy": 0.8920735716819763, "num_tokens": 698636747.0, "step": 18397 }, { "epoch": 2.340414705508205, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.732516050338745, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8890785574913025, "num_tokens": 698677885.0, "step": 18398 }, { "epoch": 2.3405419157867957, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6584863662719727, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8772868514060974, "num_tokens": 698721924.0, "step": 18399 }, { "epoch": 2.3406691260653862, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7480692863464355, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.867110013961792, "num_tokens": 698754847.0, "step": 18400 }, { "epoch": 2.3407963363439768, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.621687889099121, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8897600173950195, "num_tokens": 698798567.0, "step": 18401 }, { "epoch": 2.3409235466225673, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.735887050628662, "learning_rate": 1e-05, "loss": 0.3942, "mean_token_accuracy": 0.8854423761367798, "num_tokens": 698837574.0, "step": 18402 }, { "epoch": 2.341050756901158, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.67429256439209, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8753693699836731, "num_tokens": 698867550.0, "step": 18403 }, { "epoch": 2.341177967179748, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7253594398498535, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8895968198776245, "num_tokens": 698898859.0, "step": 18404 }, { "epoch": 2.341305177458339, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.653977870941162, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8842139840126038, "num_tokens": 698933449.0, "step": 18405 }, { "epoch": 2.341432387736929, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7098138332366943, "learning_rate": 1e-05, "loss": 0.3845, "mean_token_accuracy": 0.8913277387619019, "num_tokens": 698969045.0, "step": 18406 }, { "epoch": 2.3415595980155195, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6686501502990723, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8696192502975464, "num_tokens": 699006201.0, "step": 18407 }, { "epoch": 2.34168680829411, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6933789253234863, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.88737952709198, "num_tokens": 699044091.0, "step": 18408 }, { "epoch": 2.3418140185727006, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6660337448120117, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8786808252334595, "num_tokens": 699075519.0, "step": 18409 }, { "epoch": 2.341941228851291, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6905102729797363, "learning_rate": 1e-05, "loss": 0.4012, "mean_token_accuracy": 0.8873728513717651, "num_tokens": 699113129.0, "step": 18410 }, { "epoch": 2.3420684391298816, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6733055114746094, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8804172873497009, "num_tokens": 699147515.0, "step": 18411 }, { "epoch": 2.342195649408472, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.671631336212158, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8755168914794922, "num_tokens": 699190578.0, "step": 18412 }, { "epoch": 2.3423228596870627, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6781790256500244, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8738958835601807, "num_tokens": 699230422.0, "step": 18413 }, { "epoch": 2.342450069965653, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6680755615234375, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8816345930099487, "num_tokens": 699268662.0, "step": 18414 }, { "epoch": 2.3425772802442437, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6711337566375732, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8737881779670715, "num_tokens": 699312498.0, "step": 18415 }, { "epoch": 2.3427044905228342, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6691627502441406, "learning_rate": 1e-05, "loss": 0.3845, "mean_token_accuracy": 0.889405369758606, "num_tokens": 699354143.0, "step": 18416 }, { "epoch": 2.3428317008014248, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6792476177215576, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8859589099884033, "num_tokens": 699392099.0, "step": 18417 }, { "epoch": 2.3429589110800153, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6708486080169678, "learning_rate": 1e-05, "loss": 0.4003, "mean_token_accuracy": 0.8850116729736328, "num_tokens": 699428449.0, "step": 18418 }, { "epoch": 2.343086121358606, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.683518409729004, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.876457691192627, "num_tokens": 699469359.0, "step": 18419 }, { "epoch": 2.3432133316371964, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6719276905059814, "learning_rate": 1e-05, "loss": 0.3955, "mean_token_accuracy": 0.8876714110374451, "num_tokens": 699507481.0, "step": 18420 }, { "epoch": 2.343340541915787, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.68042254447937, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8711250424385071, "num_tokens": 699548335.0, "step": 18421 }, { "epoch": 2.3434677521943774, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6814184188842773, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8707073926925659, "num_tokens": 699587460.0, "step": 18422 }, { "epoch": 2.343594962472968, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.666689395904541, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8723153471946716, "num_tokens": 699631128.0, "step": 18423 }, { "epoch": 2.3437221727515585, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.682899236679077, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8689795136451721, "num_tokens": 699672133.0, "step": 18424 }, { "epoch": 2.343849383030149, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6953318119049072, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8754087686538696, "num_tokens": 699705280.0, "step": 18425 }, { "epoch": 2.3439765933087395, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.679696559906006, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8754105567932129, "num_tokens": 699740959.0, "step": 18426 }, { "epoch": 2.3441038035873296, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7001731395721436, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8624738454818726, "num_tokens": 699775546.0, "step": 18427 }, { "epoch": 2.3442310138659206, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6624653339385986, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8809632062911987, "num_tokens": 699809777.0, "step": 18428 }, { "epoch": 2.3443582241445107, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6933207511901855, "learning_rate": 1e-05, "loss": 0.3782, "mean_token_accuracy": 0.8927109241485596, "num_tokens": 699846140.0, "step": 18429 }, { "epoch": 2.344485434423101, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6578001976013184, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8796111345291138, "num_tokens": 699882769.0, "step": 18430 }, { "epoch": 2.3446126447016917, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6901278495788574, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8769592642784119, "num_tokens": 699923658.0, "step": 18431 }, { "epoch": 2.3447398549802823, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6581614017486572, "learning_rate": 1e-05, "loss": 0.3675, "mean_token_accuracy": 0.8912733793258667, "num_tokens": 699957457.0, "step": 18432 }, { "epoch": 2.344867065258873, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.69459867477417, "learning_rate": 1e-05, "loss": 0.3866, "mean_token_accuracy": 0.8890594840049744, "num_tokens": 699994852.0, "step": 18433 }, { "epoch": 2.3449942755374633, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.655658006668091, "learning_rate": 1e-05, "loss": 0.3908, "mean_token_accuracy": 0.8857018947601318, "num_tokens": 700030983.0, "step": 18434 }, { "epoch": 2.345121485816054, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6954119205474854, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8843625783920288, "num_tokens": 700067907.0, "step": 18435 }, { "epoch": 2.3452486960946444, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.659344434738159, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8797417283058167, "num_tokens": 700107989.0, "step": 18436 }, { "epoch": 2.345375906373235, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.695732831954956, "learning_rate": 1e-05, "loss": 0.3816, "mean_token_accuracy": 0.8928173780441284, "num_tokens": 700145870.0, "step": 18437 }, { "epoch": 2.3455031166518254, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6581156253814697, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.888978123664856, "num_tokens": 700186762.0, "step": 18438 }, { "epoch": 2.345630326930416, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6920278072357178, "learning_rate": 1e-05, "loss": 0.3968, "mean_token_accuracy": 0.8854916095733643, "num_tokens": 700227307.0, "step": 18439 }, { "epoch": 2.3457575372090065, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.667858362197876, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8755205869674683, "num_tokens": 700263660.0, "step": 18440 }, { "epoch": 2.345884747487597, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7007739543914795, "learning_rate": 1e-05, "loss": 0.3772, "mean_token_accuracy": 0.8946714401245117, "num_tokens": 700294835.0, "step": 18441 }, { "epoch": 2.3460119577661875, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.653618097305298, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8855927586555481, "num_tokens": 700335977.0, "step": 18442 }, { "epoch": 2.346139168044778, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7065269947052, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8793954849243164, "num_tokens": 700369388.0, "step": 18443 }, { "epoch": 2.3462663783233686, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.651916980743408, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8779475092887878, "num_tokens": 700409575.0, "step": 18444 }, { "epoch": 2.346393588601959, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.701111316680908, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.891554594039917, "num_tokens": 700449463.0, "step": 18445 }, { "epoch": 2.3465207988805497, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.66013765335083, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8827344179153442, "num_tokens": 700483765.0, "step": 18446 }, { "epoch": 2.34664800915914, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6952950954437256, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8854891061782837, "num_tokens": 700521953.0, "step": 18447 }, { "epoch": 2.3467752194377307, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.659428119659424, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8734275698661804, "num_tokens": 700559662.0, "step": 18448 }, { "epoch": 2.3469024297163212, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6908459663391113, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8816168308258057, "num_tokens": 700598880.0, "step": 18449 }, { "epoch": 2.3470296399949118, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6591310501098633, "learning_rate": 1e-05, "loss": 0.3612, "mean_token_accuracy": 0.8971607089042664, "num_tokens": 700631712.0, "step": 18450 }, { "epoch": 2.3471568502735023, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7113945484161377, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8689597249031067, "num_tokens": 700665420.0, "step": 18451 }, { "epoch": 2.3472840605520924, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.65019154548645, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8830716609954834, "num_tokens": 700704603.0, "step": 18452 }, { "epoch": 2.3474112708306833, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7134664058685303, "learning_rate": 1e-05, "loss": 0.3914, "mean_token_accuracy": 0.8852987289428711, "num_tokens": 700738730.0, "step": 18453 }, { "epoch": 2.3475384811092734, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6459858417510986, "learning_rate": 1e-05, "loss": 0.3711, "mean_token_accuracy": 0.8970041871070862, "num_tokens": 700773848.0, "step": 18454 }, { "epoch": 2.347665691387864, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.71891188621521, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8810795545578003, "num_tokens": 700810352.0, "step": 18455 }, { "epoch": 2.3477929016664545, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.653066873550415, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8756991624832153, "num_tokens": 700846432.0, "step": 18456 }, { "epoch": 2.347920111945045, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.710747241973877, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.8854416608810425, "num_tokens": 700883674.0, "step": 18457 }, { "epoch": 2.3480473222236355, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6407155990600586, "learning_rate": 1e-05, "loss": 0.4823, "mean_token_accuracy": 0.8605867624282837, "num_tokens": 700925096.0, "step": 18458 }, { "epoch": 2.348174532502226, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.720083236694336, "learning_rate": 1e-05, "loss": 0.3743, "mean_token_accuracy": 0.8942966461181641, "num_tokens": 700957301.0, "step": 18459 }, { "epoch": 2.3483017427808166, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.64353346824646, "learning_rate": 1e-05, "loss": 0.3899, "mean_token_accuracy": 0.8889069557189941, "num_tokens": 700990636.0, "step": 18460 }, { "epoch": 2.348428953059407, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7172141075134277, "learning_rate": 1e-05, "loss": 0.3777, "mean_token_accuracy": 0.8918443322181702, "num_tokens": 701026379.0, "step": 18461 }, { "epoch": 2.3485561633379977, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6299426555633545, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8796865940093994, "num_tokens": 701070377.0, "step": 18462 }, { "epoch": 2.348683373616588, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.731475353240967, "learning_rate": 1e-05, "loss": 0.3919, "mean_token_accuracy": 0.8869377374649048, "num_tokens": 701107691.0, "step": 18463 }, { "epoch": 2.3488105838951787, "ewc_loss": 0.07470703125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.622837543487549, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8825575709342957, "num_tokens": 701147775.0, "step": 18464 }, { "epoch": 2.3489377941737692, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7360448837280273, "learning_rate": 1e-05, "loss": 0.3873, "mean_token_accuracy": 0.8899784088134766, "num_tokens": 701187732.0, "step": 18465 }, { "epoch": 2.3490650044523598, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.632835626602173, "learning_rate": 1e-05, "loss": 0.3762, "mean_token_accuracy": 0.8915899991989136, "num_tokens": 701220916.0, "step": 18466 }, { "epoch": 2.3491922147309503, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7399754524230957, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8833023309707642, "num_tokens": 701258831.0, "step": 18467 }, { "epoch": 2.349319425009541, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.616030216217041, "learning_rate": 1e-05, "loss": 0.3948, "mean_token_accuracy": 0.8859243988990784, "num_tokens": 701299310.0, "step": 18468 }, { "epoch": 2.3494466352881314, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7825965881347656, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8713006377220154, "num_tokens": 701331471.0, "step": 18469 }, { "epoch": 2.349573845566722, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.607527256011963, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8804063200950623, "num_tokens": 701371126.0, "step": 18470 }, { "epoch": 2.3497010558453124, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7596092224121094, "learning_rate": 1e-05, "loss": 0.3489, "mean_token_accuracy": 0.903634786605835, "num_tokens": 701413612.0, "step": 18471 }, { "epoch": 2.349828266123903, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.618790626525879, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8608781099319458, "num_tokens": 701456085.0, "step": 18472 }, { "epoch": 2.3499554764024935, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.765169858932495, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8798447847366333, "num_tokens": 701494409.0, "step": 18473 }, { "epoch": 2.350082686681084, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6078381538391113, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.877876877784729, "num_tokens": 701533610.0, "step": 18474 }, { "epoch": 2.3502098969596745, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8869426250457764, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8661766648292542, "num_tokens": 701571324.0, "step": 18475 }, { "epoch": 2.350337107238265, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6040048599243164, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8676410913467407, "num_tokens": 701609077.0, "step": 18476 }, { "epoch": 2.350464317516855, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.818148612976074, "learning_rate": 1e-05, "loss": 0.359, "mean_token_accuracy": 0.900165319442749, "num_tokens": 701647281.0, "step": 18477 }, { "epoch": 2.350591527795446, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.5956411361694336, "learning_rate": 1e-05, "loss": 0.3871, "mean_token_accuracy": 0.8911150693893433, "num_tokens": 701680039.0, "step": 18478 }, { "epoch": 2.350718738074036, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000339508056640625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.839021682739258, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8767790198326111, "num_tokens": 701721679.0, "step": 18479 }, { "epoch": 2.3508459483526267, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5940654277801514, "learning_rate": 1e-05, "loss": 0.364, "mean_token_accuracy": 0.8974701166152954, "num_tokens": 701758692.0, "step": 18480 }, { "epoch": 2.3509731586312173, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8650505542755127, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.86562180519104, "num_tokens": 701793038.0, "step": 18481 }, { "epoch": 2.351100368909808, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6103179454803467, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8802503347396851, "num_tokens": 701831172.0, "step": 18482 }, { "epoch": 2.3512275791883983, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8536369800567627, "learning_rate": 1e-05, "loss": 0.3839, "mean_token_accuracy": 0.8920308947563171, "num_tokens": 701865644.0, "step": 18483 }, { "epoch": 2.351354789466989, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6129744052886963, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8815339803695679, "num_tokens": 701903928.0, "step": 18484 }, { "epoch": 2.3514819997455794, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8661906719207764, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8660772442817688, "num_tokens": 701937440.0, "step": 18485 }, { "epoch": 2.35160921002417, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6262850761413574, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8823286294937134, "num_tokens": 701975936.0, "step": 18486 }, { "epoch": 2.3517364203027604, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8543949127197266, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8781710863113403, "num_tokens": 702009504.0, "step": 18487 }, { "epoch": 2.351863630581351, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6388041973114014, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8760124444961548, "num_tokens": 702046631.0, "step": 18488 }, { "epoch": 2.3519908408599415, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.828312873840332, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8818756937980652, "num_tokens": 702086915.0, "step": 18489 }, { "epoch": 2.352118051138532, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6329400539398193, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.875352144241333, "num_tokens": 702129214.0, "step": 18490 }, { "epoch": 2.3522452614171225, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.835422992706299, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8756401538848877, "num_tokens": 702167450.0, "step": 18491 }, { "epoch": 2.352372471695713, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6381537914276123, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8811721801757812, "num_tokens": 702205967.0, "step": 18492 }, { "epoch": 2.3524996819743036, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.864610195159912, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8791463375091553, "num_tokens": 702239570.0, "step": 18493 }, { "epoch": 2.352626892252894, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.635018825531006, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8764812350273132, "num_tokens": 702279276.0, "step": 18494 }, { "epoch": 2.3527541025314846, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8216664791107178, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8865982890129089, "num_tokens": 702319695.0, "step": 18495 }, { "epoch": 2.352881312810075, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.622553825378418, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.885488748550415, "num_tokens": 702361933.0, "step": 18496 }, { "epoch": 2.3530085230886657, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.827033758163452, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8821332454681396, "num_tokens": 702398532.0, "step": 18497 }, { "epoch": 2.3531357333672562, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.634310245513916, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8810230493545532, "num_tokens": 702437880.0, "step": 18498 }, { "epoch": 2.3532629436458468, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8152856826782227, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8749406933784485, "num_tokens": 702478908.0, "step": 18499 }, { "epoch": 2.353390153924437, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6307339668273926, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8721314072608948, "num_tokens": 702523817.0, "step": 18500 }, { "epoch": 2.353517364203028, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7870078086853027, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8817650079727173, "num_tokens": 702569682.0, "step": 18501 }, { "epoch": 2.353644574481618, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6392226219177246, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.87355637550354, "num_tokens": 702604746.0, "step": 18502 }, { "epoch": 2.353771784760209, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8112804889678955, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.871526837348938, "num_tokens": 702637660.0, "step": 18503 }, { "epoch": 2.353898995038799, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.615307092666626, "learning_rate": 1e-05, "loss": 0.3636, "mean_token_accuracy": 0.8980361223220825, "num_tokens": 702676942.0, "step": 18504 }, { "epoch": 2.3540262053173895, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7990708351135254, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.867154061794281, "num_tokens": 702718629.0, "step": 18505 }, { "epoch": 2.35415341559598, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6336851119995117, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8821471929550171, "num_tokens": 702757698.0, "step": 18506 }, { "epoch": 2.3542806258745705, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7799134254455566, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8881754875183105, "num_tokens": 702797337.0, "step": 18507 }, { "epoch": 2.354407836153161, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.64981746673584, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.870406448841095, "num_tokens": 702832002.0, "step": 18508 }, { "epoch": 2.3545350464317516, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7652831077575684, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8864032030105591, "num_tokens": 702874472.0, "step": 18509 }, { "epoch": 2.354662256710342, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6649258136749268, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8779844045639038, "num_tokens": 702909867.0, "step": 18510 }, { "epoch": 2.3547894669889327, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7429466247558594, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8886019587516785, "num_tokens": 702945817.0, "step": 18511 }, { "epoch": 2.354916677267523, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.636335849761963, "learning_rate": 1e-05, "loss": 0.3768, "mean_token_accuracy": 0.8966002464294434, "num_tokens": 702986373.0, "step": 18512 }, { "epoch": 2.3550438875461137, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.745685338973999, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8781147003173828, "num_tokens": 703026077.0, "step": 18513 }, { "epoch": 2.3551710978247042, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6362853050231934, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8796395063400269, "num_tokens": 703067042.0, "step": 18514 }, { "epoch": 2.3552983081032948, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.736300468444824, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8887167572975159, "num_tokens": 703105290.0, "step": 18515 }, { "epoch": 2.3554255183818853, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6389501094818115, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.876259446144104, "num_tokens": 703144446.0, "step": 18516 }, { "epoch": 2.355552728660476, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7441208362579346, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8756296038627625, "num_tokens": 703179490.0, "step": 18517 }, { "epoch": 2.3556799389390664, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6441924571990967, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8743505477905273, "num_tokens": 703214218.0, "step": 18518 }, { "epoch": 2.355807149217657, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7281863689422607, "learning_rate": 1e-05, "loss": 0.4037, "mean_token_accuracy": 0.8844351768493652, "num_tokens": 703252576.0, "step": 18519 }, { "epoch": 2.3559343594962474, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.626441240310669, "learning_rate": 1e-05, "loss": 0.3493, "mean_token_accuracy": 0.9015403389930725, "num_tokens": 703291129.0, "step": 18520 }, { "epoch": 2.356061569774838, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.745250701904297, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8708341717720032, "num_tokens": 703326864.0, "step": 18521 }, { "epoch": 2.3561887800534285, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.636993646621704, "learning_rate": 1e-05, "loss": 0.4156, "mean_token_accuracy": 0.880392849445343, "num_tokens": 703369719.0, "step": 18522 }, { "epoch": 2.356315990332019, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.734788179397583, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8803149461746216, "num_tokens": 703415304.0, "step": 18523 }, { "epoch": 2.3564432006106095, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6309988498687744, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8740473985671997, "num_tokens": 703456273.0, "step": 18524 }, { "epoch": 2.3565704108891996, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.744900941848755, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8745877146720886, "num_tokens": 703491355.0, "step": 18525 }, { "epoch": 2.3566976211677906, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.628657579421997, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8837990760803223, "num_tokens": 703532951.0, "step": 18526 }, { "epoch": 2.3568248314463807, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.738293170928955, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8797997236251831, "num_tokens": 703575306.0, "step": 18527 }, { "epoch": 2.356952041724971, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.622812509536743, "learning_rate": 1e-05, "loss": 0.3871, "mean_token_accuracy": 0.8869544267654419, "num_tokens": 703615198.0, "step": 18528 }, { "epoch": 2.3570792520035617, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7500431537628174, "learning_rate": 1e-05, "loss": 0.3885, "mean_token_accuracy": 0.8896110653877258, "num_tokens": 703651780.0, "step": 18529 }, { "epoch": 2.3572064622821522, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6592159271240234, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8749024868011475, "num_tokens": 703685273.0, "step": 18530 }, { "epoch": 2.3573336725607428, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.729464530944824, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.8906728029251099, "num_tokens": 703722185.0, "step": 18531 }, { "epoch": 2.3574608828393333, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.643749237060547, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8858333826065063, "num_tokens": 703760229.0, "step": 18532 }, { "epoch": 2.357588093117924, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.746713161468506, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8877800703048706, "num_tokens": 703794286.0, "step": 18533 }, { "epoch": 2.3577153033965144, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.64345383644104, "learning_rate": 1e-05, "loss": 0.3988, "mean_token_accuracy": 0.8864138722419739, "num_tokens": 703827486.0, "step": 18534 }, { "epoch": 2.357842513675105, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7458078861236572, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8660805225372314, "num_tokens": 703861398.0, "step": 18535 }, { "epoch": 2.3579697239536954, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6298205852508545, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8749665021896362, "num_tokens": 703901690.0, "step": 18536 }, { "epoch": 2.358096934232286, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7381510734558105, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8821454048156738, "num_tokens": 703943828.0, "step": 18537 }, { "epoch": 2.3582241445108765, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.633810043334961, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8732884526252747, "num_tokens": 703984893.0, "step": 18538 }, { "epoch": 2.358351354789467, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.745779275894165, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8712608814239502, "num_tokens": 704023633.0, "step": 18539 }, { "epoch": 2.3584785650680575, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.633497714996338, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8792546987533569, "num_tokens": 704058762.0, "step": 18540 }, { "epoch": 2.358605775346648, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7465150356292725, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.877617597579956, "num_tokens": 704099100.0, "step": 18541 }, { "epoch": 2.3587329856252386, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6652567386627197, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8626793622970581, "num_tokens": 704131900.0, "step": 18542 }, { "epoch": 2.358860195903829, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.747342824935913, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8860626220703125, "num_tokens": 704161940.0, "step": 18543 }, { "epoch": 2.3589874061824196, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.624008893966675, "learning_rate": 1e-05, "loss": 0.3844, "mean_token_accuracy": 0.8879401087760925, "num_tokens": 704200698.0, "step": 18544 }, { "epoch": 2.35911461646101, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7405242919921875, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8789850473403931, "num_tokens": 704243317.0, "step": 18545 }, { "epoch": 2.3592418267396007, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.634896755218506, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.8899942636489868, "num_tokens": 704277258.0, "step": 18546 }, { "epoch": 2.3593690370181912, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7421414852142334, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8910550475120544, "num_tokens": 704315932.0, "step": 18547 }, { "epoch": 2.3594962472967818, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.636026620864868, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8668533563613892, "num_tokens": 704357355.0, "step": 18548 }, { "epoch": 2.3596234575753723, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.751060962677002, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8738889694213867, "num_tokens": 704392981.0, "step": 18549 }, { "epoch": 2.3597506678539624, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.632097005844116, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.882061243057251, "num_tokens": 704430530.0, "step": 18550 }, { "epoch": 2.3598778781325533, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7455759048461914, "learning_rate": 1e-05, "loss": 0.3843, "mean_token_accuracy": 0.890575647354126, "num_tokens": 704467444.0, "step": 18551 }, { "epoch": 2.3600050884111434, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.625826835632324, "learning_rate": 1e-05, "loss": 0.3792, "mean_token_accuracy": 0.8930641412734985, "num_tokens": 704504179.0, "step": 18552 }, { "epoch": 2.360132298689734, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7432377338409424, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.88372802734375, "num_tokens": 704545346.0, "step": 18553 }, { "epoch": 2.3602595089683245, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.633775234222412, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8693550825119019, "num_tokens": 704583453.0, "step": 18554 }, { "epoch": 2.360386719246915, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.754802942276001, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8793160915374756, "num_tokens": 704617210.0, "step": 18555 }, { "epoch": 2.3605139295255055, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6278414726257324, "learning_rate": 1e-05, "loss": 0.3897, "mean_token_accuracy": 0.8866738080978394, "num_tokens": 704651002.0, "step": 18556 }, { "epoch": 2.360641139804096, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.74147367477417, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.879393994808197, "num_tokens": 704694718.0, "step": 18557 }, { "epoch": 2.3607683500826866, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6305699348449707, "learning_rate": 1e-05, "loss": 0.387, "mean_token_accuracy": 0.8898391723632812, "num_tokens": 704730266.0, "step": 18558 }, { "epoch": 2.360895560361277, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.758114814758301, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8840733766555786, "num_tokens": 704766422.0, "step": 18559 }, { "epoch": 2.3610227706398677, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6268320083618164, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8821815252304077, "num_tokens": 704802133.0, "step": 18560 }, { "epoch": 2.361149980918458, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.768355131149292, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8890450596809387, "num_tokens": 704832108.0, "step": 18561 }, { "epoch": 2.3612771911970487, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.636857032775879, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.872158944606781, "num_tokens": 704867056.0, "step": 18562 }, { "epoch": 2.3614044014756392, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7576045989990234, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.867953896522522, "num_tokens": 704905565.0, "step": 18563 }, { "epoch": 2.3615316117542298, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.614908218383789, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8817604184150696, "num_tokens": 704949690.0, "step": 18564 }, { "epoch": 2.3616588220328203, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7906363010406494, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8636906147003174, "num_tokens": 704985123.0, "step": 18565 }, { "epoch": 2.361786032311411, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.616886854171753, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.8829613327980042, "num_tokens": 705020317.0, "step": 18566 }, { "epoch": 2.3619132425900013, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7803492546081543, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8854352235794067, "num_tokens": 705059025.0, "step": 18567 }, { "epoch": 2.362040452868592, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.606421947479248, "learning_rate": 1e-05, "loss": 0.3653, "mean_token_accuracy": 0.8952677249908447, "num_tokens": 705094157.0, "step": 18568 }, { "epoch": 2.3621676631471824, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.780580759048462, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8820207118988037, "num_tokens": 705133642.0, "step": 18569 }, { "epoch": 2.362294873425773, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6460866928100586, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8725117444992065, "num_tokens": 705165253.0, "step": 18570 }, { "epoch": 2.3624220837043635, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.758876085281372, "learning_rate": 1e-05, "loss": 0.3838, "mean_token_accuracy": 0.8907663822174072, "num_tokens": 705203847.0, "step": 18571 }, { "epoch": 2.362549293982954, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6369152069091797, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8748952746391296, "num_tokens": 705241984.0, "step": 18572 }, { "epoch": 2.3626765042615445, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.768388032913208, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8704714179039001, "num_tokens": 705278462.0, "step": 18573 }, { "epoch": 2.362803714540135, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.638930559158325, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8880325555801392, "num_tokens": 705313004.0, "step": 18574 }, { "epoch": 2.362930924818725, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.865448474884033, "learning_rate": 1e-05, "loss": 0.3835, "mean_token_accuracy": 0.891350269317627, "num_tokens": 705350225.0, "step": 18575 }, { "epoch": 2.363058135097316, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6178436279296875, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8858543634414673, "num_tokens": 705380213.0, "step": 18576 }, { "epoch": 2.363185345375906, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8019392490386963, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8724734783172607, "num_tokens": 705419059.0, "step": 18577 }, { "epoch": 2.3633125556544967, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.5903866291046143, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8762686252593994, "num_tokens": 705459506.0, "step": 18578 }, { "epoch": 2.3634397659330872, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8217666149139404, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8801765441894531, "num_tokens": 705495719.0, "step": 18579 }, { "epoch": 2.3635669762116778, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6046876907348633, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8665233850479126, "num_tokens": 705536040.0, "step": 18580 }, { "epoch": 2.3636941864902683, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8292934894561768, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8805320262908936, "num_tokens": 705571541.0, "step": 18581 }, { "epoch": 2.363821396768859, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.604663133621216, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8842738270759583, "num_tokens": 705610048.0, "step": 18582 }, { "epoch": 2.3639486070474494, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.837609052658081, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8880962133407593, "num_tokens": 705645098.0, "step": 18583 }, { "epoch": 2.36407581732604, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.617366075515747, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.879787266254425, "num_tokens": 705680962.0, "step": 18584 }, { "epoch": 2.3642030276046304, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.835028648376465, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8762890100479126, "num_tokens": 705716983.0, "step": 18585 }, { "epoch": 2.364330237883221, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6349501609802246, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8800034523010254, "num_tokens": 705749735.0, "step": 18586 }, { "epoch": 2.3644574481618115, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.816272497177124, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.872043788433075, "num_tokens": 705788747.0, "step": 18587 }, { "epoch": 2.364584658440402, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.631624221801758, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.875515341758728, "num_tokens": 705827475.0, "step": 18588 }, { "epoch": 2.3647118687189925, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.806427478790283, "learning_rate": 1e-05, "loss": 0.3837, "mean_token_accuracy": 0.8931870460510254, "num_tokens": 705864940.0, "step": 18589 }, { "epoch": 2.364839078997583, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6316535472869873, "learning_rate": 1e-05, "loss": 0.3585, "mean_token_accuracy": 0.8964380025863647, "num_tokens": 705898274.0, "step": 18590 }, { "epoch": 2.3649662892761736, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.79950213432312, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8774203658103943, "num_tokens": 705936619.0, "step": 18591 }, { "epoch": 2.365093499554764, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.619762897491455, "learning_rate": 1e-05, "loss": 0.3461, "mean_token_accuracy": 0.9038771390914917, "num_tokens": 705975208.0, "step": 18592 }, { "epoch": 2.3652207098333546, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8039562702178955, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8723539710044861, "num_tokens": 706019948.0, "step": 18593 }, { "epoch": 2.365347920111945, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.625274896621704, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.8835850954055786, "num_tokens": 706056862.0, "step": 18594 }, { "epoch": 2.3654751303905357, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8106701374053955, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8755959272384644, "num_tokens": 706092591.0, "step": 18595 }, { "epoch": 2.3656023406691262, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.62355375289917, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8782104253768921, "num_tokens": 706132572.0, "step": 18596 }, { "epoch": 2.3657295509477168, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8013322353363037, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8829030990600586, "num_tokens": 706168264.0, "step": 18597 }, { "epoch": 2.365856761226307, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6236236095428467, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8800240755081177, "num_tokens": 706206684.0, "step": 18598 }, { "epoch": 2.365983971504898, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7910335063934326, "learning_rate": 1e-05, "loss": 0.3718, "mean_token_accuracy": 0.8954005241394043, "num_tokens": 706247547.0, "step": 18599 }, { "epoch": 2.366111181783488, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6124916076660156, "learning_rate": 1e-05, "loss": 0.3764, "mean_token_accuracy": 0.8944563269615173, "num_tokens": 706286286.0, "step": 18600 }, { "epoch": 2.3662383920620784, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7926084995269775, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8743902444839478, "num_tokens": 706329990.0, "step": 18601 }, { "epoch": 2.366365602340669, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.619004726409912, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8767147064208984, "num_tokens": 706370423.0, "step": 18602 }, { "epoch": 2.3664928126192595, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7804274559020996, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8838700652122498, "num_tokens": 706414380.0, "step": 18603 }, { "epoch": 2.36662002289785, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.625755786895752, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8797613382339478, "num_tokens": 706455403.0, "step": 18604 }, { "epoch": 2.3667472331764405, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7862534523010254, "learning_rate": 1e-05, "loss": 0.3559, "mean_token_accuracy": 0.9003658890724182, "num_tokens": 706490114.0, "step": 18605 }, { "epoch": 2.366874443455031, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.629711389541626, "learning_rate": 1e-05, "loss": 0.3735, "mean_token_accuracy": 0.8922845125198364, "num_tokens": 706523926.0, "step": 18606 }, { "epoch": 2.3670016537336216, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7777647972106934, "learning_rate": 1e-05, "loss": 0.3759, "mean_token_accuracy": 0.8946974277496338, "num_tokens": 706562626.0, "step": 18607 }, { "epoch": 2.367128864012212, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6221742630004883, "learning_rate": 1e-05, "loss": 0.3751, "mean_token_accuracy": 0.8928491473197937, "num_tokens": 706598735.0, "step": 18608 }, { "epoch": 2.3672560742908026, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7913641929626465, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8684920072555542, "num_tokens": 706638399.0, "step": 18609 }, { "epoch": 2.367383284569393, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6179962158203125, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8800649642944336, "num_tokens": 706680274.0, "step": 18610 }, { "epoch": 2.3675104948479837, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8061728477478027, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8658796548843384, "num_tokens": 706713918.0, "step": 18611 }, { "epoch": 2.3676377051265742, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.623894214630127, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8789769411087036, "num_tokens": 706751758.0, "step": 18612 }, { "epoch": 2.3677649154051648, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.824242353439331, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.890101432800293, "num_tokens": 706788243.0, "step": 18613 }, { "epoch": 2.3678921256837553, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6197283267974854, "learning_rate": 1e-05, "loss": 0.3988, "mean_token_accuracy": 0.8888764977455139, "num_tokens": 706821960.0, "step": 18614 }, { "epoch": 2.368019335962346, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.808204412460327, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8846158981323242, "num_tokens": 706858768.0, "step": 18615 }, { "epoch": 2.3681465462409363, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6046674251556396, "learning_rate": 1e-05, "loss": 0.3707, "mean_token_accuracy": 0.894688606262207, "num_tokens": 706898889.0, "step": 18616 }, { "epoch": 2.368273756519527, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8159730434417725, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8806885480880737, "num_tokens": 706934245.0, "step": 18617 }, { "epoch": 2.3684009667981174, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.60251522064209, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8765923976898193, "num_tokens": 706982092.0, "step": 18618 }, { "epoch": 2.368528177076708, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.831112861633301, "learning_rate": 1e-05, "loss": 0.3958, "mean_token_accuracy": 0.8856222629547119, "num_tokens": 707014748.0, "step": 18619 }, { "epoch": 2.3686553873552985, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6237313747406006, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8850822448730469, "num_tokens": 707048363.0, "step": 18620 }, { "epoch": 2.368782597633889, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.825648069381714, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8741505146026611, "num_tokens": 707084089.0, "step": 18621 }, { "epoch": 2.3689098079124795, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.617046356201172, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8716542720794678, "num_tokens": 707127334.0, "step": 18622 }, { "epoch": 2.3690370181910696, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.829833507537842, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8832167387008667, "num_tokens": 707166485.0, "step": 18623 }, { "epoch": 2.3691642284696606, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.633373975753784, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8818687200546265, "num_tokens": 707202031.0, "step": 18624 }, { "epoch": 2.3692914387482507, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8189947605133057, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.8835103511810303, "num_tokens": 707240649.0, "step": 18625 }, { "epoch": 2.369418649026841, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.632876396179199, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8806166052818298, "num_tokens": 707275996.0, "step": 18626 }, { "epoch": 2.3695458593054317, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8281450271606445, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8662817478179932, "num_tokens": 707312776.0, "step": 18627 }, { "epoch": 2.3696730695840222, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.638450860977173, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8749885559082031, "num_tokens": 707349519.0, "step": 18628 }, { "epoch": 2.3698002798626128, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7982022762298584, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8781620264053345, "num_tokens": 707389180.0, "step": 18629 }, { "epoch": 2.3699274901412033, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6243417263031006, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8819304704666138, "num_tokens": 707428603.0, "step": 18630 }, { "epoch": 2.370054700419794, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.800368547439575, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8690759539604187, "num_tokens": 707470297.0, "step": 18631 }, { "epoch": 2.3701819106983844, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6346311569213867, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8815356492996216, "num_tokens": 707507353.0, "step": 18632 }, { "epoch": 2.370309120976975, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.797041654586792, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8874421119689941, "num_tokens": 707541795.0, "step": 18633 }, { "epoch": 2.3704363312555654, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.638913154602051, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8791624307632446, "num_tokens": 707577333.0, "step": 18634 }, { "epoch": 2.370563541534156, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.813601016998291, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8671717643737793, "num_tokens": 707611306.0, "step": 18635 }, { "epoch": 2.3706907518127465, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.63283109664917, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8745017647743225, "num_tokens": 707647698.0, "step": 18636 }, { "epoch": 2.370817962091337, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7898688316345215, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8849438428878784, "num_tokens": 707683934.0, "step": 18637 }, { "epoch": 2.3709451723699275, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.625288486480713, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8758242130279541, "num_tokens": 707723216.0, "step": 18638 }, { "epoch": 2.371072382648518, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7875735759735107, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8844236135482788, "num_tokens": 707761194.0, "step": 18639 }, { "epoch": 2.3711995929271086, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6180827617645264, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8646612167358398, "num_tokens": 707807327.0, "step": 18640 }, { "epoch": 2.371326803205699, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7847418785095215, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8811277151107788, "num_tokens": 707846394.0, "step": 18641 }, { "epoch": 2.3714540134842896, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6219234466552734, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8765805959701538, "num_tokens": 707889099.0, "step": 18642 }, { "epoch": 2.37158122376288, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7830300331115723, "learning_rate": 1e-05, "loss": 0.3838, "mean_token_accuracy": 0.8879238367080688, "num_tokens": 707924729.0, "step": 18643 }, { "epoch": 2.3717084340414707, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6369738578796387, "learning_rate": 1e-05, "loss": 0.4859, "mean_token_accuracy": 0.8578723073005676, "num_tokens": 707965096.0, "step": 18644 }, { "epoch": 2.371835644320061, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7861196994781494, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8829638957977295, "num_tokens": 708002277.0, "step": 18645 }, { "epoch": 2.3719628545986517, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6433703899383545, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8742939233779907, "num_tokens": 708038978.0, "step": 18646 }, { "epoch": 2.3720900648772423, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.763032913208008, "learning_rate": 1e-05, "loss": 0.3781, "mean_token_accuracy": 0.8941912055015564, "num_tokens": 708076022.0, "step": 18647 }, { "epoch": 2.3722172751558324, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6388237476348877, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8678969740867615, "num_tokens": 708119606.0, "step": 18648 }, { "epoch": 2.3723444854344233, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7513270378112793, "learning_rate": 1e-05, "loss": 0.3836, "mean_token_accuracy": 0.8911443948745728, "num_tokens": 708159570.0, "step": 18649 }, { "epoch": 2.3724716957130134, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6440014839172363, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8657116889953613, "num_tokens": 708201092.0, "step": 18650 }, { "epoch": 2.372598905991604, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7455062866210938, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8833801746368408, "num_tokens": 708235808.0, "step": 18651 }, { "epoch": 2.3727261162701945, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6345055103302, "learning_rate": 1e-05, "loss": 0.3698, "mean_token_accuracy": 0.8982247114181519, "num_tokens": 708270863.0, "step": 18652 }, { "epoch": 2.372853326548785, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7540483474731445, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8674776554107666, "num_tokens": 708309653.0, "step": 18653 }, { "epoch": 2.3729805368273755, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.640190839767456, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8834912776947021, "num_tokens": 708346242.0, "step": 18654 }, { "epoch": 2.373107747105966, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.760437250137329, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8737547993659973, "num_tokens": 708383391.0, "step": 18655 }, { "epoch": 2.3732349573845566, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6436147689819336, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8867489695549011, "num_tokens": 708417094.0, "step": 18656 }, { "epoch": 2.373362167663147, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.746035575866699, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8777358531951904, "num_tokens": 708456648.0, "step": 18657 }, { "epoch": 2.3734893779417376, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6346094608306885, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8708770275115967, "num_tokens": 708499125.0, "step": 18658 }, { "epoch": 2.373616588220328, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.738483428955078, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8825360536575317, "num_tokens": 708544844.0, "step": 18659 }, { "epoch": 2.3737437984989187, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6319382190704346, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8778707981109619, "num_tokens": 708586852.0, "step": 18660 }, { "epoch": 2.3738710087775092, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7501280307769775, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8811838030815125, "num_tokens": 708624108.0, "step": 18661 }, { "epoch": 2.3739982190560998, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6289217472076416, "learning_rate": 1e-05, "loss": 0.3627, "mean_token_accuracy": 0.8987072706222534, "num_tokens": 708658214.0, "step": 18662 }, { "epoch": 2.3741254293346903, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7412266731262207, "learning_rate": 1e-05, "loss": 0.3795, "mean_token_accuracy": 0.8941183686256409, "num_tokens": 708695315.0, "step": 18663 }, { "epoch": 2.374252639613281, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.636697292327881, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8714921474456787, "num_tokens": 708734034.0, "step": 18664 }, { "epoch": 2.3743798498918713, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.759218215942383, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8787568807601929, "num_tokens": 708769758.0, "step": 18665 }, { "epoch": 2.374507060170462, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6463329792022705, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.87076735496521, "num_tokens": 708803639.0, "step": 18666 }, { "epoch": 2.3746342704490524, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.762690305709839, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8760098218917847, "num_tokens": 708837085.0, "step": 18667 }, { "epoch": 2.374761480727643, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6290063858032227, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.881992757320404, "num_tokens": 708876102.0, "step": 18668 }, { "epoch": 2.3748886910062335, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.750246524810791, "learning_rate": 1e-05, "loss": 0.3961, "mean_token_accuracy": 0.8903919458389282, "num_tokens": 708914260.0, "step": 18669 }, { "epoch": 2.375015901284824, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6463425159454346, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8680360317230225, "num_tokens": 708952071.0, "step": 18670 }, { "epoch": 2.3751431115634145, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.753960132598877, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8748461008071899, "num_tokens": 708989394.0, "step": 18671 }, { "epoch": 2.375270321842005, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6408138275146484, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8571959137916565, "num_tokens": 709033558.0, "step": 18672 }, { "epoch": 2.375397532120595, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.74908185005188, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8765139579772949, "num_tokens": 709071420.0, "step": 18673 }, { "epoch": 2.375524742399186, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.639051675796509, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.8897337317466736, "num_tokens": 709104810.0, "step": 18674 }, { "epoch": 2.375651952677776, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.74739933013916, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8732991218566895, "num_tokens": 709142789.0, "step": 18675 }, { "epoch": 2.3757791629563667, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.637714147567749, "learning_rate": 1e-05, "loss": 0.3891, "mean_token_accuracy": 0.8888194561004639, "num_tokens": 709177590.0, "step": 18676 }, { "epoch": 2.3759063732349572, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.739586591720581, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8887008428573608, "num_tokens": 709217177.0, "step": 18677 }, { "epoch": 2.3760335835135478, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.664236545562744, "learning_rate": 1e-05, "loss": 0.5112, "mean_token_accuracy": 0.8500864505767822, "num_tokens": 709252566.0, "step": 18678 }, { "epoch": 2.3761607937921383, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7362427711486816, "learning_rate": 1e-05, "loss": 0.3726, "mean_token_accuracy": 0.8927944898605347, "num_tokens": 709290386.0, "step": 18679 }, { "epoch": 2.376288004070729, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.636720895767212, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8801097869873047, "num_tokens": 709331110.0, "step": 18680 }, { "epoch": 2.3764152143493193, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7251060009002686, "learning_rate": 1e-05, "loss": 0.3843, "mean_token_accuracy": 0.8916724920272827, "num_tokens": 709371390.0, "step": 18681 }, { "epoch": 2.37654242462791, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6451897621154785, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8783190250396729, "num_tokens": 709413428.0, "step": 18682 }, { "epoch": 2.3766696349065004, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.734396457672119, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8814630508422852, "num_tokens": 709450332.0, "step": 18683 }, { "epoch": 2.376796845185091, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6526951789855957, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8577089905738831, "num_tokens": 709490039.0, "step": 18684 }, { "epoch": 2.3769240554636815, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7254977226257324, "learning_rate": 1e-05, "loss": 0.3975, "mean_token_accuracy": 0.8884280920028687, "num_tokens": 709531839.0, "step": 18685 }, { "epoch": 2.377051265742272, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.643267869949341, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8844447731971741, "num_tokens": 709568126.0, "step": 18686 }, { "epoch": 2.3771784760208625, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7270522117614746, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8832526803016663, "num_tokens": 709610546.0, "step": 18687 }, { "epoch": 2.377305686299453, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.076957702636719e-05, "grad_norm": 2.6262450218200684, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8729051947593689, "num_tokens": 709656869.0, "step": 18688 }, { "epoch": 2.3774328965780436, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7332897186279297, "learning_rate": 1e-05, "loss": 0.3906, "mean_token_accuracy": 0.8881863355636597, "num_tokens": 709695691.0, "step": 18689 }, { "epoch": 2.377560106856634, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.629833698272705, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8794581890106201, "num_tokens": 709736691.0, "step": 18690 }, { "epoch": 2.3776873171352246, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.736957550048828, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8739206790924072, "num_tokens": 709776277.0, "step": 18691 }, { "epoch": 2.377814527413815, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.623473882675171, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8716486096382141, "num_tokens": 709817222.0, "step": 18692 }, { "epoch": 2.3779417376924057, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7483294010162354, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8766863346099854, "num_tokens": 709859060.0, "step": 18693 }, { "epoch": 2.378068947970996, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6234207153320312, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.877962589263916, "num_tokens": 709897396.0, "step": 18694 }, { "epoch": 2.3781961582495867, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.74957537651062, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8827983140945435, "num_tokens": 709939797.0, "step": 18695 }, { "epoch": 2.378323368528177, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.631704092025757, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8762357234954834, "num_tokens": 709977102.0, "step": 18696 }, { "epoch": 2.378450578806768, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.748307704925537, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8776070475578308, "num_tokens": 710020824.0, "step": 18697 }, { "epoch": 2.378577789085358, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.642348527908325, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.881085991859436, "num_tokens": 710052263.0, "step": 18698 }, { "epoch": 2.3787049993639484, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7372748851776123, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8815414309501648, "num_tokens": 710095196.0, "step": 18699 }, { "epoch": 2.378832209642539, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6569268703460693, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8742407560348511, "num_tokens": 710127255.0, "step": 18700 }, { "epoch": 2.3789594199211295, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7517716884613037, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8773897886276245, "num_tokens": 710159129.0, "step": 18701 }, { "epoch": 2.37908663019972, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.634063482284546, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8771147727966309, "num_tokens": 710197244.0, "step": 18702 }, { "epoch": 2.3792138404783105, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7393927574157715, "learning_rate": 1e-05, "loss": 0.4787, "mean_token_accuracy": 0.8630142211914062, "num_tokens": 710241133.0, "step": 18703 }, { "epoch": 2.379341050756901, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.637587308883667, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8827011585235596, "num_tokens": 710283264.0, "step": 18704 }, { "epoch": 2.3794682610354916, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.740736961364746, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.87312912940979, "num_tokens": 710329203.0, "step": 18705 }, { "epoch": 2.379595471314082, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6338725090026855, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8837867975234985, "num_tokens": 710366772.0, "step": 18706 }, { "epoch": 2.3797226815926726, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.742117166519165, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8770087957382202, "num_tokens": 710407458.0, "step": 18707 }, { "epoch": 2.379849891871263, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.623307466506958, "learning_rate": 1e-05, "loss": 0.4033, "mean_token_accuracy": 0.8823007941246033, "num_tokens": 710449200.0, "step": 18708 }, { "epoch": 2.3799771021498537, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7539381980895996, "learning_rate": 1e-05, "loss": 0.3655, "mean_token_accuracy": 0.9004066586494446, "num_tokens": 710490764.0, "step": 18709 }, { "epoch": 2.3801043124284442, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6323931217193604, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8800500631332397, "num_tokens": 710526282.0, "step": 18710 }, { "epoch": 2.3802315227070348, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7492051124572754, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8875036239624023, "num_tokens": 710566523.0, "step": 18711 }, { "epoch": 2.3803587329856253, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6155307292938232, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.8834095001220703, "num_tokens": 710607640.0, "step": 18712 }, { "epoch": 2.380485943264216, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.768282413482666, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8776062726974487, "num_tokens": 710644785.0, "step": 18713 }, { "epoch": 2.3806131535428063, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.623230218887329, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8821485638618469, "num_tokens": 710681857.0, "step": 18714 }, { "epoch": 2.380740363821397, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7684760093688965, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8885849714279175, "num_tokens": 710718119.0, "step": 18715 }, { "epoch": 2.3808675740999874, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6219334602355957, "learning_rate": 1e-05, "loss": 0.469, "mean_token_accuracy": 0.8670524954795837, "num_tokens": 710758573.0, "step": 18716 }, { "epoch": 2.380994784378578, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7810463905334473, "learning_rate": 1e-05, "loss": 0.3819, "mean_token_accuracy": 0.8930203914642334, "num_tokens": 710791567.0, "step": 18717 }, { "epoch": 2.3811219946571685, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.606689214706421, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8860284090042114, "num_tokens": 710833060.0, "step": 18718 }, { "epoch": 2.381249204935759, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.793515205383301, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8762029409408569, "num_tokens": 710872175.0, "step": 18719 }, { "epoch": 2.3813764152143495, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.605584144592285, "learning_rate": 1e-05, "loss": 0.3975, "mean_token_accuracy": 0.8839308619499207, "num_tokens": 710910442.0, "step": 18720 }, { "epoch": 2.3815036254929396, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8013103008270264, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8806267380714417, "num_tokens": 710948656.0, "step": 18721 }, { "epoch": 2.3816308357715306, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6145987510681152, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8760948777198792, "num_tokens": 710989300.0, "step": 18722 }, { "epoch": 2.3817580460501206, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8061647415161133, "learning_rate": 1e-05, "loss": 0.3936, "mean_token_accuracy": 0.891625702381134, "num_tokens": 711027661.0, "step": 18723 }, { "epoch": 2.381885256328711, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.615307569503784, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8705282211303711, "num_tokens": 711067895.0, "step": 18724 }, { "epoch": 2.3820124666073017, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8232414722442627, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8739144802093506, "num_tokens": 711105567.0, "step": 18725 }, { "epoch": 2.3821396768858922, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6036832332611084, "learning_rate": 1e-05, "loss": 0.3698, "mean_token_accuracy": 0.8943397998809814, "num_tokens": 711145263.0, "step": 18726 }, { "epoch": 2.3822668871644828, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.810014009475708, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8844382762908936, "num_tokens": 711187024.0, "step": 18727 }, { "epoch": 2.3823940974430733, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6183629035949707, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8785691261291504, "num_tokens": 711226261.0, "step": 18728 }, { "epoch": 2.382521307721664, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003414154052734375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.813117027282715, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8717902898788452, "num_tokens": 711267005.0, "step": 18729 }, { "epoch": 2.3826485180002543, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6220216751098633, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8736814856529236, "num_tokens": 711307145.0, "step": 18730 }, { "epoch": 2.382775728278845, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.797752618789673, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8840780258178711, "num_tokens": 711351128.0, "step": 18731 }, { "epoch": 2.3829029385574354, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6300458908081055, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8870363235473633, "num_tokens": 711385344.0, "step": 18732 }, { "epoch": 2.383030148836026, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.807861328125, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8750963807106018, "num_tokens": 711421548.0, "step": 18733 }, { "epoch": 2.3831573591146165, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.625206232070923, "learning_rate": 1e-05, "loss": 0.4033, "mean_token_accuracy": 0.8828920125961304, "num_tokens": 711456952.0, "step": 18734 }, { "epoch": 2.383284569393207, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8095200061798096, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8849714398384094, "num_tokens": 711486985.0, "step": 18735 }, { "epoch": 2.3834117796717975, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6281237602233887, "learning_rate": 1e-05, "loss": 0.3819, "mean_token_accuracy": 0.8956289291381836, "num_tokens": 711521047.0, "step": 18736 }, { "epoch": 2.383538989950388, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.796651601791382, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.888079047203064, "num_tokens": 711561151.0, "step": 18737 }, { "epoch": 2.3836662002289786, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6362743377685547, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.861660361289978, "num_tokens": 711597975.0, "step": 18738 }, { "epoch": 2.383793410507569, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7788567543029785, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8791612386703491, "num_tokens": 711637923.0, "step": 18739 }, { "epoch": 2.3839206207861596, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6489250659942627, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.878235399723053, "num_tokens": 711675408.0, "step": 18740 }, { "epoch": 2.38404783106475, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7718937397003174, "learning_rate": 1e-05, "loss": 0.3975, "mean_token_accuracy": 0.8866201639175415, "num_tokens": 711712781.0, "step": 18741 }, { "epoch": 2.3841750413433407, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.631202459335327, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8819254636764526, "num_tokens": 711752974.0, "step": 18742 }, { "epoch": 2.384302251621931, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.767848253250122, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8817691802978516, "num_tokens": 711793526.0, "step": 18743 }, { "epoch": 2.3844294619005217, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.951516628265381, "learning_rate": 1e-05, "loss": 0.3653, "mean_token_accuracy": 0.8955875635147095, "num_tokens": 711830409.0, "step": 18744 }, { "epoch": 2.3845566721791123, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6329598426818848, "learning_rate": 1e-05, "loss": 0.3901, "mean_token_accuracy": 0.8887302875518799, "num_tokens": 711873476.0, "step": 18745 }, { "epoch": 2.3846838824577024, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.780646562576294, "learning_rate": 1e-05, "loss": 0.4037, "mean_token_accuracy": 0.8885565400123596, "num_tokens": 711912507.0, "step": 18746 }, { "epoch": 2.3848110927362933, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.618903160095215, "learning_rate": 1e-05, "loss": 0.3885, "mean_token_accuracy": 0.8882962465286255, "num_tokens": 711946884.0, "step": 18747 }, { "epoch": 2.3849383030148834, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7924227714538574, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8688907027244568, "num_tokens": 711987060.0, "step": 18748 }, { "epoch": 2.385065513293474, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.614818572998047, "learning_rate": 1e-05, "loss": 0.3678, "mean_token_accuracy": 0.8954058885574341, "num_tokens": 712023197.0, "step": 18749 }, { "epoch": 2.3851927235720645, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.793264389038086, "learning_rate": 1e-05, "loss": 0.384, "mean_token_accuracy": 0.8919345140457153, "num_tokens": 712060337.0, "step": 18750 }, { "epoch": 2.385319933850655, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.60619854927063, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8903719186782837, "num_tokens": 712101601.0, "step": 18751 }, { "epoch": 2.3854471441292455, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8143348693847656, "learning_rate": 1e-05, "loss": 0.3879, "mean_token_accuracy": 0.8892184495925903, "num_tokens": 712136311.0, "step": 18752 }, { "epoch": 2.385574354407836, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.621676206588745, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8729004859924316, "num_tokens": 712174386.0, "step": 18753 }, { "epoch": 2.3857015646864266, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8157637119293213, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8845944404602051, "num_tokens": 712209540.0, "step": 18754 }, { "epoch": 2.385828774965017, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6293458938598633, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8735455274581909, "num_tokens": 712245405.0, "step": 18755 }, { "epoch": 2.3859559852436076, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8194782733917236, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8606985211372375, "num_tokens": 712285013.0, "step": 18756 }, { "epoch": 2.386083195522198, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.620516777038574, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.863601803779602, "num_tokens": 712324850.0, "step": 18757 }, { "epoch": 2.3862104058007887, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.813673257827759, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8768702149391174, "num_tokens": 712364336.0, "step": 18758 }, { "epoch": 2.386337616079379, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6271729469299316, "learning_rate": 1e-05, "loss": 0.4691, "mean_token_accuracy": 0.8651153445243835, "num_tokens": 712403773.0, "step": 18759 }, { "epoch": 2.3864648263579697, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.818239212036133, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8671424388885498, "num_tokens": 712440307.0, "step": 18760 }, { "epoch": 2.3865920366365603, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.624582290649414, "learning_rate": 1e-05, "loss": 0.4058, "mean_token_accuracy": 0.882143497467041, "num_tokens": 712472008.0, "step": 18761 }, { "epoch": 2.386719246915151, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.796109676361084, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8859796524047852, "num_tokens": 712512935.0, "step": 18762 }, { "epoch": 2.3868464571937413, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6157474517822266, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.8882017135620117, "num_tokens": 712552783.0, "step": 18763 }, { "epoch": 2.386973667472332, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8112826347351074, "learning_rate": 1e-05, "loss": 0.3551, "mean_token_accuracy": 0.8994146585464478, "num_tokens": 712583752.0, "step": 18764 }, { "epoch": 2.3871008777509224, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6385490894317627, "learning_rate": 1e-05, "loss": 0.4646, "mean_token_accuracy": 0.8687642812728882, "num_tokens": 712623125.0, "step": 18765 }, { "epoch": 2.387228088029513, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7994141578674316, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8733493089675903, "num_tokens": 712664965.0, "step": 18766 }, { "epoch": 2.3873552983081034, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6305935382843018, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.876548171043396, "num_tokens": 712703571.0, "step": 18767 }, { "epoch": 2.387482508586694, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7907466888427734, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8835150003433228, "num_tokens": 712739830.0, "step": 18768 }, { "epoch": 2.3876097188652845, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6225626468658447, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8734058737754822, "num_tokens": 712782367.0, "step": 18769 }, { "epoch": 2.387736929143875, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8020293712615967, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8836266398429871, "num_tokens": 712816937.0, "step": 18770 }, { "epoch": 2.387864139422465, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6307449340820312, "learning_rate": 1e-05, "loss": 0.384, "mean_token_accuracy": 0.8905760645866394, "num_tokens": 712850198.0, "step": 18771 }, { "epoch": 2.387991349701056, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8070080280303955, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8818429708480835, "num_tokens": 712886201.0, "step": 18772 }, { "epoch": 2.388118559979646, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6278235912323, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8740270137786865, "num_tokens": 712924309.0, "step": 18773 }, { "epoch": 2.3882457702582367, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7823235988616943, "learning_rate": 1e-05, "loss": 0.3769, "mean_token_accuracy": 0.8957263231277466, "num_tokens": 712964508.0, "step": 18774 }, { "epoch": 2.3883729805368272, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6172637939453125, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8734908103942871, "num_tokens": 713002737.0, "step": 18775 }, { "epoch": 2.3885001908154178, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.795670747756958, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8809666633605957, "num_tokens": 713040859.0, "step": 18776 }, { "epoch": 2.3886274010940083, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6255714893341064, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8834598064422607, "num_tokens": 713073504.0, "step": 18777 }, { "epoch": 2.388754611372599, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7974393367767334, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8838374614715576, "num_tokens": 713110285.0, "step": 18778 }, { "epoch": 2.3888818216511893, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6439473628997803, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8738846778869629, "num_tokens": 713146227.0, "step": 18779 }, { "epoch": 2.38900903192978, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.777578353881836, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8765720725059509, "num_tokens": 713187632.0, "step": 18780 }, { "epoch": 2.3891362422083704, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.641295909881592, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8776544332504272, "num_tokens": 713223023.0, "step": 18781 }, { "epoch": 2.389263452486961, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7787106037139893, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8703188896179199, "num_tokens": 713264390.0, "step": 18782 }, { "epoch": 2.3893906627655515, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.650420665740967, "learning_rate": 1e-05, "loss": 0.3925, "mean_token_accuracy": 0.8877954483032227, "num_tokens": 713306504.0, "step": 18783 }, { "epoch": 2.389517873044142, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7584939002990723, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8816356062889099, "num_tokens": 713347556.0, "step": 18784 }, { "epoch": 2.3896450833227325, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6549222469329834, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.854932427406311, "num_tokens": 713386645.0, "step": 18785 }, { "epoch": 2.389772293601323, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7566163539886475, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8847267627716064, "num_tokens": 713419754.0, "step": 18786 }, { "epoch": 2.3898995038799136, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.668069362640381, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.879035472869873, "num_tokens": 713456374.0, "step": 18787 }, { "epoch": 2.390026714158504, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7462775707244873, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8823176622390747, "num_tokens": 713487961.0, "step": 18788 }, { "epoch": 2.3901539244370946, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6462671756744385, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8790375590324402, "num_tokens": 713525697.0, "step": 18789 }, { "epoch": 2.390281134715685, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.739611864089966, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8789688944816589, "num_tokens": 713563902.0, "step": 18790 }, { "epoch": 2.3904083449942757, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6455976963043213, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8696235418319702, "num_tokens": 713602996.0, "step": 18791 }, { "epoch": 2.390535555272866, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7428061962127686, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8800791501998901, "num_tokens": 713641725.0, "step": 18792 }, { "epoch": 2.3906627655514567, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6509885787963867, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8593161106109619, "num_tokens": 713682973.0, "step": 18793 }, { "epoch": 2.390789975830047, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.737027883529663, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8856768608093262, "num_tokens": 713718919.0, "step": 18794 }, { "epoch": 2.390917186108638, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.649395704269409, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8775045275688171, "num_tokens": 713757416.0, "step": 18795 }, { "epoch": 2.391044396387228, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.736567735671997, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8638710975646973, "num_tokens": 713795648.0, "step": 18796 }, { "epoch": 2.3911716066658184, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6458189487457275, "learning_rate": 1e-05, "loss": 0.3667, "mean_token_accuracy": 0.8981027603149414, "num_tokens": 713830095.0, "step": 18797 }, { "epoch": 2.391298816944409, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7340807914733887, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8728631734848022, "num_tokens": 713868958.0, "step": 18798 }, { "epoch": 2.3914260272229995, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6646041870117188, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.869225263595581, "num_tokens": 713911533.0, "step": 18799 }, { "epoch": 2.39155323750159, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7356433868408203, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8721796274185181, "num_tokens": 713952585.0, "step": 18800 }, { "epoch": 2.3916804477801805, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6346018314361572, "learning_rate": 1e-05, "loss": 0.372, "mean_token_accuracy": 0.8957202434539795, "num_tokens": 713990929.0, "step": 18801 }, { "epoch": 2.391807658058771, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7477855682373047, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8697855472564697, "num_tokens": 714027098.0, "step": 18802 }, { "epoch": 2.3919348683373616, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.645298480987549, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8714866638183594, "num_tokens": 714062444.0, "step": 18803 }, { "epoch": 2.392062078615952, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7321903705596924, "learning_rate": 1e-05, "loss": 0.3662, "mean_token_accuracy": 0.8972560167312622, "num_tokens": 714099091.0, "step": 18804 }, { "epoch": 2.3921892888945426, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6409902572631836, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8737603425979614, "num_tokens": 714135694.0, "step": 18805 }, { "epoch": 2.392316499173133, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.737527370452881, "learning_rate": 1e-05, "loss": 0.3871, "mean_token_accuracy": 0.8922381401062012, "num_tokens": 714170324.0, "step": 18806 }, { "epoch": 2.3924437094517237, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6466169357299805, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8826135396957397, "num_tokens": 714205080.0, "step": 18807 }, { "epoch": 2.392570919730314, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.72660493850708, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8776842355728149, "num_tokens": 714245297.0, "step": 18808 }, { "epoch": 2.3926981300089047, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6358413696289062, "learning_rate": 1e-05, "loss": 0.3834, "mean_token_accuracy": 0.8922490477561951, "num_tokens": 714284155.0, "step": 18809 }, { "epoch": 2.3928253402874953, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7396504878997803, "learning_rate": 1e-05, "loss": 0.4017, "mean_token_accuracy": 0.8841357231140137, "num_tokens": 714318688.0, "step": 18810 }, { "epoch": 2.392952550566086, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6252105236053467, "learning_rate": 1e-05, "loss": 0.3635, "mean_token_accuracy": 0.896098256111145, "num_tokens": 714353730.0, "step": 18811 }, { "epoch": 2.3930797608446763, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7452166080474854, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8795859813690186, "num_tokens": 714390156.0, "step": 18812 }, { "epoch": 2.393206971123267, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.637281656265259, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.864864706993103, "num_tokens": 714429435.0, "step": 18813 }, { "epoch": 2.3933341814018574, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7429754734039307, "learning_rate": 1e-05, "loss": 0.3672, "mean_token_accuracy": 0.8974526524543762, "num_tokens": 714470034.0, "step": 18814 }, { "epoch": 2.393461391680448, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6340126991271973, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8775794506072998, "num_tokens": 714505989.0, "step": 18815 }, { "epoch": 2.3935886019590384, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7511026859283447, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8743219375610352, "num_tokens": 714541727.0, "step": 18816 }, { "epoch": 2.393715812237629, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6404013633728027, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8817651867866516, "num_tokens": 714581429.0, "step": 18817 }, { "epoch": 2.3938430225162195, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7408268451690674, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8793137073516846, "num_tokens": 714621839.0, "step": 18818 }, { "epoch": 2.3939702327948096, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6292059421539307, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8768230080604553, "num_tokens": 714665459.0, "step": 18819 }, { "epoch": 2.3940974430734006, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.76065731048584, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.886772871017456, "num_tokens": 714699259.0, "step": 18820 }, { "epoch": 2.3942246533519906, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6293389797210693, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8719344139099121, "num_tokens": 714740330.0, "step": 18821 }, { "epoch": 2.394351863630581, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.768413543701172, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.883179247379303, "num_tokens": 714773330.0, "step": 18822 }, { "epoch": 2.3944790739091717, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6265833377838135, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8713014721870422, "num_tokens": 714812707.0, "step": 18823 }, { "epoch": 2.3946062841877622, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7614810466766357, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8838944435119629, "num_tokens": 714853496.0, "step": 18824 }, { "epoch": 2.3947334944663528, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6173198223114014, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8855754137039185, "num_tokens": 714895766.0, "step": 18825 }, { "epoch": 2.3948607047449433, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7631306648254395, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8823592662811279, "num_tokens": 714936089.0, "step": 18826 }, { "epoch": 2.394987915023534, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6087918281555176, "learning_rate": 1e-05, "loss": 0.3725, "mean_token_accuracy": 0.8940133452415466, "num_tokens": 714980759.0, "step": 18827 }, { "epoch": 2.3951151253021243, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7768194675445557, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8852877616882324, "num_tokens": 715019671.0, "step": 18828 }, { "epoch": 2.395242335580715, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.626244306564331, "learning_rate": 1e-05, "loss": 0.4738, "mean_token_accuracy": 0.8629173636436462, "num_tokens": 715059639.0, "step": 18829 }, { "epoch": 2.3953695458593054, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7761545181274414, "learning_rate": 1e-05, "loss": 0.3878, "mean_token_accuracy": 0.892570436000824, "num_tokens": 715096315.0, "step": 18830 }, { "epoch": 2.395496756137896, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.626687526702881, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.881178617477417, "num_tokens": 715130800.0, "step": 18831 }, { "epoch": 2.3956239664164865, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7625856399536133, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8742614984512329, "num_tokens": 715175369.0, "step": 18832 }, { "epoch": 2.395751176695077, "ewc_loss": 0.0751953125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.100799560546875e-05, "grad_norm": 2.6194581985473633, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8764781951904297, "num_tokens": 715216480.0, "step": 18833 }, { "epoch": 2.3958783869736675, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7640697956085205, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8875131011009216, "num_tokens": 715257182.0, "step": 18834 }, { "epoch": 2.396005597252258, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6420624256134033, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8743389844894409, "num_tokens": 715288793.0, "step": 18835 }, { "epoch": 2.3961328075308486, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7733964920043945, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8796247243881226, "num_tokens": 715323720.0, "step": 18836 }, { "epoch": 2.396260017809439, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.653045177459717, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.874183177947998, "num_tokens": 715356104.0, "step": 18837 }, { "epoch": 2.3963872280880296, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.752185344696045, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8888012170791626, "num_tokens": 715390727.0, "step": 18838 }, { "epoch": 2.39651443836662, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6472537517547607, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8849475383758545, "num_tokens": 715426340.0, "step": 18839 }, { "epoch": 2.3966416486452107, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.742199659347534, "learning_rate": 1e-05, "loss": 0.3372, "mean_token_accuracy": 0.9058784246444702, "num_tokens": 715458195.0, "step": 18840 }, { "epoch": 2.396768858923801, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.633999824523926, "learning_rate": 1e-05, "loss": 0.3888, "mean_token_accuracy": 0.8916182518005371, "num_tokens": 715496896.0, "step": 18841 }, { "epoch": 2.3968960692023917, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.743518114089966, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8678022027015686, "num_tokens": 715537985.0, "step": 18842 }, { "epoch": 2.3970232794809823, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.630784273147583, "learning_rate": 1e-05, "loss": 0.3609, "mean_token_accuracy": 0.8937727212905884, "num_tokens": 715574767.0, "step": 18843 }, { "epoch": 2.3971504897595723, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.763425588607788, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.8788304328918457, "num_tokens": 715608649.0, "step": 18844 }, { "epoch": 2.3972777000381633, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.619849681854248, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8808174133300781, "num_tokens": 715650278.0, "step": 18845 }, { "epoch": 2.3974049103167534, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.766890048980713, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8758134841918945, "num_tokens": 715692678.0, "step": 18846 }, { "epoch": 2.397532120595344, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6250929832458496, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8716453313827515, "num_tokens": 715733128.0, "step": 18847 }, { "epoch": 2.3976593308739345, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7889881134033203, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8781881332397461, "num_tokens": 715767272.0, "step": 18848 }, { "epoch": 2.397786541152525, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6195714473724365, "learning_rate": 1e-05, "loss": 0.3547, "mean_token_accuracy": 0.8978744149208069, "num_tokens": 715803497.0, "step": 18849 }, { "epoch": 2.3979137514311155, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.770777940750122, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8835239410400391, "num_tokens": 715847069.0, "step": 18850 }, { "epoch": 2.398040961709706, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.628892421722412, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8848527073860168, "num_tokens": 715887639.0, "step": 18851 }, { "epoch": 2.3981681719882966, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7773091793060303, "learning_rate": 1e-05, "loss": 0.3518, "mean_token_accuracy": 0.9010621309280396, "num_tokens": 715919427.0, "step": 18852 }, { "epoch": 2.398295382266887, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6281042098999023, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8675826787948608, "num_tokens": 715961287.0, "step": 18853 }, { "epoch": 2.3984225925454776, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.793444871902466, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8773864507675171, "num_tokens": 715995021.0, "step": 18854 }, { "epoch": 2.398549802824068, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6047754287719727, "learning_rate": 1e-05, "loss": 0.3779, "mean_token_accuracy": 0.8947218060493469, "num_tokens": 716039517.0, "step": 18855 }, { "epoch": 2.3986770131026587, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8085107803344727, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8653799295425415, "num_tokens": 716077116.0, "step": 18856 }, { "epoch": 2.398804223381249, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6278316974639893, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8735769987106323, "num_tokens": 716112800.0, "step": 18857 }, { "epoch": 2.3989314336598397, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8026840686798096, "learning_rate": 1e-05, "loss": 0.391, "mean_token_accuracy": 0.89085453748703, "num_tokens": 716148165.0, "step": 18858 }, { "epoch": 2.3990586439384303, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.620271682739258, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.875101625919342, "num_tokens": 716185745.0, "step": 18859 }, { "epoch": 2.399185854217021, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.789419412612915, "learning_rate": 1e-05, "loss": 0.3752, "mean_token_accuracy": 0.8941553235054016, "num_tokens": 716225177.0, "step": 18860 }, { "epoch": 2.3993130644956113, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.607609748840332, "learning_rate": 1e-05, "loss": 0.3664, "mean_token_accuracy": 0.8983632922172546, "num_tokens": 716265969.0, "step": 18861 }, { "epoch": 2.399440274774202, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8074066638946533, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8803197741508484, "num_tokens": 716302056.0, "step": 18862 }, { "epoch": 2.3995674850527924, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.619828701019287, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8698334693908691, "num_tokens": 716339452.0, "step": 18863 }, { "epoch": 2.399694695331383, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8008363246917725, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8817669749259949, "num_tokens": 716379930.0, "step": 18864 }, { "epoch": 2.3998219056099734, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.617449998855591, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8777021169662476, "num_tokens": 716416367.0, "step": 18865 }, { "epoch": 2.399949115888564, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8070592880249023, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8644076585769653, "num_tokens": 716460112.0, "step": 18866 }, { "epoch": 2.4000763261671545, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.622394561767578, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.8764544725418091, "num_tokens": 716497085.0, "step": 18867 }, { "epoch": 2.400203536445745, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7857847213745117, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8813326358795166, "num_tokens": 716542892.0, "step": 18868 }, { "epoch": 2.400330746724335, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6331894397735596, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8680962324142456, "num_tokens": 716578765.0, "step": 18869 }, { "epoch": 2.400457957002926, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7854557037353516, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8835560083389282, "num_tokens": 716618881.0, "step": 18870 }, { "epoch": 2.400585167281516, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6359810829162598, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8802720308303833, "num_tokens": 716651583.0, "step": 18871 }, { "epoch": 2.4007123775601067, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.794180154800415, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8697279691696167, "num_tokens": 716687118.0, "step": 18872 }, { "epoch": 2.400839587838697, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.642805576324463, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.870353102684021, "num_tokens": 716722543.0, "step": 18873 }, { "epoch": 2.4009667981172877, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.776479482650757, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8823211789131165, "num_tokens": 716757447.0, "step": 18874 }, { "epoch": 2.4010940083958783, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6230461597442627, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8789748549461365, "num_tokens": 716799736.0, "step": 18875 }, { "epoch": 2.401221218674469, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.77603816986084, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.875541090965271, "num_tokens": 716844850.0, "step": 18876 }, { "epoch": 2.4013484289530593, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6561999320983887, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8777533173561096, "num_tokens": 716878801.0, "step": 18877 }, { "epoch": 2.40147563923165, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7633299827575684, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8795982003211975, "num_tokens": 716914271.0, "step": 18878 }, { "epoch": 2.4016028495102404, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.640031337738037, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8779719471931458, "num_tokens": 716954094.0, "step": 18879 }, { "epoch": 2.401730059788831, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7526283264160156, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8836030960083008, "num_tokens": 716995061.0, "step": 18880 }, { "epoch": 2.4018572700674214, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6359341144561768, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.878516674041748, "num_tokens": 717032271.0, "step": 18881 }, { "epoch": 2.401984480346012, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7679617404937744, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8842862844467163, "num_tokens": 717068551.0, "step": 18882 }, { "epoch": 2.4021116906246025, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6275475025177, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8831465244293213, "num_tokens": 717107460.0, "step": 18883 }, { "epoch": 2.402238900903193, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7648677825927734, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8757601380348206, "num_tokens": 717144774.0, "step": 18884 }, { "epoch": 2.4023661111817836, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6350109577178955, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.879387617111206, "num_tokens": 717180808.0, "step": 18885 }, { "epoch": 2.402493321460374, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7541229724884033, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.8852384686470032, "num_tokens": 717220788.0, "step": 18886 }, { "epoch": 2.4026205317389646, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.630276679992676, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8850079774856567, "num_tokens": 717260257.0, "step": 18887 }, { "epoch": 2.402747742017555, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7744288444519043, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8871278166770935, "num_tokens": 717291911.0, "step": 18888 }, { "epoch": 2.4028749522961457, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.647477865219116, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8775813579559326, "num_tokens": 717322356.0, "step": 18889 }, { "epoch": 2.403002162574736, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7495338916778564, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8777898550033569, "num_tokens": 717363439.0, "step": 18890 }, { "epoch": 2.4031293728533267, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6365392208099365, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.876435399055481, "num_tokens": 717401050.0, "step": 18891 }, { "epoch": 2.403256583131917, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.747328042984009, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8783565759658813, "num_tokens": 717442346.0, "step": 18892 }, { "epoch": 2.403383793410508, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6275172233581543, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.884118914604187, "num_tokens": 717483387.0, "step": 18893 }, { "epoch": 2.403511003689098, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7511534690856934, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8812860250473022, "num_tokens": 717522208.0, "step": 18894 }, { "epoch": 2.4036382139676884, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6210274696350098, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8895618915557861, "num_tokens": 717563338.0, "step": 18895 }, { "epoch": 2.403765424246279, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.758007526397705, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8796061873435974, "num_tokens": 717603351.0, "step": 18896 }, { "epoch": 2.4038926345248695, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.626391887664795, "learning_rate": 1e-05, "loss": 0.3765, "mean_token_accuracy": 0.8913928270339966, "num_tokens": 717639886.0, "step": 18897 }, { "epoch": 2.40401984480346, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7984111309051514, "learning_rate": 1e-05, "loss": 0.4936, "mean_token_accuracy": 0.8585890531539917, "num_tokens": 717672228.0, "step": 18898 }, { "epoch": 2.4041470550820505, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6195104122161865, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8711707592010498, "num_tokens": 717709703.0, "step": 18899 }, { "epoch": 2.404274265360641, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8025195598602295, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8776556253433228, "num_tokens": 717746785.0, "step": 18900 }, { "epoch": 2.4044014756392316, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6081764698028564, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8751450777053833, "num_tokens": 717785786.0, "step": 18901 }, { "epoch": 2.404528685917822, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.814401388168335, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.877486526966095, "num_tokens": 717821833.0, "step": 18902 }, { "epoch": 2.4046558961964126, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.612346887588501, "learning_rate": 1e-05, "loss": 0.4015, "mean_token_accuracy": 0.8878146409988403, "num_tokens": 717858107.0, "step": 18903 }, { "epoch": 2.404783106475003, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.80773663520813, "learning_rate": 1e-05, "loss": 0.3758, "mean_token_accuracy": 0.8946451544761658, "num_tokens": 717893090.0, "step": 18904 }, { "epoch": 2.4049103167535937, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6431832313537598, "learning_rate": 1e-05, "loss": 0.3887, "mean_token_accuracy": 0.8899836540222168, "num_tokens": 717919703.0, "step": 18905 }, { "epoch": 2.405037527032184, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.794191360473633, "learning_rate": 1e-05, "loss": 0.3906, "mean_token_accuracy": 0.8916746377944946, "num_tokens": 717956877.0, "step": 18906 }, { "epoch": 2.4051647373107747, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6255996227264404, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8713006973266602, "num_tokens": 718002812.0, "step": 18907 }, { "epoch": 2.4052919475893653, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7950994968414307, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8817554712295532, "num_tokens": 718041170.0, "step": 18908 }, { "epoch": 2.405419157867956, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6203207969665527, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8767332434654236, "num_tokens": 718084168.0, "step": 18909 }, { "epoch": 2.4055463681465463, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8099679946899414, "learning_rate": 1e-05, "loss": 0.3873, "mean_token_accuracy": 0.8900460004806519, "num_tokens": 718118663.0, "step": 18910 }, { "epoch": 2.405673578425137, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6177051067352295, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8802501559257507, "num_tokens": 718155490.0, "step": 18911 }, { "epoch": 2.4058007887037274, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.812405824661255, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8773409128189087, "num_tokens": 718192301.0, "step": 18912 }, { "epoch": 2.405927998982318, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.614839553833008, "learning_rate": 1e-05, "loss": 0.3887, "mean_token_accuracy": 0.8891977667808533, "num_tokens": 718231390.0, "step": 18913 }, { "epoch": 2.4060552092609084, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8038039207458496, "learning_rate": 1e-05, "loss": 0.3861, "mean_token_accuracy": 0.8915594816207886, "num_tokens": 718269248.0, "step": 18914 }, { "epoch": 2.406182419539499, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6169753074645996, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8834876418113708, "num_tokens": 718305961.0, "step": 18915 }, { "epoch": 2.4063096298180895, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.828050136566162, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8751319646835327, "num_tokens": 718344659.0, "step": 18916 }, { "epoch": 2.4064368400966796, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.624544143676758, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8678946495056152, "num_tokens": 718386624.0, "step": 18917 }, { "epoch": 2.4065640503752705, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8089418411254883, "learning_rate": 1e-05, "loss": 0.3747, "mean_token_accuracy": 0.8930783271789551, "num_tokens": 718422154.0, "step": 18918 }, { "epoch": 2.4066912606538606, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.625635862350464, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8655648827552795, "num_tokens": 718460253.0, "step": 18919 }, { "epoch": 2.406818470932451, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.81215763092041, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8826556205749512, "num_tokens": 718495328.0, "step": 18920 }, { "epoch": 2.4069456812110417, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6310179233551025, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8835440874099731, "num_tokens": 718531244.0, "step": 18921 }, { "epoch": 2.407072891489632, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.794555187225342, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8826842904090881, "num_tokens": 718569766.0, "step": 18922 }, { "epoch": 2.4072001017682227, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.618496894836426, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.879281759262085, "num_tokens": 718612496.0, "step": 18923 }, { "epoch": 2.4073273120468133, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7945008277893066, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.881035566329956, "num_tokens": 718651251.0, "step": 18924 }, { "epoch": 2.407454522325404, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6420717239379883, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8797117471694946, "num_tokens": 718683204.0, "step": 18925 }, { "epoch": 2.4075817326039943, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7852554321289062, "learning_rate": 1e-05, "loss": 0.384, "mean_token_accuracy": 0.8893939256668091, "num_tokens": 718717114.0, "step": 18926 }, { "epoch": 2.407708942882585, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6398727893829346, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8716724514961243, "num_tokens": 718754756.0, "step": 18927 }, { "epoch": 2.4078361531611754, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.784320592880249, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8729866147041321, "num_tokens": 718792312.0, "step": 18928 }, { "epoch": 2.407963363439766, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6267764568328857, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8768680691719055, "num_tokens": 718834771.0, "step": 18929 }, { "epoch": 2.4080905737183564, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.784454584121704, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8762195706367493, "num_tokens": 718871124.0, "step": 18930 }, { "epoch": 2.408217783996947, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.622986316680908, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8782248497009277, "num_tokens": 718912037.0, "step": 18931 }, { "epoch": 2.4083449942755375, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.799154281616211, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8723764419555664, "num_tokens": 718953048.0, "step": 18932 }, { "epoch": 2.408472204554128, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6196787357330322, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8732567429542542, "num_tokens": 718991825.0, "step": 18933 }, { "epoch": 2.4085994148327186, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.805399179458618, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8784529566764832, "num_tokens": 719026847.0, "step": 18934 }, { "epoch": 2.408726625111309, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.623377799987793, "learning_rate": 1e-05, "loss": 0.3987, "mean_token_accuracy": 0.8847622871398926, "num_tokens": 719061415.0, "step": 18935 }, { "epoch": 2.4088538353898996, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8061349391937256, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.886390209197998, "num_tokens": 719097432.0, "step": 18936 }, { "epoch": 2.40898104566849, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.635061502456665, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.8822623491287231, "num_tokens": 719129301.0, "step": 18937 }, { "epoch": 2.4091082559470807, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.787743330001831, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8792682886123657, "num_tokens": 719171772.0, "step": 18938 }, { "epoch": 2.409235466225671, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6344707012176514, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8746421933174133, "num_tokens": 719211751.0, "step": 18939 }, { "epoch": 2.4093626765042617, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7897136211395264, "learning_rate": 1e-05, "loss": 0.3718, "mean_token_accuracy": 0.8969855308532715, "num_tokens": 719246812.0, "step": 18940 }, { "epoch": 2.4094898867828523, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.626495122909546, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8623256683349609, "num_tokens": 719290962.0, "step": 18941 }, { "epoch": 2.4096170970614423, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.79662823677063, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8793434500694275, "num_tokens": 719323506.0, "step": 18942 }, { "epoch": 2.4097443073400333, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.00034332275390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.727386713027954, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8724290728569031, "num_tokens": 719361435.0, "step": 18943 }, { "epoch": 2.4098715176186234, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.727215528488159, "learning_rate": 1e-05, "loss": 0.3872, "mean_token_accuracy": 0.8904308676719666, "num_tokens": 719405617.0, "step": 18944 }, { "epoch": 2.409998727897214, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.649472713470459, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8787221908569336, "num_tokens": 719450588.0, "step": 18945 }, { "epoch": 2.4101259381758044, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.723782777786255, "learning_rate": 1e-05, "loss": 0.3842, "mean_token_accuracy": 0.8880295753479004, "num_tokens": 719489028.0, "step": 18946 }, { "epoch": 2.410253148454395, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.675309658050537, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8751399517059326, "num_tokens": 719527921.0, "step": 18947 }, { "epoch": 2.4103803587329855, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.705202341079712, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8815929889678955, "num_tokens": 719572461.0, "step": 18948 }, { "epoch": 2.410507569011576, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.65940523147583, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8773943185806274, "num_tokens": 719613478.0, "step": 18949 }, { "epoch": 2.4106347792901666, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7212185859680176, "learning_rate": 1e-05, "loss": 0.3803, "mean_token_accuracy": 0.8922039270401001, "num_tokens": 719643665.0, "step": 18950 }, { "epoch": 2.410761989568757, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.675107479095459, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8704153895378113, "num_tokens": 719684168.0, "step": 18951 }, { "epoch": 2.4108891998473476, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7054426670074463, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8800946474075317, "num_tokens": 719722484.0, "step": 18952 }, { "epoch": 2.411016410125938, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6687705516815186, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8810200691223145, "num_tokens": 719756659.0, "step": 18953 }, { "epoch": 2.4111436204045287, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.725870132446289, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8824512958526611, "num_tokens": 719787302.0, "step": 18954 }, { "epoch": 2.411270830683119, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7474379539489746, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8830711841583252, "num_tokens": 719823359.0, "step": 18955 }, { "epoch": 2.4113980409617097, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.674579620361328, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8802233934402466, "num_tokens": 719861263.0, "step": 18956 }, { "epoch": 2.4115252512403003, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.687997579574585, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8839776515960693, "num_tokens": 719897987.0, "step": 18957 }, { "epoch": 2.411652461518891, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.67680287361145, "learning_rate": 1e-05, "loss": 0.3937, "mean_token_accuracy": 0.8872989416122437, "num_tokens": 719931067.0, "step": 18958 }, { "epoch": 2.4117796717974813, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.687438488006592, "learning_rate": 1e-05, "loss": 0.3795, "mean_token_accuracy": 0.8908337354660034, "num_tokens": 719971174.0, "step": 18959 }, { "epoch": 2.411906882076072, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.677687168121338, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.8863472938537598, "num_tokens": 720005781.0, "step": 18960 }, { "epoch": 2.4120340923546624, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6778194904327393, "learning_rate": 1e-05, "loss": 0.3827, "mean_token_accuracy": 0.8875211477279663, "num_tokens": 720040917.0, "step": 18961 }, { "epoch": 2.412161302633253, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.68404483795166, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8683900833129883, "num_tokens": 720080490.0, "step": 18962 }, { "epoch": 2.4122885129118434, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.675630569458008, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8808753490447998, "num_tokens": 720116461.0, "step": 18963 }, { "epoch": 2.412415723190434, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6597540378570557, "learning_rate": 1e-05, "loss": 0.3487, "mean_token_accuracy": 0.9003100991249084, "num_tokens": 720156756.0, "step": 18964 }, { "epoch": 2.4125429334690245, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6858723163604736, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8790735006332397, "num_tokens": 720195271.0, "step": 18965 }, { "epoch": 2.412670143747615, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6718642711639404, "learning_rate": 1e-05, "loss": 0.3859, "mean_token_accuracy": 0.890152096748352, "num_tokens": 720232159.0, "step": 18966 }, { "epoch": 2.412797354026205, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6939566135406494, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8868141174316406, "num_tokens": 720264591.0, "step": 18967 }, { "epoch": 2.412924564304796, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6636857986450195, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8826016187667847, "num_tokens": 720302957.0, "step": 18968 }, { "epoch": 2.413051774583386, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7176544666290283, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8727117776870728, "num_tokens": 720335679.0, "step": 18969 }, { "epoch": 2.4131789848619767, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6524600982666016, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8776285648345947, "num_tokens": 720377261.0, "step": 18970 }, { "epoch": 2.413306195140567, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7109131813049316, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8799976706504822, "num_tokens": 720415344.0, "step": 18971 }, { "epoch": 2.4134334054191577, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6709656715393066, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.87633216381073, "num_tokens": 720449133.0, "step": 18972 }, { "epoch": 2.4135606156977483, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.717836380004883, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8715764284133911, "num_tokens": 720484250.0, "step": 18973 }, { "epoch": 2.413687825976339, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6625654697418213, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8751702904701233, "num_tokens": 720520466.0, "step": 18974 }, { "epoch": 2.4138150362549293, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7118613719940186, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8823913931846619, "num_tokens": 720557013.0, "step": 18975 }, { "epoch": 2.41394224653352, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.654411554336548, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8837218880653381, "num_tokens": 720594957.0, "step": 18976 }, { "epoch": 2.4140694568121104, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.715482234954834, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8772886395454407, "num_tokens": 720630531.0, "step": 18977 }, { "epoch": 2.414196667090701, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.647714376449585, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.875406801700592, "num_tokens": 720673502.0, "step": 18978 }, { "epoch": 2.4143238773692914, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7076239585876465, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8802501559257507, "num_tokens": 720716765.0, "step": 18979 }, { "epoch": 2.414451087647882, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6522252559661865, "learning_rate": 1e-05, "loss": 0.3783, "mean_token_accuracy": 0.8920737504959106, "num_tokens": 720754711.0, "step": 18980 }, { "epoch": 2.4145782979264725, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7302231788635254, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8749353289604187, "num_tokens": 720789067.0, "step": 18981 }, { "epoch": 2.414705508205063, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.670808792114258, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8725327849388123, "num_tokens": 720821725.0, "step": 18982 }, { "epoch": 2.4148327184836536, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7228260040283203, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8850463032722473, "num_tokens": 720857587.0, "step": 18983 }, { "epoch": 2.414959928762244, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6311988830566406, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8844702243804932, "num_tokens": 720900481.0, "step": 18984 }, { "epoch": 2.4150871390408346, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7411751747131348, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8799422979354858, "num_tokens": 720933131.0, "step": 18985 }, { "epoch": 2.415214349319425, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6238105297088623, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8782892227172852, "num_tokens": 720972142.0, "step": 18986 }, { "epoch": 2.4153415595980157, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7397234439849854, "learning_rate": 1e-05, "loss": 0.3886, "mean_token_accuracy": 0.8899860382080078, "num_tokens": 721014251.0, "step": 18987 }, { "epoch": 2.415468769876606, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6310698986053467, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8701674342155457, "num_tokens": 721053327.0, "step": 18988 }, { "epoch": 2.4155959801551967, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.734297513961792, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8844231367111206, "num_tokens": 721092047.0, "step": 18989 }, { "epoch": 2.415723190433787, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6330721378326416, "learning_rate": 1e-05, "loss": 0.3587, "mean_token_accuracy": 0.8994643688201904, "num_tokens": 721125318.0, "step": 18990 }, { "epoch": 2.4158504007123778, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7422380447387695, "learning_rate": 1e-05, "loss": 0.383, "mean_token_accuracy": 0.8939033150672913, "num_tokens": 721162058.0, "step": 18991 }, { "epoch": 2.415977610990968, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6381685733795166, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8829249739646912, "num_tokens": 721199219.0, "step": 18992 }, { "epoch": 2.4161048212695584, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.736760377883911, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8758244514465332, "num_tokens": 721239696.0, "step": 18993 }, { "epoch": 2.416232031548149, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6346733570098877, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8728752136230469, "num_tokens": 721277660.0, "step": 18994 }, { "epoch": 2.4163592418267394, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7510826587677, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8791054487228394, "num_tokens": 721311805.0, "step": 18995 }, { "epoch": 2.41648645210533, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.623044967651367, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8607883453369141, "num_tokens": 721355968.0, "step": 18996 }, { "epoch": 2.4166136623839205, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.748124599456787, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8738170862197876, "num_tokens": 721397930.0, "step": 18997 }, { "epoch": 2.416740872662511, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6216366291046143, "learning_rate": 1e-05, "loss": 0.3815, "mean_token_accuracy": 0.8906208276748657, "num_tokens": 721432408.0, "step": 18998 }, { "epoch": 2.4168680829411016, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.750777244567871, "learning_rate": 1e-05, "loss": 0.3628, "mean_token_accuracy": 0.898351788520813, "num_tokens": 721468903.0, "step": 18999 }, { "epoch": 2.416995293219692, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6096463203430176, "learning_rate": 1e-05, "loss": 0.3718, "mean_token_accuracy": 0.8962457180023193, "num_tokens": 721509687.0, "step": 19000 }, { "epoch": 2.4171225034982826, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7763893604278564, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.873059868812561, "num_tokens": 721548130.0, "step": 19001 }, { "epoch": 2.417249713776873, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.615945816040039, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8868206739425659, "num_tokens": 721588504.0, "step": 19002 }, { "epoch": 2.4173769240554637, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.769040822982788, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8774442076683044, "num_tokens": 721632621.0, "step": 19003 }, { "epoch": 2.417504134334054, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6097395420074463, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8793089985847473, "num_tokens": 721670898.0, "step": 19004 }, { "epoch": 2.4176313446126447, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.773697853088379, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.888592004776001, "num_tokens": 721709012.0, "step": 19005 }, { "epoch": 2.4177585548912353, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6126954555511475, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8808705806732178, "num_tokens": 721747135.0, "step": 19006 }, { "epoch": 2.417885765169826, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7848503589630127, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8804067969322205, "num_tokens": 721784447.0, "step": 19007 }, { "epoch": 2.4180129754484163, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6314759254455566, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8706867694854736, "num_tokens": 721817853.0, "step": 19008 }, { "epoch": 2.418140185727007, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7901370525360107, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8770095109939575, "num_tokens": 721852537.0, "step": 19009 }, { "epoch": 2.4182673960055974, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6131112575531006, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8831591606140137, "num_tokens": 721892156.0, "step": 19010 }, { "epoch": 2.418394606284188, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.784600257873535, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8810880780220032, "num_tokens": 721929425.0, "step": 19011 }, { "epoch": 2.4185218165627784, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6707065105438232, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8725155591964722, "num_tokens": 721967492.0, "step": 19012 }, { "epoch": 2.418649026841369, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7600579261779785, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.876427173614502, "num_tokens": 722007754.0, "step": 19013 }, { "epoch": 2.4187762371199595, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.647078275680542, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8811256289482117, "num_tokens": 722043380.0, "step": 19014 }, { "epoch": 2.4189034473985496, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7473666667938232, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8762024641036987, "num_tokens": 722086200.0, "step": 19015 }, { "epoch": 2.4190306576771405, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.657092332839966, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8842000365257263, "num_tokens": 722122907.0, "step": 19016 }, { "epoch": 2.4191578679557306, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.741668701171875, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8841201663017273, "num_tokens": 722155951.0, "step": 19017 }, { "epoch": 2.419285078234321, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6514928340911865, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8764631748199463, "num_tokens": 722194479.0, "step": 19018 }, { "epoch": 2.4194122885129117, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7359323501586914, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8836156129837036, "num_tokens": 722228470.0, "step": 19019 }, { "epoch": 2.419539498791502, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6543846130371094, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8711227178573608, "num_tokens": 722265960.0, "step": 19020 }, { "epoch": 2.4196667090700927, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.780352830886841, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8751084804534912, "num_tokens": 722301377.0, "step": 19021 }, { "epoch": 2.4197939193486833, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6242141723632812, "learning_rate": 1e-05, "loss": 0.3983, "mean_token_accuracy": 0.8836791515350342, "num_tokens": 722338534.0, "step": 19022 }, { "epoch": 2.419921129627274, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.76469087600708, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8764920234680176, "num_tokens": 722377460.0, "step": 19023 }, { "epoch": 2.4200483399058643, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6293954849243164, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8710963726043701, "num_tokens": 722415167.0, "step": 19024 }, { "epoch": 2.420175550184455, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.772113084793091, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8670454025268555, "num_tokens": 722456114.0, "step": 19025 }, { "epoch": 2.4203027604630454, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6214048862457275, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8860366344451904, "num_tokens": 722496378.0, "step": 19026 }, { "epoch": 2.420429970741636, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7782187461853027, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8713772892951965, "num_tokens": 722534903.0, "step": 19027 }, { "epoch": 2.4205571810202264, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.618734836578369, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8754289746284485, "num_tokens": 722576649.0, "step": 19028 }, { "epoch": 2.420684391298817, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8008694648742676, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8721179962158203, "num_tokens": 722618555.0, "step": 19029 }, { "epoch": 2.4208116015774075, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.614269733428955, "learning_rate": 1e-05, "loss": 0.3612, "mean_token_accuracy": 0.8979638814926147, "num_tokens": 722652045.0, "step": 19030 }, { "epoch": 2.420938811855998, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.805042266845703, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8784564733505249, "num_tokens": 722686987.0, "step": 19031 }, { "epoch": 2.4210660221345885, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.624798059463501, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8742780685424805, "num_tokens": 722724161.0, "step": 19032 }, { "epoch": 2.421193232413179, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7812280654907227, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.886982798576355, "num_tokens": 722766473.0, "step": 19033 }, { "epoch": 2.4213204426917696, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.610201597213745, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8810960650444031, "num_tokens": 722809499.0, "step": 19034 }, { "epoch": 2.42144765297036, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.789480686187744, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8846835494041443, "num_tokens": 722847295.0, "step": 19035 }, { "epoch": 2.4215748632489507, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6183080673217773, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8810952305793762, "num_tokens": 722884313.0, "step": 19036 }, { "epoch": 2.421702073527541, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.788968801498413, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8840804100036621, "num_tokens": 722924530.0, "step": 19037 }, { "epoch": 2.4218292838061317, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6304755210876465, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8695856332778931, "num_tokens": 722962968.0, "step": 19038 }, { "epoch": 2.4219564940847222, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.792179822921753, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.888941764831543, "num_tokens": 722997801.0, "step": 19039 }, { "epoch": 2.4220837043633123, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6286568641662598, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8692656755447388, "num_tokens": 723037724.0, "step": 19040 }, { "epoch": 2.4222109146419033, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7892470359802246, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8712657690048218, "num_tokens": 723078199.0, "step": 19041 }, { "epoch": 2.4223381249204934, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6545331478118896, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8720840215682983, "num_tokens": 723110894.0, "step": 19042 }, { "epoch": 2.422465335199084, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7726638317108154, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8695988655090332, "num_tokens": 723150688.0, "step": 19043 }, { "epoch": 2.4225925454776744, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.637488842010498, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8785341382026672, "num_tokens": 723189863.0, "step": 19044 }, { "epoch": 2.422719755756265, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7602481842041016, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8700418472290039, "num_tokens": 723232333.0, "step": 19045 }, { "epoch": 2.4228469660348555, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6515376567840576, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.866671085357666, "num_tokens": 723267469.0, "step": 19046 }, { "epoch": 2.422974176313446, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.751620292663574, "learning_rate": 1e-05, "loss": 0.3799, "mean_token_accuracy": 0.8922892808914185, "num_tokens": 723304367.0, "step": 19047 }, { "epoch": 2.4231013865920366, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6293740272521973, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.87635338306427, "num_tokens": 723343247.0, "step": 19048 }, { "epoch": 2.423228596870627, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.75209903717041, "learning_rate": 1e-05, "loss": 0.3883, "mean_token_accuracy": 0.8908429145812988, "num_tokens": 723378022.0, "step": 19049 }, { "epoch": 2.4233558071492176, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.650692939758301, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8822999000549316, "num_tokens": 723409827.0, "step": 19050 }, { "epoch": 2.423483017427808, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.742527723312378, "learning_rate": 1e-05, "loss": 0.393, "mean_token_accuracy": 0.8887628316879272, "num_tokens": 723447192.0, "step": 19051 }, { "epoch": 2.4236102277063987, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6265599727630615, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8841335773468018, "num_tokens": 723492123.0, "step": 19052 }, { "epoch": 2.423737437984989, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.753262996673584, "learning_rate": 1e-05, "loss": 0.3705, "mean_token_accuracy": 0.8953981399536133, "num_tokens": 723525680.0, "step": 19053 }, { "epoch": 2.4238646482635797, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6355676651000977, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8779193162918091, "num_tokens": 723565667.0, "step": 19054 }, { "epoch": 2.4239918585421703, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7672221660614014, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8751457929611206, "num_tokens": 723599486.0, "step": 19055 }, { "epoch": 2.424119068820761, "ewc_loss": 0.07568359375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.635960817337036, "learning_rate": 1e-05, "loss": 0.3932, "mean_token_accuracy": 0.8864028453826904, "num_tokens": 723631992.0, "step": 19056 }, { "epoch": 2.4242462790993513, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7594175338745117, "learning_rate": 1e-05, "loss": 0.384, "mean_token_accuracy": 0.8882342576980591, "num_tokens": 723667654.0, "step": 19057 }, { "epoch": 2.424373489377942, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6257777214050293, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.8851860165596008, "num_tokens": 723705368.0, "step": 19058 }, { "epoch": 2.4245006996565324, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.9500253200531006, "learning_rate": 1e-05, "loss": 0.3961, "mean_token_accuracy": 0.8824491500854492, "num_tokens": 723744973.0, "step": 19059 }, { "epoch": 2.424627909935123, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6145811080932617, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8705717325210571, "num_tokens": 723778868.0, "step": 19060 }, { "epoch": 2.4247551202137134, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8313941955566406, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8829756379127502, "num_tokens": 723822413.0, "step": 19061 }, { "epoch": 2.424882330492304, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.588113307952881, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8735212683677673, "num_tokens": 723865826.0, "step": 19062 }, { "epoch": 2.4250095407708945, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.839897155761719e-05, "grad_norm": 2.8555686473846436, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8861916065216064, "num_tokens": 723901839.0, "step": 19063 }, { "epoch": 2.425136751049485, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.5933828353881836, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8874461650848389, "num_tokens": 723944883.0, "step": 19064 }, { "epoch": 2.425263961328075, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.857280969619751, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8864922523498535, "num_tokens": 723985084.0, "step": 19065 }, { "epoch": 2.425391171606666, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6039927005767822, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8777352571487427, "num_tokens": 724027919.0, "step": 19066 }, { "epoch": 2.425518381885256, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.9114227294921875e-05, "grad_norm": 2.874547243118286, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8857054114341736, "num_tokens": 724064593.0, "step": 19067 }, { "epoch": 2.4256455921638467, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6277689933776855, "learning_rate": 1e-05, "loss": 0.3878, "mean_token_accuracy": 0.8899379372596741, "num_tokens": 724095967.0, "step": 19068 }, { "epoch": 2.425772802442437, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.864144802093506, "learning_rate": 1e-05, "loss": 0.3806, "mean_token_accuracy": 0.8934985995292664, "num_tokens": 724128420.0, "step": 19069 }, { "epoch": 2.4259000127210277, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6392180919647217, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8714103698730469, "num_tokens": 724165073.0, "step": 19070 }, { "epoch": 2.4260272229996183, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8511359691619873, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.878095805644989, "num_tokens": 724201377.0, "step": 19071 }, { "epoch": 2.426154433278209, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6233980655670166, "learning_rate": 1e-05, "loss": 0.3764, "mean_token_accuracy": 0.8951694369316101, "num_tokens": 724241783.0, "step": 19072 }, { "epoch": 2.4262816435567993, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8448290824890137, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8796991109848022, "num_tokens": 724277903.0, "step": 19073 }, { "epoch": 2.42640885383539, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6190950870513916, "learning_rate": 1e-05, "loss": 0.3879, "mean_token_accuracy": 0.8881428241729736, "num_tokens": 724318834.0, "step": 19074 }, { "epoch": 2.4265360641139804, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.839897155761719e-05, "grad_norm": 2.846712589263916, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8844029307365417, "num_tokens": 724356406.0, "step": 19075 }, { "epoch": 2.426663274392571, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.620906114578247, "learning_rate": 1e-05, "loss": 0.3868, "mean_token_accuracy": 0.8903800845146179, "num_tokens": 724394774.0, "step": 19076 }, { "epoch": 2.4267904846711614, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.849557876586914, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8739044070243835, "num_tokens": 724431825.0, "step": 19077 }, { "epoch": 2.426917694949752, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6462788581848145, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8805802464485168, "num_tokens": 724466402.0, "step": 19078 }, { "epoch": 2.4270449052283425, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8281476497650146, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8708941340446472, "num_tokens": 724503353.0, "step": 19079 }, { "epoch": 2.427172115506933, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6426868438720703, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8868014216423035, "num_tokens": 724540348.0, "step": 19080 }, { "epoch": 2.4272993257855235, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.828552484512329, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8773982524871826, "num_tokens": 724577999.0, "step": 19081 }, { "epoch": 2.427426536064114, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.625596284866333, "learning_rate": 1e-05, "loss": 0.3617, "mean_token_accuracy": 0.8949690461158752, "num_tokens": 724614754.0, "step": 19082 }, { "epoch": 2.4275537463427046, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.819037914276123, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.870930552482605, "num_tokens": 724653232.0, "step": 19083 }, { "epoch": 2.427680956621295, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6514787673950195, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.871871829032898, "num_tokens": 724686461.0, "step": 19084 }, { "epoch": 2.4278081668998857, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.786679267883301, "learning_rate": 1e-05, "loss": 0.3915, "mean_token_accuracy": 0.8911653757095337, "num_tokens": 724727330.0, "step": 19085 }, { "epoch": 2.427935377178476, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6514832973480225, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8723092079162598, "num_tokens": 724766509.0, "step": 19086 }, { "epoch": 2.4280625874570667, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.785276412963867, "learning_rate": 1e-05, "loss": 0.3746, "mean_token_accuracy": 0.8956997394561768, "num_tokens": 724795128.0, "step": 19087 }, { "epoch": 2.428189797735657, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.648193597793579, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8748047947883606, "num_tokens": 724833941.0, "step": 19088 }, { "epoch": 2.4283170080142478, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.773078680038452, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8862216472625732, "num_tokens": 724870429.0, "step": 19089 }, { "epoch": 2.428444218292838, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.645273208618164, "learning_rate": 1e-05, "loss": 0.4025, "mean_token_accuracy": 0.8829350471496582, "num_tokens": 724905574.0, "step": 19090 }, { "epoch": 2.4285714285714284, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7630460262298584, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8874548673629761, "num_tokens": 724949845.0, "step": 19091 }, { "epoch": 2.428698638850019, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6420984268188477, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8733742833137512, "num_tokens": 724987586.0, "step": 19092 }, { "epoch": 2.4288258491286094, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.752873420715332, "learning_rate": 1e-05, "loss": 0.3891, "mean_token_accuracy": 0.886659562587738, "num_tokens": 725026432.0, "step": 19093 }, { "epoch": 2.4289530594072, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6403520107269287, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8604440689086914, "num_tokens": 725070627.0, "step": 19094 }, { "epoch": 2.4290802696857905, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.762279748916626, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8813592791557312, "num_tokens": 725106005.0, "step": 19095 }, { "epoch": 2.429207479964381, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6281628608703613, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8866647481918335, "num_tokens": 725149667.0, "step": 19096 }, { "epoch": 2.4293346902429716, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.773912191390991, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8839508295059204, "num_tokens": 725185526.0, "step": 19097 }, { "epoch": 2.429461900521562, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6255178451538086, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8846109509468079, "num_tokens": 725221789.0, "step": 19098 }, { "epoch": 2.4295891108001526, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7773947715759277, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8734373450279236, "num_tokens": 725258275.0, "step": 19099 }, { "epoch": 2.429716321078743, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.632328510284424, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8659051656723022, "num_tokens": 725297753.0, "step": 19100 }, { "epoch": 2.4298435313573337, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.77154803276062, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8874461650848389, "num_tokens": 725331978.0, "step": 19101 }, { "epoch": 2.429970741635924, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.620034694671631, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8849263191223145, "num_tokens": 725373876.0, "step": 19102 }, { "epoch": 2.4300979519145147, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7835495471954346, "learning_rate": 1e-05, "loss": 0.3877, "mean_token_accuracy": 0.8877426385879517, "num_tokens": 725411280.0, "step": 19103 }, { "epoch": 2.4302251621931052, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6283528804779053, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8681797385215759, "num_tokens": 725452050.0, "step": 19104 }, { "epoch": 2.4303523724716958, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.776535987854004, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8872492909431458, "num_tokens": 725489922.0, "step": 19105 }, { "epoch": 2.4304795827502863, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6290409564971924, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8772258758544922, "num_tokens": 725527871.0, "step": 19106 }, { "epoch": 2.430606793028877, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.774113416671753, "learning_rate": 1e-05, "loss": 0.3998, "mean_token_accuracy": 0.8906843662261963, "num_tokens": 725565045.0, "step": 19107 }, { "epoch": 2.4307340033074674, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6381356716156006, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.8916904926300049, "num_tokens": 725596604.0, "step": 19108 }, { "epoch": 2.430861213586058, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7817442417144775, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.880050539970398, "num_tokens": 725632299.0, "step": 19109 }, { "epoch": 2.4309884238646484, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6210665702819824, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8797957897186279, "num_tokens": 725673778.0, "step": 19110 }, { "epoch": 2.431115634143239, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7834339141845703, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8690011501312256, "num_tokens": 725715087.0, "step": 19111 }, { "epoch": 2.4312428444218295, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.615061044692993, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8831276893615723, "num_tokens": 725754047.0, "step": 19112 }, { "epoch": 2.4313700547004196, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7906851768493652, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8701427578926086, "num_tokens": 725793168.0, "step": 19113 }, { "epoch": 2.4314972649790105, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.605048179626465, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8840482831001282, "num_tokens": 725834713.0, "step": 19114 }, { "epoch": 2.4316244752576006, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.809818744659424, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8819608688354492, "num_tokens": 725872759.0, "step": 19115 }, { "epoch": 2.431751685536191, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6015024185180664, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8849568963050842, "num_tokens": 725916192.0, "step": 19116 }, { "epoch": 2.4318788958147817, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8115134239196777, "learning_rate": 1e-05, "loss": 0.3667, "mean_token_accuracy": 0.896699070930481, "num_tokens": 725949838.0, "step": 19117 }, { "epoch": 2.432006106093372, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.627286195755005, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8747684955596924, "num_tokens": 725986538.0, "step": 19118 }, { "epoch": 2.4321333163719627, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.796339988708496, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8850406408309937, "num_tokens": 726029007.0, "step": 19119 }, { "epoch": 2.4322605266505533, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.618330717086792, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8818724155426025, "num_tokens": 726066745.0, "step": 19120 }, { "epoch": 2.432387736929144, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8068251609802246, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8747389912605286, "num_tokens": 726106768.0, "step": 19121 }, { "epoch": 2.4325149472077343, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6287052631378174, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8811005353927612, "num_tokens": 726144879.0, "step": 19122 }, { "epoch": 2.432642157486325, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.807084798812866, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8809953331947327, "num_tokens": 726181278.0, "step": 19123 }, { "epoch": 2.4327693677649154, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6152520179748535, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.882951021194458, "num_tokens": 726226090.0, "step": 19124 }, { "epoch": 2.432896578043506, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8146440982818604, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8786753416061401, "num_tokens": 726261567.0, "step": 19125 }, { "epoch": 2.4330237883220964, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6223762035369873, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8808410167694092, "num_tokens": 726300527.0, "step": 19126 }, { "epoch": 2.433150998600687, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8046867847442627, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.886182427406311, "num_tokens": 726341240.0, "step": 19127 }, { "epoch": 2.4332782088792775, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6386470794677734, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8718358874320984, "num_tokens": 726375286.0, "step": 19128 }, { "epoch": 2.433405419157868, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7837469577789307, "learning_rate": 1e-05, "loss": 0.3805, "mean_token_accuracy": 0.8941231369972229, "num_tokens": 726414810.0, "step": 19129 }, { "epoch": 2.4335326294364585, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.630178213119507, "learning_rate": 1e-05, "loss": 0.3703, "mean_token_accuracy": 0.8972707986831665, "num_tokens": 726451204.0, "step": 19130 }, { "epoch": 2.433659839715049, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7768211364746094, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8808965086936951, "num_tokens": 726491107.0, "step": 19131 }, { "epoch": 2.4337870499936396, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.630598545074463, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8782581686973572, "num_tokens": 726535297.0, "step": 19132 }, { "epoch": 2.43391426027223, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.780635356903076, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.871309757232666, "num_tokens": 726574841.0, "step": 19133 }, { "epoch": 2.4340414705508207, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6343300342559814, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8752011060714722, "num_tokens": 726611101.0, "step": 19134 }, { "epoch": 2.434168680829411, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7706809043884277, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8755744695663452, "num_tokens": 726654323.0, "step": 19135 }, { "epoch": 2.4342958911080017, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6355929374694824, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8830749988555908, "num_tokens": 726691504.0, "step": 19136 }, { "epoch": 2.4344231013865922, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7562074661254883, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8759948015213013, "num_tokens": 726737257.0, "step": 19137 }, { "epoch": 2.4345503116651823, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6238656044006348, "learning_rate": 1e-05, "loss": 0.3883, "mean_token_accuracy": 0.8902289271354675, "num_tokens": 726779917.0, "step": 19138 }, { "epoch": 2.4346775219437733, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7731380462646484, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.883270800113678, "num_tokens": 726823130.0, "step": 19139 }, { "epoch": 2.4348047322223634, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.616048574447632, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8837000131607056, "num_tokens": 726861787.0, "step": 19140 }, { "epoch": 2.434931942500954, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7770283222198486, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8656201362609863, "num_tokens": 726904413.0, "step": 19141 }, { "epoch": 2.4350591527795444, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6104555130004883, "learning_rate": 1e-05, "loss": 0.3897, "mean_token_accuracy": 0.8911834955215454, "num_tokens": 726946984.0, "step": 19142 }, { "epoch": 2.435186363058135, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7971251010894775, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8797667026519775, "num_tokens": 726982587.0, "step": 19143 }, { "epoch": 2.4353135733367255, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6211183071136475, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8846028447151184, "num_tokens": 727017255.0, "step": 19144 }, { "epoch": 2.435440783615316, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.806692123413086, "learning_rate": 1e-05, "loss": 0.3924, "mean_token_accuracy": 0.8870805501937866, "num_tokens": 727046147.0, "step": 19145 }, { "epoch": 2.4355679938939065, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.615332841873169, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8799343109130859, "num_tokens": 727085982.0, "step": 19146 }, { "epoch": 2.435695204172497, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.794461965560913, "learning_rate": 1e-05, "loss": 0.3843, "mean_token_accuracy": 0.8936054706573486, "num_tokens": 727123991.0, "step": 19147 }, { "epoch": 2.4358224144510876, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.623537540435791, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8641543388366699, "num_tokens": 727162155.0, "step": 19148 }, { "epoch": 2.435949624729678, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.801130533218384, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8705788850784302, "num_tokens": 727201234.0, "step": 19149 }, { "epoch": 2.4360768350082687, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6158554553985596, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8792878985404968, "num_tokens": 727243429.0, "step": 19150 }, { "epoch": 2.436204045286859, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.802385091781616, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.881902813911438, "num_tokens": 727279705.0, "step": 19151 }, { "epoch": 2.4363312555654497, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6204988956451416, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8862793445587158, "num_tokens": 727315503.0, "step": 19152 }, { "epoch": 2.4364584658440402, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7985546588897705, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8773519396781921, "num_tokens": 727355272.0, "step": 19153 }, { "epoch": 2.4365856761226308, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6326303482055664, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8669417500495911, "num_tokens": 727391058.0, "step": 19154 }, { "epoch": 2.4367128864012213, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003452301025390625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7899088859558105, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8740088939666748, "num_tokens": 727433011.0, "step": 19155 }, { "epoch": 2.436840096679812, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6468465328216553, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8874164819717407, "num_tokens": 727462282.0, "step": 19156 }, { "epoch": 2.4369673069584024, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.791508197784424, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8774731159210205, "num_tokens": 727501636.0, "step": 19157 }, { "epoch": 2.437094517236993, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6205081939697266, "learning_rate": 1e-05, "loss": 0.3562, "mean_token_accuracy": 0.8977433443069458, "num_tokens": 727536828.0, "step": 19158 }, { "epoch": 2.4372217275155834, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.79994535446167, "learning_rate": 1e-05, "loss": 0.4902, "mean_token_accuracy": 0.8546103239059448, "num_tokens": 727574858.0, "step": 19159 }, { "epoch": 2.437348937794174, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.601478099822998, "learning_rate": 1e-05, "loss": 0.3731, "mean_token_accuracy": 0.8963642120361328, "num_tokens": 727618168.0, "step": 19160 }, { "epoch": 2.4374761480727645, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8177807331085205, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8696686029434204, "num_tokens": 727660205.0, "step": 19161 }, { "epoch": 2.437603358351355, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.5967018604278564, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.8870981335639954, "num_tokens": 727701980.0, "step": 19162 }, { "epoch": 2.437730568629945, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8354532718658447, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.868786096572876, "num_tokens": 727743766.0, "step": 19163 }, { "epoch": 2.437857778908536, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6041576862335205, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8826072216033936, "num_tokens": 727785504.0, "step": 19164 }, { "epoch": 2.437984989187126, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.828291893005371, "learning_rate": 1e-05, "loss": 0.4025, "mean_token_accuracy": 0.8872954249382019, "num_tokens": 727831296.0, "step": 19165 }, { "epoch": 2.4381121994657167, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6147260665893555, "learning_rate": 1e-05, "loss": 0.3879, "mean_token_accuracy": 0.8868438601493835, "num_tokens": 727867343.0, "step": 19166 }, { "epoch": 2.438239409744307, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.844329833984375, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8801279067993164, "num_tokens": 727903856.0, "step": 19167 }, { "epoch": 2.4383666200228977, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.628109931945801, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8777198195457458, "num_tokens": 727938385.0, "step": 19168 }, { "epoch": 2.4384938303014883, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.859264612197876, "learning_rate": 1e-05, "loss": 0.3896, "mean_token_accuracy": 0.8890061974525452, "num_tokens": 727965566.0, "step": 19169 }, { "epoch": 2.438621040580079, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6206326484680176, "learning_rate": 1e-05, "loss": 0.3842, "mean_token_accuracy": 0.8952968120574951, "num_tokens": 728002203.0, "step": 19170 }, { "epoch": 2.4387482508586693, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.8349685668945312, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.8869483470916748, "num_tokens": 728037055.0, "step": 19171 }, { "epoch": 2.43887546113726, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.627492666244507, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8738389015197754, "num_tokens": 728077226.0, "step": 19172 }, { "epoch": 2.4390026714158504, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8397889137268066, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8850313425064087, "num_tokens": 728111396.0, "step": 19173 }, { "epoch": 2.439129881694441, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.63260817527771, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8781843185424805, "num_tokens": 728149609.0, "step": 19174 }, { "epoch": 2.4392570919730314, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.835773468017578, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8808354139328003, "num_tokens": 728186668.0, "step": 19175 }, { "epoch": 2.439384302251622, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6447970867156982, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8648888468742371, "num_tokens": 728226504.0, "step": 19176 }, { "epoch": 2.4395115125302125, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.817389726638794, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8782202005386353, "num_tokens": 728264648.0, "step": 19177 }, { "epoch": 2.439638722808803, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.643096446990967, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8826274871826172, "num_tokens": 728300292.0, "step": 19178 }, { "epoch": 2.4397659330873935, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8193843364715576, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8784263134002686, "num_tokens": 728336169.0, "step": 19179 }, { "epoch": 2.439893143365984, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6304707527160645, "learning_rate": 1e-05, "loss": 0.3986, "mean_token_accuracy": 0.885563850402832, "num_tokens": 728380160.0, "step": 19180 }, { "epoch": 2.4400203536445746, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.803558826446533, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8817635178565979, "num_tokens": 728417978.0, "step": 19181 }, { "epoch": 2.440147563923165, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.630857467651367, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8844283223152161, "num_tokens": 728455264.0, "step": 19182 }, { "epoch": 2.4402747742017556, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.799278974533081, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8694887161254883, "num_tokens": 728498880.0, "step": 19183 }, { "epoch": 2.440401984480346, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6574151515960693, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8767066597938538, "num_tokens": 728531925.0, "step": 19184 }, { "epoch": 2.4405291947589367, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.790599822998047, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8765813112258911, "num_tokens": 728570573.0, "step": 19185 }, { "epoch": 2.440656405037527, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.639700412750244, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8816800117492676, "num_tokens": 728610877.0, "step": 19186 }, { "epoch": 2.4407836153161178, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7723019123077393, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8797447085380554, "num_tokens": 728650963.0, "step": 19187 }, { "epoch": 2.440910825594708, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6376681327819824, "learning_rate": 1e-05, "loss": 0.3855, "mean_token_accuracy": 0.8902202248573303, "num_tokens": 728688699.0, "step": 19188 }, { "epoch": 2.4410380358732984, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7734451293945312, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8690856695175171, "num_tokens": 728730142.0, "step": 19189 }, { "epoch": 2.441165246151889, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.648639678955078, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8730928897857666, "num_tokens": 728767808.0, "step": 19190 }, { "epoch": 2.4412924564304794, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.764011859893799, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8666171431541443, "num_tokens": 728816171.0, "step": 19191 }, { "epoch": 2.44141966670907, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.654914617538452, "learning_rate": 1e-05, "loss": 0.4726, "mean_token_accuracy": 0.8651797771453857, "num_tokens": 728858350.0, "step": 19192 }, { "epoch": 2.4415468769876605, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7563796043395996, "learning_rate": 1e-05, "loss": 0.3773, "mean_token_accuracy": 0.8933873176574707, "num_tokens": 728894408.0, "step": 19193 }, { "epoch": 2.441674087266251, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6553871631622314, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8607540130615234, "num_tokens": 728936563.0, "step": 19194 }, { "epoch": 2.4418012975448415, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.8053321838378906, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8721874356269836, "num_tokens": 728978796.0, "step": 19195 }, { "epoch": 2.441928507823432, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.622466564178467, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8872459530830383, "num_tokens": 729018524.0, "step": 19196 }, { "epoch": 2.4420557181020226, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.793550729751587, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8776575326919556, "num_tokens": 729055336.0, "step": 19197 }, { "epoch": 2.442182928380613, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.618246555328369, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8712803721427917, "num_tokens": 729095505.0, "step": 19198 }, { "epoch": 2.4423101386592037, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.792612314224243, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8772072792053223, "num_tokens": 729138384.0, "step": 19199 }, { "epoch": 2.442437348937794, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.625385046005249, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8592758178710938, "num_tokens": 729179942.0, "step": 19200 }, { "epoch": 2.4425645592163847, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.793008327484131, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8622669577598572, "num_tokens": 729223051.0, "step": 19201 }, { "epoch": 2.4426917694949752, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.622856855392456, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8826630115509033, "num_tokens": 729259368.0, "step": 19202 }, { "epoch": 2.4428189797735658, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8127024173736572, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8745896816253662, "num_tokens": 729295148.0, "step": 19203 }, { "epoch": 2.4429461900521563, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6289255619049072, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8715142607688904, "num_tokens": 729330649.0, "step": 19204 }, { "epoch": 2.443073400330747, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.790482759475708, "learning_rate": 1e-05, "loss": 0.4021, "mean_token_accuracy": 0.8867651224136353, "num_tokens": 729369260.0, "step": 19205 }, { "epoch": 2.4432006106093374, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.629457950592041, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8800917267799377, "num_tokens": 729406029.0, "step": 19206 }, { "epoch": 2.443327820887928, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.791175603866577, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8759068250656128, "num_tokens": 729444102.0, "step": 19207 }, { "epoch": 2.4434550311665184, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.625495195388794, "learning_rate": 1e-05, "loss": 0.3478, "mean_token_accuracy": 0.9022105932235718, "num_tokens": 729477121.0, "step": 19208 }, { "epoch": 2.443582241445109, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.809133529663086, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8696606159210205, "num_tokens": 729511682.0, "step": 19209 }, { "epoch": 2.4437094517236995, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.637058734893799, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8856879472732544, "num_tokens": 729545675.0, "step": 19210 }, { "epoch": 2.4438366620022896, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7900519371032715, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8791760802268982, "num_tokens": 729585323.0, "step": 19211 }, { "epoch": 2.4439638722808805, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.635219097137451, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8739222288131714, "num_tokens": 729623881.0, "step": 19212 }, { "epoch": 2.4440910825594706, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.790989637374878, "learning_rate": 1e-05, "loss": 0.3803, "mean_token_accuracy": 0.8941279649734497, "num_tokens": 729655486.0, "step": 19213 }, { "epoch": 2.444218292838061, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.628464698791504, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8803569078445435, "num_tokens": 729693606.0, "step": 19214 }, { "epoch": 2.4443455031166517, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7941248416900635, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8707321286201477, "num_tokens": 729729467.0, "step": 19215 }, { "epoch": 2.444472713395242, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6260592937469482, "learning_rate": 1e-05, "loss": 0.3889, "mean_token_accuracy": 0.8888761401176453, "num_tokens": 729767264.0, "step": 19216 }, { "epoch": 2.4445999236738327, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.78901743888855, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8830129504203796, "num_tokens": 729807352.0, "step": 19217 }, { "epoch": 2.4447271339524232, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6424202919006348, "learning_rate": 1e-05, "loss": 0.4781, "mean_token_accuracy": 0.8615367412567139, "num_tokens": 729847489.0, "step": 19218 }, { "epoch": 2.4448543442310138, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7683663368225098, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8860669732093811, "num_tokens": 729889675.0, "step": 19219 }, { "epoch": 2.4449815545096043, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6394636631011963, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8782333731651306, "num_tokens": 729926562.0, "step": 19220 }, { "epoch": 2.445108764788195, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.785248041152954, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8656134605407715, "num_tokens": 729964533.0, "step": 19221 }, { "epoch": 2.4452359750667854, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.628506898880005, "learning_rate": 1e-05, "loss": 0.3887, "mean_token_accuracy": 0.8905249238014221, "num_tokens": 730004206.0, "step": 19222 }, { "epoch": 2.445363185345376, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.779200553894043, "learning_rate": 1e-05, "loss": 0.4411, "mean_token_accuracy": 0.8734771013259888, "num_tokens": 730046779.0, "step": 19223 }, { "epoch": 2.4454903956239664, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.63226318359375, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8752642273902893, "num_tokens": 730085882.0, "step": 19224 }, { "epoch": 2.445617605902557, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.779156446456909, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8772364258766174, "num_tokens": 730126072.0, "step": 19225 }, { "epoch": 2.4457448161811475, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.637033700942993, "learning_rate": 1e-05, "loss": 0.3718, "mean_token_accuracy": 0.8969799280166626, "num_tokens": 730160038.0, "step": 19226 }, { "epoch": 2.445872026459738, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7886440753936768, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8814976215362549, "num_tokens": 730195701.0, "step": 19227 }, { "epoch": 2.4459992367383285, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6186516284942627, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8860891461372375, "num_tokens": 730233938.0, "step": 19228 }, { "epoch": 2.446126447016919, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7803618907928467, "learning_rate": 1e-05, "loss": 0.3738, "mean_token_accuracy": 0.8950585126876831, "num_tokens": 730272333.0, "step": 19229 }, { "epoch": 2.4462536572955096, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6209611892700195, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8697664737701416, "num_tokens": 730312967.0, "step": 19230 }, { "epoch": 2.4463808675741, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7899467945098877, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8817673921585083, "num_tokens": 730345525.0, "step": 19231 }, { "epoch": 2.4465080778526906, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6369848251342773, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8837034702301025, "num_tokens": 730384329.0, "step": 19232 }, { "epoch": 2.446635288131281, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.769092559814453, "learning_rate": 1e-05, "loss": 0.371, "mean_token_accuracy": 0.8958824276924133, "num_tokens": 730423139.0, "step": 19233 }, { "epoch": 2.4467624984098717, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.629422426223755, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.8801020979881287, "num_tokens": 730463989.0, "step": 19234 }, { "epoch": 2.4468897086884622, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7813079357147217, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8732216954231262, "num_tokens": 730500854.0, "step": 19235 }, { "epoch": 2.4470169189670523, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6187796592712402, "learning_rate": 1e-05, "loss": 0.3674, "mean_token_accuracy": 0.8974270224571228, "num_tokens": 730537742.0, "step": 19236 }, { "epoch": 2.4471441292456433, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.794867992401123, "learning_rate": 1e-05, "loss": 0.3638, "mean_token_accuracy": 0.8981807827949524, "num_tokens": 730573090.0, "step": 19237 }, { "epoch": 2.4472713395242334, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.615879774093628, "learning_rate": 1e-05, "loss": 0.3696, "mean_token_accuracy": 0.8973168134689331, "num_tokens": 730611865.0, "step": 19238 }, { "epoch": 2.447398549802824, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8157284259796143, "learning_rate": 1e-05, "loss": 0.4836, "mean_token_accuracy": 0.8608002066612244, "num_tokens": 730646831.0, "step": 19239 }, { "epoch": 2.4475257600814144, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6078758239746094, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8839462995529175, "num_tokens": 730689789.0, "step": 19240 }, { "epoch": 2.447652970360005, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8302783966064453, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8833168745040894, "num_tokens": 730720289.0, "step": 19241 }, { "epoch": 2.4477801806385955, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6056299209594727, "learning_rate": 1e-05, "loss": 0.3928, "mean_token_accuracy": 0.8868800401687622, "num_tokens": 730757499.0, "step": 19242 }, { "epoch": 2.447907390917186, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8197181224823, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8842142820358276, "num_tokens": 730799897.0, "step": 19243 }, { "epoch": 2.4480346011957765, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6077773571014404, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8867300748825073, "num_tokens": 730837709.0, "step": 19244 }, { "epoch": 2.448161811474367, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8252251148223877, "learning_rate": 1e-05, "loss": 0.3659, "mean_token_accuracy": 0.8985903263092041, "num_tokens": 730874128.0, "step": 19245 }, { "epoch": 2.4482890217529576, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6164543628692627, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8754360675811768, "num_tokens": 730917027.0, "step": 19246 }, { "epoch": 2.448416232031548, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.832942008972168, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8686364889144897, "num_tokens": 730955107.0, "step": 19247 }, { "epoch": 2.4485434423101387, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6268036365509033, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8692541718482971, "num_tokens": 730995807.0, "step": 19248 }, { "epoch": 2.448670652588729, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.819481134414673, "learning_rate": 1e-05, "loss": 0.3831, "mean_token_accuracy": 0.8929775953292847, "num_tokens": 731034935.0, "step": 19249 }, { "epoch": 2.4487978628673197, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6378848552703857, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8773691058158875, "num_tokens": 731070934.0, "step": 19250 }, { "epoch": 2.4489250731459102, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8002874851226807, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8886427283287048, "num_tokens": 731112625.0, "step": 19251 }, { "epoch": 2.4490522834245008, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6350417137145996, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8745253682136536, "num_tokens": 731151011.0, "step": 19252 }, { "epoch": 2.4491794937030913, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7880806922912598, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8826184868812561, "num_tokens": 731190885.0, "step": 19253 }, { "epoch": 2.449306703981682, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6374151706695557, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.8822874426841736, "num_tokens": 731229056.0, "step": 19254 }, { "epoch": 2.4494339142602723, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.78779673576355, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8802872896194458, "num_tokens": 731268098.0, "step": 19255 }, { "epoch": 2.449561124538863, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.646544933319092, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8664072751998901, "num_tokens": 731307809.0, "step": 19256 }, { "epoch": 2.4496883348174534, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7870309352874756, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8732680082321167, "num_tokens": 731343904.0, "step": 19257 }, { "epoch": 2.449815545096044, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6262481212615967, "learning_rate": 1e-05, "loss": 0.3746, "mean_token_accuracy": 0.8945425748825073, "num_tokens": 731383120.0, "step": 19258 }, { "epoch": 2.4499427553746345, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7840538024902344, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.882278561592102, "num_tokens": 731423020.0, "step": 19259 }, { "epoch": 2.450069965653225, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.625239610671997, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.87678062915802, "num_tokens": 731463522.0, "step": 19260 }, { "epoch": 2.450197175931815, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.79213547706604, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8706933259963989, "num_tokens": 731502327.0, "step": 19261 }, { "epoch": 2.450324386210406, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.619593381881714, "learning_rate": 1e-05, "loss": 0.3831, "mean_token_accuracy": 0.8910645246505737, "num_tokens": 731545253.0, "step": 19262 }, { "epoch": 2.450451596488996, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.799800157546997, "learning_rate": 1e-05, "loss": 0.3893, "mean_token_accuracy": 0.8874070644378662, "num_tokens": 731575388.0, "step": 19263 }, { "epoch": 2.4505788067675867, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6176559925079346, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8787989616394043, "num_tokens": 731618195.0, "step": 19264 }, { "epoch": 2.450706017046177, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.805687665939331, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8728865385055542, "num_tokens": 731658551.0, "step": 19265 }, { "epoch": 2.4508332273247677, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6024396419525146, "learning_rate": 1e-05, "loss": 0.3773, "mean_token_accuracy": 0.8919957876205444, "num_tokens": 731700477.0, "step": 19266 }, { "epoch": 2.4509604376033582, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.830152750015259, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.88332200050354, "num_tokens": 731735841.0, "step": 19267 }, { "epoch": 2.4510876478819488, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6235153675079346, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8753823041915894, "num_tokens": 731774974.0, "step": 19268 }, { "epoch": 2.4512148581605393, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.827033042907715, "learning_rate": 1e-05, "loss": 0.4058, "mean_token_accuracy": 0.8852648735046387, "num_tokens": 731806010.0, "step": 19269 }, { "epoch": 2.45134206843913, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6189961433410645, "learning_rate": 1e-05, "loss": 0.3668, "mean_token_accuracy": 0.8969237804412842, "num_tokens": 731841773.0, "step": 19270 }, { "epoch": 2.4514692787177204, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.829418659210205, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.8883970975875854, "num_tokens": 731877974.0, "step": 19271 }, { "epoch": 2.451596488996311, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6150853633880615, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.8816107511520386, "num_tokens": 731919375.0, "step": 19272 }, { "epoch": 2.4517236992749014, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.828594923019409, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8814506530761719, "num_tokens": 731957518.0, "step": 19273 }, { "epoch": 2.451850909553492, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6169703006744385, "learning_rate": 1e-05, "loss": 0.3873, "mean_token_accuracy": 0.8889508843421936, "num_tokens": 731996597.0, "step": 19274 }, { "epoch": 2.4519781198320825, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8230066299438477, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8802241086959839, "num_tokens": 732035888.0, "step": 19275 }, { "epoch": 2.452105330110673, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.8766515254974365, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8779393434524536, "num_tokens": 732073026.0, "step": 19276 }, { "epoch": 2.4522325403892635, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7262606620788574, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8572152853012085, "num_tokens": 732110794.0, "step": 19277 }, { "epoch": 2.452359750667854, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7222514152526855, "learning_rate": 1e-05, "loss": 0.3895, "mean_token_accuracy": 0.8890775442123413, "num_tokens": 732141051.0, "step": 19278 }, { "epoch": 2.4524869609464446, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6587727069854736, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8849419355392456, "num_tokens": 732187023.0, "step": 19279 }, { "epoch": 2.452614171225035, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.744948625564575, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8685778379440308, "num_tokens": 732220960.0, "step": 19280 }, { "epoch": 2.4527413815036256, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6571764945983887, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8811913728713989, "num_tokens": 732260656.0, "step": 19281 }, { "epoch": 2.452868591782216, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.748751401901245, "learning_rate": 1e-05, "loss": 0.497, "mean_token_accuracy": 0.8549543619155884, "num_tokens": 732301771.0, "step": 19282 }, { "epoch": 2.4529958020608067, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.646472215652466, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8734086155891418, "num_tokens": 732341618.0, "step": 19283 }, { "epoch": 2.453123012339397, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7438602447509766, "learning_rate": 1e-05, "loss": 0.3328, "mean_token_accuracy": 0.9099125266075134, "num_tokens": 732375797.0, "step": 19284 }, { "epoch": 2.4532502226179878, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.635061264038086, "learning_rate": 1e-05, "loss": 0.391, "mean_token_accuracy": 0.8891580700874329, "num_tokens": 732414234.0, "step": 19285 }, { "epoch": 2.453377432896578, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7372841835021973, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8761200904846191, "num_tokens": 732461063.0, "step": 19286 }, { "epoch": 2.4535046431751684, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6430814266204834, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8825294971466064, "num_tokens": 732501622.0, "step": 19287 }, { "epoch": 2.453631853453759, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7382829189300537, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8755909204483032, "num_tokens": 732546938.0, "step": 19288 }, { "epoch": 2.4537590637323494, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6488051414489746, "learning_rate": 1e-05, "loss": 0.3709, "mean_token_accuracy": 0.8965542912483215, "num_tokens": 732581250.0, "step": 19289 }, { "epoch": 2.45388627401094, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.739980697631836, "learning_rate": 1e-05, "loss": 0.3907, "mean_token_accuracy": 0.8912071585655212, "num_tokens": 732621081.0, "step": 19290 }, { "epoch": 2.4540134842895305, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.641472578048706, "learning_rate": 1e-05, "loss": 0.393, "mean_token_accuracy": 0.8880996108055115, "num_tokens": 732660625.0, "step": 19291 }, { "epoch": 2.454140694568121, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.735292911529541, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.883877158164978, "num_tokens": 732700100.0, "step": 19292 }, { "epoch": 2.4542679048467115, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.650618553161621, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8805403113365173, "num_tokens": 732735252.0, "step": 19293 }, { "epoch": 2.454395115125302, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.746406078338623, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8748862743377686, "num_tokens": 732767912.0, "step": 19294 }, { "epoch": 2.4545223254038926, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6379876136779785, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8734562397003174, "num_tokens": 732808499.0, "step": 19295 }, { "epoch": 2.454649535682483, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7582170963287354, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8812641501426697, "num_tokens": 732840754.0, "step": 19296 }, { "epoch": 2.4547767459610736, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6372592449188232, "learning_rate": 1e-05, "loss": 0.4825, "mean_token_accuracy": 0.8589522838592529, "num_tokens": 732882432.0, "step": 19297 }, { "epoch": 2.454903956239664, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7626688480377197, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.889079213142395, "num_tokens": 732912418.0, "step": 19298 }, { "epoch": 2.4550311665182547, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6351706981658936, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8833045363426208, "num_tokens": 732946050.0, "step": 19299 }, { "epoch": 2.4551583767968452, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.768507957458496, "learning_rate": 1e-05, "loss": 0.3972, "mean_token_accuracy": 0.8877947926521301, "num_tokens": 732980667.0, "step": 19300 }, { "epoch": 2.4552855870754358, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.62725830078125, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.882238507270813, "num_tokens": 733016169.0, "step": 19301 }, { "epoch": 2.4554127973540263, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000347137451171875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7685842514038086, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.875019907951355, "num_tokens": 733055120.0, "step": 19302 }, { "epoch": 2.455540007632617, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6348819732666016, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8706873655319214, "num_tokens": 733094550.0, "step": 19303 }, { "epoch": 2.4556672179112073, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7659904956817627, "learning_rate": 1e-05, "loss": 0.3705, "mean_token_accuracy": 0.8940556049346924, "num_tokens": 733132338.0, "step": 19304 }, { "epoch": 2.455794428189798, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.636946439743042, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8755914568901062, "num_tokens": 733167053.0, "step": 19305 }, { "epoch": 2.4559216384683884, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7697718143463135, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8706448674201965, "num_tokens": 733208175.0, "step": 19306 }, { "epoch": 2.456048848746979, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.640778064727783, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.868895411491394, "num_tokens": 733248336.0, "step": 19307 }, { "epoch": 2.4561760590255695, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7813313007354736, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8778484463691711, "num_tokens": 733283314.0, "step": 19308 }, { "epoch": 2.4563032693041595, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.632495164871216, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8793596029281616, "num_tokens": 733317541.0, "step": 19309 }, { "epoch": 2.4564304795827505, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.777272939682007, "learning_rate": 1e-05, "loss": 0.3602, "mean_token_accuracy": 0.8989672660827637, "num_tokens": 733349975.0, "step": 19310 }, { "epoch": 2.4565576898613406, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6221978664398193, "learning_rate": 1e-05, "loss": 0.3897, "mean_token_accuracy": 0.8909828662872314, "num_tokens": 733388085.0, "step": 19311 }, { "epoch": 2.456684900139931, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7824416160583496, "learning_rate": 1e-05, "loss": 0.3875, "mean_token_accuracy": 0.8898200988769531, "num_tokens": 733428420.0, "step": 19312 }, { "epoch": 2.4568121104185217, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6374855041503906, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.8853387832641602, "num_tokens": 733460271.0, "step": 19313 }, { "epoch": 2.456939320697112, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7758560180664062, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8803918361663818, "num_tokens": 733498610.0, "step": 19314 }, { "epoch": 2.4570665309757027, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.654022693634033, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8729245662689209, "num_tokens": 733530501.0, "step": 19315 }, { "epoch": 2.4571937412542932, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7700083255767822, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.8849204778671265, "num_tokens": 733563208.0, "step": 19316 }, { "epoch": 2.4573209515328838, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.650589942932129, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8884446024894714, "num_tokens": 733598729.0, "step": 19317 }, { "epoch": 2.4574481618114743, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.767158031463623, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8788876533508301, "num_tokens": 733632033.0, "step": 19318 }, { "epoch": 2.457575372090065, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6353201866149902, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8741353154182434, "num_tokens": 733673088.0, "step": 19319 }, { "epoch": 2.4577025823686554, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7662785053253174, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8718361854553223, "num_tokens": 733714096.0, "step": 19320 }, { "epoch": 2.457829792647246, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6320347785949707, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.872443675994873, "num_tokens": 733754495.0, "step": 19321 }, { "epoch": 2.4579570029258364, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7667226791381836, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8816001415252686, "num_tokens": 733788904.0, "step": 19322 }, { "epoch": 2.458084213204427, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6229071617126465, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8841207027435303, "num_tokens": 733831041.0, "step": 19323 }, { "epoch": 2.4582114234830175, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.775080919265747, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8702076077461243, "num_tokens": 733872209.0, "step": 19324 }, { "epoch": 2.458338633761608, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.639991283416748, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8729267120361328, "num_tokens": 733906884.0, "step": 19325 }, { "epoch": 2.4584658440401985, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7628490924835205, "learning_rate": 1e-05, "loss": 0.3756, "mean_token_accuracy": 0.895715594291687, "num_tokens": 733949262.0, "step": 19326 }, { "epoch": 2.458593054318789, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6309306621551514, "learning_rate": 1e-05, "loss": 0.3958, "mean_token_accuracy": 0.8878142833709717, "num_tokens": 733991111.0, "step": 19327 }, { "epoch": 2.4587202645973796, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.775111198425293, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8767014145851135, "num_tokens": 734030968.0, "step": 19328 }, { "epoch": 2.45884747487597, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.635655641555786, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8712029457092285, "num_tokens": 734070612.0, "step": 19329 }, { "epoch": 2.4589746851545606, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.76389217376709, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8837370872497559, "num_tokens": 734108682.0, "step": 19330 }, { "epoch": 2.459101895433151, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6412193775177, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8856750726699829, "num_tokens": 734141565.0, "step": 19331 }, { "epoch": 2.4592291057117417, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7589564323425293, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8849887847900391, "num_tokens": 734183048.0, "step": 19332 }, { "epoch": 2.459356315990332, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6503102779388428, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8737072348594666, "num_tokens": 734216141.0, "step": 19333 }, { "epoch": 2.4594835262689223, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7491674423217773, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8805813193321228, "num_tokens": 734256365.0, "step": 19334 }, { "epoch": 2.4596107365475133, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6254689693450928, "learning_rate": 1e-05, "loss": 0.3826, "mean_token_accuracy": 0.8910406827926636, "num_tokens": 734297679.0, "step": 19335 }, { "epoch": 2.4597379468261034, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.769965648651123, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8678216934204102, "num_tokens": 734333245.0, "step": 19336 }, { "epoch": 2.459865157104694, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.63203501701355, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8712365031242371, "num_tokens": 734374408.0, "step": 19337 }, { "epoch": 2.4599923673832844, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7810895442962646, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8854740858078003, "num_tokens": 734408105.0, "step": 19338 }, { "epoch": 2.460119577661875, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6132445335388184, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8817791938781738, "num_tokens": 734451700.0, "step": 19339 }, { "epoch": 2.4602467879404655, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7867722511291504, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8729802966117859, "num_tokens": 734492343.0, "step": 19340 }, { "epoch": 2.460373998219056, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6194043159484863, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8804546594619751, "num_tokens": 734531720.0, "step": 19341 }, { "epoch": 2.4605012084976465, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.794412136077881, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8851032853126526, "num_tokens": 734567809.0, "step": 19342 }, { "epoch": 2.460628418776237, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.60402250289917, "learning_rate": 1e-05, "loss": 0.3507, "mean_token_accuracy": 0.8979567289352417, "num_tokens": 734605491.0, "step": 19343 }, { "epoch": 2.4607556290548276, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8178417682647705, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8875728845596313, "num_tokens": 734637169.0, "step": 19344 }, { "epoch": 2.460882839333418, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.625396251678467, "learning_rate": 1e-05, "loss": 0.3928, "mean_token_accuracy": 0.8846822381019592, "num_tokens": 734668167.0, "step": 19345 }, { "epoch": 2.4610100496120086, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7978789806365967, "learning_rate": 1e-05, "loss": 0.3776, "mean_token_accuracy": 0.895287275314331, "num_tokens": 734705641.0, "step": 19346 }, { "epoch": 2.461137259890599, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.623537540435791, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8759174346923828, "num_tokens": 734742412.0, "step": 19347 }, { "epoch": 2.4612644701691897, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7894527912139893, "learning_rate": 1e-05, "loss": 0.384, "mean_token_accuracy": 0.8925710916519165, "num_tokens": 734782930.0, "step": 19348 }, { "epoch": 2.4613916804477802, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6165285110473633, "learning_rate": 1e-05, "loss": 0.3911, "mean_token_accuracy": 0.8892635107040405, "num_tokens": 734825051.0, "step": 19349 }, { "epoch": 2.4615188907263708, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.80206036567688, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8713058829307556, "num_tokens": 734863524.0, "step": 19350 }, { "epoch": 2.4616461010049613, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.630232095718384, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8738152980804443, "num_tokens": 734899692.0, "step": 19351 }, { "epoch": 2.461773311283552, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8160204887390137, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8635561466217041, "num_tokens": 734933114.0, "step": 19352 }, { "epoch": 2.4619005215621423, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6252105236053467, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8626923561096191, "num_tokens": 734975896.0, "step": 19353 }, { "epoch": 2.462027731840733, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.799046754837036, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8705687522888184, "num_tokens": 735017969.0, "step": 19354 }, { "epoch": 2.4621549421193234, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.634892463684082, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8771910071372986, "num_tokens": 735057052.0, "step": 19355 }, { "epoch": 2.462282152397914, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.795933485031128, "learning_rate": 1e-05, "loss": 0.3753, "mean_token_accuracy": 0.8956894278526306, "num_tokens": 735094211.0, "step": 19356 }, { "epoch": 2.4624093626765045, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6303176879882812, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.861204981803894, "num_tokens": 735134870.0, "step": 19357 }, { "epoch": 2.462536572955095, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7941348552703857, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8687182664871216, "num_tokens": 735177218.0, "step": 19358 }, { "epoch": 2.462663783233685, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.660828113555908, "learning_rate": 1e-05, "loss": 0.5014, "mean_token_accuracy": 0.8553391695022583, "num_tokens": 735208310.0, "step": 19359 }, { "epoch": 2.462790993512276, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.77687406539917, "learning_rate": 1e-05, "loss": 0.3878, "mean_token_accuracy": 0.8884316682815552, "num_tokens": 735243514.0, "step": 19360 }, { "epoch": 2.462918203790866, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6533567905426025, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8731188774108887, "num_tokens": 735278765.0, "step": 19361 }, { "epoch": 2.4630454140694567, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7582008838653564, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8651415705680847, "num_tokens": 735317966.0, "step": 19362 }, { "epoch": 2.463172624348047, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6495656967163086, "learning_rate": 1e-05, "loss": 0.3934, "mean_token_accuracy": 0.8855109214782715, "num_tokens": 735355552.0, "step": 19363 }, { "epoch": 2.4632998346266377, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7458009719848633, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8732714653015137, "num_tokens": 735398011.0, "step": 19364 }, { "epoch": 2.4634270449052282, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.645470380783081, "learning_rate": 1e-05, "loss": 0.3842, "mean_token_accuracy": 0.8936498761177063, "num_tokens": 735436743.0, "step": 19365 }, { "epoch": 2.4635542551838188, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.771947145462036, "learning_rate": 1e-05, "loss": 0.3793, "mean_token_accuracy": 0.8943753242492676, "num_tokens": 735467932.0, "step": 19366 }, { "epoch": 2.4636814654624093, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6243886947631836, "learning_rate": 1e-05, "loss": 0.389, "mean_token_accuracy": 0.8898073434829712, "num_tokens": 735510057.0, "step": 19367 }, { "epoch": 2.463808675741, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7519681453704834, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.875809907913208, "num_tokens": 735553700.0, "step": 19368 }, { "epoch": 2.4639358860195903, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.634242534637451, "learning_rate": 1e-05, "loss": 0.4792, "mean_token_accuracy": 0.8626179695129395, "num_tokens": 735598490.0, "step": 19369 }, { "epoch": 2.464063096298181, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.762758493423462, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8759965896606445, "num_tokens": 735637987.0, "step": 19370 }, { "epoch": 2.4641903065767714, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6239333152770996, "learning_rate": 1e-05, "loss": 0.3563, "mean_token_accuracy": 0.9014778137207031, "num_tokens": 735676327.0, "step": 19371 }, { "epoch": 2.464317516855362, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7727949619293213, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8768293857574463, "num_tokens": 735713015.0, "step": 19372 }, { "epoch": 2.4644447271339525, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.636428117752075, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8821419477462769, "num_tokens": 735750561.0, "step": 19373 }, { "epoch": 2.464571937412543, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7785210609436035, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.871866762638092, "num_tokens": 735788264.0, "step": 19374 }, { "epoch": 2.4646991476911335, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6273279190063477, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.880579948425293, "num_tokens": 735829384.0, "step": 19375 }, { "epoch": 2.464826357969724, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7601966857910156, "learning_rate": 1e-05, "loss": 0.3742, "mean_token_accuracy": 0.8959360122680664, "num_tokens": 735868323.0, "step": 19376 }, { "epoch": 2.4649535682483146, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6288344860076904, "learning_rate": 1e-05, "loss": 0.3832, "mean_token_accuracy": 0.8903466463088989, "num_tokens": 735902107.0, "step": 19377 }, { "epoch": 2.465080778526905, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7686102390289307, "learning_rate": 1e-05, "loss": 0.4143, "mean_token_accuracy": 0.8826217651367188, "num_tokens": 735941501.0, "step": 19378 }, { "epoch": 2.4652079888054956, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6236214637756348, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8744252920150757, "num_tokens": 735987721.0, "step": 19379 }, { "epoch": 2.465335199084086, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.78328013420105, "learning_rate": 1e-05, "loss": 0.3845, "mean_token_accuracy": 0.893817126750946, "num_tokens": 736022126.0, "step": 19380 }, { "epoch": 2.4654624093626767, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6305649280548096, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8868682384490967, "num_tokens": 736059782.0, "step": 19381 }, { "epoch": 2.4655896196412668, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7759456634521484, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8675798177719116, "num_tokens": 736101514.0, "step": 19382 }, { "epoch": 2.4657168299198577, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.624521017074585, "learning_rate": 1e-05, "loss": 0.3881, "mean_token_accuracy": 0.8879790902137756, "num_tokens": 736137861.0, "step": 19383 }, { "epoch": 2.465844040198448, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7804479598999023, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8763902187347412, "num_tokens": 736177567.0, "step": 19384 }, { "epoch": 2.4659712504770384, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.633275270462036, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.881548285484314, "num_tokens": 736214893.0, "step": 19385 }, { "epoch": 2.466098460755629, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.790123462677002, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8680485486984253, "num_tokens": 736250539.0, "step": 19386 }, { "epoch": 2.4662256710342194, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6235437393188477, "learning_rate": 1e-05, "loss": 0.3797, "mean_token_accuracy": 0.8941982984542847, "num_tokens": 736280924.0, "step": 19387 }, { "epoch": 2.46635288131281, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7832534313201904, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.880683422088623, "num_tokens": 736322446.0, "step": 19388 }, { "epoch": 2.4664800915914005, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.635457754135132, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8843783736228943, "num_tokens": 736361370.0, "step": 19389 }, { "epoch": 2.466607301869991, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7787094116210938, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8822181224822998, "num_tokens": 736401931.0, "step": 19390 }, { "epoch": 2.4667345121485815, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.63015079498291, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8837184906005859, "num_tokens": 736445125.0, "step": 19391 }, { "epoch": 2.466861722427172, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.769214391708374, "learning_rate": 1e-05, "loss": 0.3693, "mean_token_accuracy": 0.9001076817512512, "num_tokens": 736488409.0, "step": 19392 }, { "epoch": 2.4669889327057626, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6358566284179688, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8769969940185547, "num_tokens": 736526979.0, "step": 19393 }, { "epoch": 2.467116142984353, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.75752329826355, "learning_rate": 1e-05, "loss": 0.3916, "mean_token_accuracy": 0.888877272605896, "num_tokens": 736571933.0, "step": 19394 }, { "epoch": 2.4672433532629436, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6506659984588623, "learning_rate": 1e-05, "loss": 0.4702, "mean_token_accuracy": 0.8644417524337769, "num_tokens": 736610423.0, "step": 19395 }, { "epoch": 2.467370563541534, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.761796474456787, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8708059787750244, "num_tokens": 736648432.0, "step": 19396 }, { "epoch": 2.4674977738201247, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.655238151550293, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8627946376800537, "num_tokens": 736688032.0, "step": 19397 }, { "epoch": 2.4676249840987152, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7505342960357666, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8805277347564697, "num_tokens": 736725785.0, "step": 19398 }, { "epoch": 2.4677521943773058, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.663393020629883, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8598818778991699, "num_tokens": 736760229.0, "step": 19399 }, { "epoch": 2.4678794046558963, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.739743709564209, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8805176019668579, "num_tokens": 736794981.0, "step": 19400 }, { "epoch": 2.468006614934487, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.644843101501465, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.879814863204956, "num_tokens": 736835387.0, "step": 19401 }, { "epoch": 2.4681338252130773, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.723804235458374, "learning_rate": 1e-05, "loss": 0.3666, "mean_token_accuracy": 0.8987075090408325, "num_tokens": 736877047.0, "step": 19402 }, { "epoch": 2.468261035491668, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.652869939804077, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8728322982788086, "num_tokens": 736916637.0, "step": 19403 }, { "epoch": 2.4683882457702584, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7252309322357178, "learning_rate": 1e-05, "loss": 0.3818, "mean_token_accuracy": 0.8884060978889465, "num_tokens": 736954905.0, "step": 19404 }, { "epoch": 2.468515456048849, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.649205446243286, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8897867202758789, "num_tokens": 736992264.0, "step": 19405 }, { "epoch": 2.4686426663274394, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.726658821105957, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8636693954467773, "num_tokens": 737032896.0, "step": 19406 }, { "epoch": 2.4687698766060295, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6449873447418213, "learning_rate": 1e-05, "loss": 0.3922, "mean_token_accuracy": 0.8869008421897888, "num_tokens": 737073734.0, "step": 19407 }, { "epoch": 2.4688970868846205, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7314772605895996, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8853318095207214, "num_tokens": 737112246.0, "step": 19408 }, { "epoch": 2.4690242971632106, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.650632858276367, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8732733726501465, "num_tokens": 737148378.0, "step": 19409 }, { "epoch": 2.469151507441801, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.751868963241577, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8669465780258179, "num_tokens": 737182815.0, "step": 19410 }, { "epoch": 2.4692787177203916, "ewc_loss": 0.076171875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.124641418457031e-05, "grad_norm": 2.6265530586242676, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8844330310821533, "num_tokens": 737223301.0, "step": 19411 }, { "epoch": 2.469405927998982, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.775514602661133, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8741119503974915, "num_tokens": 737254184.0, "step": 19412 }, { "epoch": 2.4695331382775727, "ewc_loss": 0.07666015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6171059608459473, "learning_rate": 1e-05, "loss": 0.3584, "mean_token_accuracy": 0.8988109230995178, "num_tokens": 737292540.0, "step": 19413 }, { "epoch": 2.4696603485561632, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.780369758605957, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8654952645301819, "num_tokens": 737329822.0, "step": 19414 }, { "epoch": 2.4697875588347538, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6175527572631836, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8880115747451782, "num_tokens": 737365416.0, "step": 19415 }, { "epoch": 2.4699147691133443, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7715797424316406, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8828588724136353, "num_tokens": 737411498.0, "step": 19416 }, { "epoch": 2.470041979391935, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6321568489074707, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8875436782836914, "num_tokens": 737453667.0, "step": 19417 }, { "epoch": 2.4701691896705253, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.782601833343506, "learning_rate": 1e-05, "loss": 0.3946, "mean_token_accuracy": 0.8865764141082764, "num_tokens": 737487095.0, "step": 19418 }, { "epoch": 2.470296399949116, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6242969036102295, "learning_rate": 1e-05, "loss": 0.3864, "mean_token_accuracy": 0.8887215256690979, "num_tokens": 737521513.0, "step": 19419 }, { "epoch": 2.4704236102277064, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7805941104888916, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.889032244682312, "num_tokens": 737559218.0, "step": 19420 }, { "epoch": 2.470550820506297, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6309404373168945, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8689295053482056, "num_tokens": 737600486.0, "step": 19421 }, { "epoch": 2.4706780307848875, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.78422212600708, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8689851760864258, "num_tokens": 737639035.0, "step": 19422 }, { "epoch": 2.470805241063478, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6142876148223877, "learning_rate": 1e-05, "loss": 0.3584, "mean_token_accuracy": 0.8978018760681152, "num_tokens": 737676582.0, "step": 19423 }, { "epoch": 2.4709324513420685, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.801034450531006, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8826067447662354, "num_tokens": 737714809.0, "step": 19424 }, { "epoch": 2.471059661620659, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6101980209350586, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.8758695125579834, "num_tokens": 737757803.0, "step": 19425 }, { "epoch": 2.4711868718992496, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.810854911804199, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8790733218193054, "num_tokens": 737796563.0, "step": 19426 }, { "epoch": 2.47131408217784, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6313440799713135, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8742414712905884, "num_tokens": 737829711.0, "step": 19427 }, { "epoch": 2.4714412924564306, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.823756694793701, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8826228976249695, "num_tokens": 737862632.0, "step": 19428 }, { "epoch": 2.471568502735021, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6300368309020996, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8798160552978516, "num_tokens": 737900188.0, "step": 19429 }, { "epoch": 2.4716957130136117, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.816243886947632, "learning_rate": 1e-05, "loss": 0.4004, "mean_token_accuracy": 0.8861727714538574, "num_tokens": 737934667.0, "step": 19430 }, { "epoch": 2.471822923292202, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.631535530090332, "learning_rate": 1e-05, "loss": 0.3866, "mean_token_accuracy": 0.8867793679237366, "num_tokens": 737968386.0, "step": 19431 }, { "epoch": 2.4719501335707923, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.796984910964966, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8787705898284912, "num_tokens": 738011761.0, "step": 19432 }, { "epoch": 2.4720773438493833, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6335272789001465, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8868178129196167, "num_tokens": 738049319.0, "step": 19433 }, { "epoch": 2.4722045541279734, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.806720495223999, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8755472898483276, "num_tokens": 738083725.0, "step": 19434 }, { "epoch": 2.472331764406564, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.636338710784912, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8731285929679871, "num_tokens": 738126330.0, "step": 19435 }, { "epoch": 2.4724589746851544, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.807699203491211, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8775478005409241, "num_tokens": 738163564.0, "step": 19436 }, { "epoch": 2.472586184963745, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.650527000427246, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8732216954231262, "num_tokens": 738197014.0, "step": 19437 }, { "epoch": 2.4727133952423355, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.796597957611084, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8818619847297668, "num_tokens": 738232263.0, "step": 19438 }, { "epoch": 2.472840605520926, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6210556030273438, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.8897958397865295, "num_tokens": 738273839.0, "step": 19439 }, { "epoch": 2.4729678157995165, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7831339836120605, "learning_rate": 1e-05, "loss": 0.3752, "mean_token_accuracy": 0.8948823809623718, "num_tokens": 738314062.0, "step": 19440 }, { "epoch": 2.473095026078107, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6423821449279785, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8751270771026611, "num_tokens": 738352330.0, "step": 19441 }, { "epoch": 2.4732222363566976, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7951815128326416, "learning_rate": 1e-05, "loss": 0.4762, "mean_token_accuracy": 0.8606899380683899, "num_tokens": 738388527.0, "step": 19442 }, { "epoch": 2.473349446635288, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6548094749450684, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8747667670249939, "num_tokens": 738422913.0, "step": 19443 }, { "epoch": 2.4734766569138786, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.77603816986084, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8730260133743286, "num_tokens": 738463739.0, "step": 19444 }, { "epoch": 2.473603867192469, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6445472240448, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.870913028717041, "num_tokens": 738499160.0, "step": 19445 }, { "epoch": 2.4737310774710597, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.748903274536133, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.8892183303833008, "num_tokens": 738546634.0, "step": 19446 }, { "epoch": 2.47385828774965, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.658233880996704, "learning_rate": 1e-05, "loss": 0.3968, "mean_token_accuracy": 0.8877354860305786, "num_tokens": 738581021.0, "step": 19447 }, { "epoch": 2.4739854980282407, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.744415521621704, "learning_rate": 1e-05, "loss": 0.4113, "mean_token_accuracy": 0.8805373907089233, "num_tokens": 738622606.0, "step": 19448 }, { "epoch": 2.4741127083068313, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.660768985748291, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8788843750953674, "num_tokens": 738656244.0, "step": 19449 }, { "epoch": 2.474239918585422, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.745378255844116, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8700572848320007, "num_tokens": 738694743.0, "step": 19450 }, { "epoch": 2.4743671288640123, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6613194942474365, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8747400641441345, "num_tokens": 738731447.0, "step": 19451 }, { "epoch": 2.474494339142603, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7369701862335205, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8903552293777466, "num_tokens": 738764631.0, "step": 19452 }, { "epoch": 2.4746215494211934, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.634512424468994, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8888149857521057, "num_tokens": 738807685.0, "step": 19453 }, { "epoch": 2.474748759699784, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.733187198638916, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8764657974243164, "num_tokens": 738851717.0, "step": 19454 }, { "epoch": 2.4748759699783744, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6638245582580566, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8650490045547485, "num_tokens": 738889432.0, "step": 19455 }, { "epoch": 2.475003180256965, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7338995933532715, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8728102445602417, "num_tokens": 738923197.0, "step": 19456 }, { "epoch": 2.475130390535555, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.657270908355713, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8843314051628113, "num_tokens": 738959699.0, "step": 19457 }, { "epoch": 2.475257600814146, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7457146644592285, "learning_rate": 1e-05, "loss": 0.3805, "mean_token_accuracy": 0.894611120223999, "num_tokens": 738990444.0, "step": 19458 }, { "epoch": 2.475384811092736, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6379919052124023, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8905431032180786, "num_tokens": 739030864.0, "step": 19459 }, { "epoch": 2.4755120213713266, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.754176378250122, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8752090930938721, "num_tokens": 739068407.0, "step": 19460 }, { "epoch": 2.475639231649917, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.636366367340088, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.882436215877533, "num_tokens": 739105560.0, "step": 19461 }, { "epoch": 2.4757664419285077, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.76584529876709, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8695424199104309, "num_tokens": 739142056.0, "step": 19462 }, { "epoch": 2.4758936522070982, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.656580686569214, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8605223894119263, "num_tokens": 739185407.0, "step": 19463 }, { "epoch": 2.4760208624856888, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7525994777679443, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8749665021896362, "num_tokens": 739222594.0, "step": 19464 }, { "epoch": 2.4761480727642793, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.629441499710083, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8826490640640259, "num_tokens": 739261021.0, "step": 19465 }, { "epoch": 2.47627528304287, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.761770248413086, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8759809732437134, "num_tokens": 739300828.0, "step": 19466 }, { "epoch": 2.4764024933214603, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.633467435836792, "learning_rate": 1e-05, "loss": 0.3901, "mean_token_accuracy": 0.8887863755226135, "num_tokens": 739338724.0, "step": 19467 }, { "epoch": 2.476529703600051, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.769575357437134, "learning_rate": 1e-05, "loss": 0.3936, "mean_token_accuracy": 0.8883378505706787, "num_tokens": 739377821.0, "step": 19468 }, { "epoch": 2.4766569138786414, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.629288673400879, "learning_rate": 1e-05, "loss": 0.3792, "mean_token_accuracy": 0.8920999765396118, "num_tokens": 739411438.0, "step": 19469 }, { "epoch": 2.476784124157232, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7672135829925537, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8699976205825806, "num_tokens": 739457836.0, "step": 19470 }, { "epoch": 2.4769113344358225, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.634770393371582, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8704790472984314, "num_tokens": 739496336.0, "step": 19471 }, { "epoch": 2.477038544714413, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7663912773132324, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8816426992416382, "num_tokens": 739539777.0, "step": 19472 }, { "epoch": 2.4771657549930035, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6313424110412598, "learning_rate": 1e-05, "loss": 0.3765, "mean_token_accuracy": 0.8955510854721069, "num_tokens": 739572560.0, "step": 19473 }, { "epoch": 2.477292965271594, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.780683755874634, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.8862217664718628, "num_tokens": 739604260.0, "step": 19474 }, { "epoch": 2.4774201755501846, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.625091075897217, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8796377182006836, "num_tokens": 739645241.0, "step": 19475 }, { "epoch": 2.477547385828775, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.782188653945923, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8831697106361389, "num_tokens": 739683516.0, "step": 19476 }, { "epoch": 2.4776745961073656, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6244096755981445, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8783345222473145, "num_tokens": 739720422.0, "step": 19477 }, { "epoch": 2.477801806385956, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7783994674682617, "learning_rate": 1e-05, "loss": 0.377, "mean_token_accuracy": 0.894263744354248, "num_tokens": 739758071.0, "step": 19478 }, { "epoch": 2.4779290166645467, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.622744083404541, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8842341899871826, "num_tokens": 739796143.0, "step": 19479 }, { "epoch": 2.4780562269431368, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8009519577026367, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8760179281234741, "num_tokens": 739832038.0, "step": 19480 }, { "epoch": 2.4781834372217277, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6143593788146973, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8835610747337341, "num_tokens": 739873232.0, "step": 19481 }, { "epoch": 2.478310647500318, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7989487648010254, "learning_rate": 1e-05, "loss": 0.3972, "mean_token_accuracy": 0.8864379525184631, "num_tokens": 739912213.0, "step": 19482 }, { "epoch": 2.4784378577789083, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.628167152404785, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8770297765731812, "num_tokens": 739949363.0, "step": 19483 }, { "epoch": 2.478565068057499, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.824885129928589, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8842180371284485, "num_tokens": 739979360.0, "step": 19484 }, { "epoch": 2.4786922783360894, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6256790161132812, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8648444414138794, "num_tokens": 740019691.0, "step": 19485 }, { "epoch": 2.47881948861468, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.844958782196045, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8861376047134399, "num_tokens": 740053604.0, "step": 19486 }, { "epoch": 2.4789466988932705, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.627927303314209, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8756373524665833, "num_tokens": 740090220.0, "step": 19487 }, { "epoch": 2.479073909171861, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.817138671875, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8918159008026123, "num_tokens": 740126361.0, "step": 19488 }, { "epoch": 2.4792011194504515, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6198084354400635, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8793400526046753, "num_tokens": 740164752.0, "step": 19489 }, { "epoch": 2.479328329729042, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.821546792984009, "learning_rate": 1e-05, "loss": 0.4033, "mean_token_accuracy": 0.8838354349136353, "num_tokens": 740199622.0, "step": 19490 }, { "epoch": 2.4794555400076326, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6296002864837646, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8795424699783325, "num_tokens": 740236457.0, "step": 19491 }, { "epoch": 2.479582750286223, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.821010112762451, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8735823035240173, "num_tokens": 740275094.0, "step": 19492 }, { "epoch": 2.4797099605648136, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.638291358947754, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8770308494567871, "num_tokens": 740308564.0, "step": 19493 }, { "epoch": 2.479837170843404, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.80967378616333, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8805249929428101, "num_tokens": 740354994.0, "step": 19494 }, { "epoch": 2.4799643811219947, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6240720748901367, "learning_rate": 1e-05, "loss": 0.4025, "mean_token_accuracy": 0.8905007243156433, "num_tokens": 740394924.0, "step": 19495 }, { "epoch": 2.480091591400585, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.823564052581787, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.8901936411857605, "num_tokens": 740428362.0, "step": 19496 }, { "epoch": 2.4802188016791757, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6160662174224854, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8708926439285278, "num_tokens": 740473987.0, "step": 19497 }, { "epoch": 2.4803460119577663, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8282878398895264, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8649681806564331, "num_tokens": 740512458.0, "step": 19498 }, { "epoch": 2.480473222236357, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.625176429748535, "learning_rate": 1e-05, "loss": 0.4789, "mean_token_accuracy": 0.8616816401481628, "num_tokens": 740558744.0, "step": 19499 }, { "epoch": 2.4806004325149473, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.823751926422119, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.869391143321991, "num_tokens": 740599889.0, "step": 19500 }, { "epoch": 2.480727642793538, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6260621547698975, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8805646896362305, "num_tokens": 740636162.0, "step": 19501 }, { "epoch": 2.4808548530721284, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.812429428100586, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8633161783218384, "num_tokens": 740679595.0, "step": 19502 }, { "epoch": 2.480982063350719, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.641291856765747, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8872097730636597, "num_tokens": 740714830.0, "step": 19503 }, { "epoch": 2.4811092736293094, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.804478645324707, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8844817876815796, "num_tokens": 740749748.0, "step": 19504 }, { "epoch": 2.4812364839078995, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6413893699645996, "learning_rate": 1e-05, "loss": 0.4543, "mean_token_accuracy": 0.8669118881225586, "num_tokens": 740785110.0, "step": 19505 }, { "epoch": 2.4813636941864905, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8146352767944336, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8843668699264526, "num_tokens": 740824432.0, "step": 19506 }, { "epoch": 2.4814909044650806, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6396565437316895, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8773949146270752, "num_tokens": 740861403.0, "step": 19507 }, { "epoch": 2.481618114743671, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8018462657928467, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.878095269203186, "num_tokens": 740899299.0, "step": 19508 }, { "epoch": 2.4817453250222616, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6421620845794678, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8862507343292236, "num_tokens": 740934178.0, "step": 19509 }, { "epoch": 2.481872535300852, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8000311851501465, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8763329982757568, "num_tokens": 740967010.0, "step": 19510 }, { "epoch": 2.4819997455794427, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6338701248168945, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8768004179000854, "num_tokens": 741004655.0, "step": 19511 }, { "epoch": 2.4821269558580332, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8049416542053223, "learning_rate": 1e-05, "loss": 0.3898, "mean_token_accuracy": 0.8912547826766968, "num_tokens": 741037365.0, "step": 19512 }, { "epoch": 2.4822541661366238, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6395533084869385, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8747180700302124, "num_tokens": 741072304.0, "step": 19513 }, { "epoch": 2.4823813764152143, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7933201789855957, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8822730183601379, "num_tokens": 741107187.0, "step": 19514 }, { "epoch": 2.482508586693805, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.65098237991333, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.878519594669342, "num_tokens": 741142152.0, "step": 19515 }, { "epoch": 2.4826357969723953, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7876956462860107, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.886854887008667, "num_tokens": 741175704.0, "step": 19516 }, { "epoch": 2.482763007250986, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.642489433288574, "learning_rate": 1e-05, "loss": 0.3876, "mean_token_accuracy": 0.8908831477165222, "num_tokens": 741212187.0, "step": 19517 }, { "epoch": 2.4828902175295764, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7832438945770264, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8803466558456421, "num_tokens": 741246830.0, "step": 19518 }, { "epoch": 2.483017427808167, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.649184226989746, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8767075538635254, "num_tokens": 741279672.0, "step": 19519 }, { "epoch": 2.4831446380867574, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7663612365722656, "learning_rate": 1e-05, "loss": 0.3893, "mean_token_accuracy": 0.8914787769317627, "num_tokens": 741318205.0, "step": 19520 }, { "epoch": 2.483271848365348, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6375412940979004, "learning_rate": 1e-05, "loss": 0.3986, "mean_token_accuracy": 0.8850399255752563, "num_tokens": 741356312.0, "step": 19521 }, { "epoch": 2.4833990586439385, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.780203104019165, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8716534376144409, "num_tokens": 741393370.0, "step": 19522 }, { "epoch": 2.483526268922529, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6426920890808105, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.869166374206543, "num_tokens": 741428533.0, "step": 19523 }, { "epoch": 2.4836534792011196, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7775042057037354, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.873013973236084, "num_tokens": 741464614.0, "step": 19524 }, { "epoch": 2.48378068947971, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.641859292984009, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8777191638946533, "num_tokens": 741499126.0, "step": 19525 }, { "epoch": 2.4839078997583006, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7542896270751953, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8851820826530457, "num_tokens": 741544381.0, "step": 19526 }, { "epoch": 2.484035110036891, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6250574588775635, "learning_rate": 1e-05, "loss": 0.3781, "mean_token_accuracy": 0.89534991979599, "num_tokens": 741588627.0, "step": 19527 }, { "epoch": 2.4841623203154817, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7648494243621826, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8860927820205688, "num_tokens": 741631165.0, "step": 19528 }, { "epoch": 2.484289530594072, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.633798837661743, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8821796178817749, "num_tokens": 741668667.0, "step": 19529 }, { "epoch": 2.4844167408726623, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.778730869293213, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8808770179748535, "num_tokens": 741703625.0, "step": 19530 }, { "epoch": 2.4845439511512533, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6304614543914795, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8767039179801941, "num_tokens": 741744371.0, "step": 19531 }, { "epoch": 2.4846711614298433, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.773257255554199, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8714478015899658, "num_tokens": 741786733.0, "step": 19532 }, { "epoch": 2.484798371708434, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.622349262237549, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8807228803634644, "num_tokens": 741825693.0, "step": 19533 }, { "epoch": 2.4849255819870244, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.788851499557495, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8625388145446777, "num_tokens": 741866176.0, "step": 19534 }, { "epoch": 2.485052792265615, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6400272846221924, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8706157803535461, "num_tokens": 741902523.0, "step": 19535 }, { "epoch": 2.4851800025442055, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.784172534942627, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8707761764526367, "num_tokens": 741939005.0, "step": 19536 }, { "epoch": 2.485307212822796, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6362385749816895, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.8658767938613892, "num_tokens": 741975012.0, "step": 19537 }, { "epoch": 2.4854344231013865, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.895411252975464, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8810492753982544, "num_tokens": 742008351.0, "step": 19538 }, { "epoch": 2.485561633379977, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.601055860519409, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8831596374511719, "num_tokens": 742045477.0, "step": 19539 }, { "epoch": 2.4856888436585676, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8525493144989014, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.86655592918396, "num_tokens": 742080070.0, "step": 19540 }, { "epoch": 2.485816053937158, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.612807273864746, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.883116602897644, "num_tokens": 742116735.0, "step": 19541 }, { "epoch": 2.4859432642157486, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.858851194381714, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8787205219268799, "num_tokens": 742151492.0, "step": 19542 }, { "epoch": 2.486070474494339, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.5964226722717285, "learning_rate": 1e-05, "loss": 0.3856, "mean_token_accuracy": 0.8924464583396912, "num_tokens": 742196913.0, "step": 19543 }, { "epoch": 2.4861976847729297, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8497977256774902, "learning_rate": 1e-05, "loss": 0.3785, "mean_token_accuracy": 0.8957343697547913, "num_tokens": 742241894.0, "step": 19544 }, { "epoch": 2.48632489505152, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.630660057067871, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8860217332839966, "num_tokens": 742276256.0, "step": 19545 }, { "epoch": 2.4864521053301107, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8372182846069336, "learning_rate": 1e-05, "loss": 0.3816, "mean_token_accuracy": 0.895104169845581, "num_tokens": 742316834.0, "step": 19546 }, { "epoch": 2.4865793156087013, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6250922679901123, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8818799257278442, "num_tokens": 742359132.0, "step": 19547 }, { "epoch": 2.486706525887292, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.844338893890381, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.864582896232605, "num_tokens": 742399746.0, "step": 19548 }, { "epoch": 2.4868337361658823, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.645514726638794, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8687351942062378, "num_tokens": 742435400.0, "step": 19549 }, { "epoch": 2.486960946444473, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.824850559234619, "learning_rate": 1e-05, "loss": 0.3971, "mean_token_accuracy": 0.8865507245063782, "num_tokens": 742472935.0, "step": 19550 }, { "epoch": 2.4870881567230634, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.628483295440674, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8846994638442993, "num_tokens": 742516841.0, "step": 19551 }, { "epoch": 2.487215367001654, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.83450984954834, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8744012117385864, "num_tokens": 742551057.0, "step": 19552 }, { "epoch": 2.4873425772802444, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6486239433288574, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8765245676040649, "num_tokens": 742584891.0, "step": 19553 }, { "epoch": 2.487469787558835, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8027758598327637, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8866010308265686, "num_tokens": 742622574.0, "step": 19554 }, { "epoch": 2.487596997837425, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6359312534332275, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8794535398483276, "num_tokens": 742664458.0, "step": 19555 }, { "epoch": 2.487724208116016, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8077309131622314, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8826413154602051, "num_tokens": 742704006.0, "step": 19556 }, { "epoch": 2.487851418394606, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.658524751663208, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8697195053100586, "num_tokens": 742736885.0, "step": 19557 }, { "epoch": 2.4879786286731966, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.791090726852417, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8770159482955933, "num_tokens": 742773496.0, "step": 19558 }, { "epoch": 2.488105838951787, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6502416133880615, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8838299512863159, "num_tokens": 742812444.0, "step": 19559 }, { "epoch": 2.4882330492303777, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.784716844558716, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.879403829574585, "num_tokens": 742849910.0, "step": 19560 }, { "epoch": 2.488360259508968, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.621398448944092, "learning_rate": 1e-05, "loss": 0.3887, "mean_token_accuracy": 0.8906991481781006, "num_tokens": 742891412.0, "step": 19561 }, { "epoch": 2.4884874697875587, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8014349937438965, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8756283521652222, "num_tokens": 742929510.0, "step": 19562 }, { "epoch": 2.4886146800661493, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.633701801300049, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8704580068588257, "num_tokens": 742969767.0, "step": 19563 }, { "epoch": 2.48874189034474, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8141086101531982, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8875388503074646, "num_tokens": 743001416.0, "step": 19564 }, { "epoch": 2.4888691006233303, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6323750019073486, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8761171102523804, "num_tokens": 743037273.0, "step": 19565 }, { "epoch": 2.488996310901921, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.811997652053833, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8763867020606995, "num_tokens": 743072774.0, "step": 19566 }, { "epoch": 2.4891235211805114, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6279420852661133, "learning_rate": 1e-05, "loss": 0.4507, "mean_token_accuracy": 0.8682149052619934, "num_tokens": 743115931.0, "step": 19567 }, { "epoch": 2.489250731459102, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8081672191619873, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.8905797004699707, "num_tokens": 743158500.0, "step": 19568 }, { "epoch": 2.4893779417376924, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.658597469329834, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8829306960105896, "num_tokens": 743195311.0, "step": 19569 }, { "epoch": 2.489505152016283, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.798844575881958, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8839201927185059, "num_tokens": 743227978.0, "step": 19570 }, { "epoch": 2.4896323622948735, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.64312481880188, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8857370615005493, "num_tokens": 743264297.0, "step": 19571 }, { "epoch": 2.489759572573464, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7809829711914062, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8757180571556091, "num_tokens": 743305351.0, "step": 19572 }, { "epoch": 2.4898867828520546, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.646676778793335, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8745814561843872, "num_tokens": 743345105.0, "step": 19573 }, { "epoch": 2.490013993130645, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.772029161453247, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.889089822769165, "num_tokens": 743383889.0, "step": 19574 }, { "epoch": 2.4901412034092356, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6457817554473877, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8809384703636169, "num_tokens": 743420045.0, "step": 19575 }, { "epoch": 2.490268413687826, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.765272378921509, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8784779906272888, "num_tokens": 743455358.0, "step": 19576 }, { "epoch": 2.4903956239664167, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6600146293640137, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8610671162605286, "num_tokens": 743492337.0, "step": 19577 }, { "epoch": 2.4905228342450068, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.759678840637207, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.8871423006057739, "num_tokens": 743526023.0, "step": 19578 }, { "epoch": 2.4906500445235977, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6451151371002197, "learning_rate": 1e-05, "loss": 0.4044, "mean_token_accuracy": 0.8815542459487915, "num_tokens": 743560224.0, "step": 19579 }, { "epoch": 2.490777254802188, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.761213779449463, "learning_rate": 1e-05, "loss": 0.4975, "mean_token_accuracy": 0.8537027835845947, "num_tokens": 743601403.0, "step": 19580 }, { "epoch": 2.4909044650807783, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6402995586395264, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8823553323745728, "num_tokens": 743641677.0, "step": 19581 }, { "epoch": 2.491031675359369, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7470407485961914, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8784450888633728, "num_tokens": 743684615.0, "step": 19582 }, { "epoch": 2.4911588856379594, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6455893516540527, "learning_rate": 1e-05, "loss": 0.3707, "mean_token_accuracy": 0.8948211669921875, "num_tokens": 743718960.0, "step": 19583 }, { "epoch": 2.49128609591655, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.744227886199951, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.889726459980011, "num_tokens": 743758336.0, "step": 19584 }, { "epoch": 2.4914133061951405, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.643662929534912, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8701563477516174, "num_tokens": 743803243.0, "step": 19585 }, { "epoch": 2.491540516473731, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7428271770477295, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8856157064437866, "num_tokens": 743840945.0, "step": 19586 }, { "epoch": 2.4916677267523215, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6380648612976074, "learning_rate": 1e-05, "loss": 0.4129, "mean_token_accuracy": 0.8828179836273193, "num_tokens": 743884338.0, "step": 19587 }, { "epoch": 2.491794937030912, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7547786235809326, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8801314830780029, "num_tokens": 743925499.0, "step": 19588 }, { "epoch": 2.4919221473095026, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6495368480682373, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8667656779289246, "num_tokens": 743965016.0, "step": 19589 }, { "epoch": 2.492049357588093, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.770648717880249, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8742802143096924, "num_tokens": 744000523.0, "step": 19590 }, { "epoch": 2.4921765678666836, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6201820373535156, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8762520551681519, "num_tokens": 744043862.0, "step": 19591 }, { "epoch": 2.492303778145274, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.775120973587036, "learning_rate": 1e-05, "loss": 0.4732, "mean_token_accuracy": 0.863276481628418, "num_tokens": 744085161.0, "step": 19592 }, { "epoch": 2.4924309884238647, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.637078046798706, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8830139636993408, "num_tokens": 744117575.0, "step": 19593 }, { "epoch": 2.492558198702455, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7850029468536377, "learning_rate": 1e-05, "loss": 0.5048, "mean_token_accuracy": 0.8543202877044678, "num_tokens": 744156312.0, "step": 19594 }, { "epoch": 2.4926854089810457, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6269760131835938, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8823258876800537, "num_tokens": 744191177.0, "step": 19595 }, { "epoch": 2.4928126192596363, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.780205249786377, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8746465444564819, "num_tokens": 744230407.0, "step": 19596 }, { "epoch": 2.492939829538227, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6212494373321533, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.876188337802887, "num_tokens": 744270265.0, "step": 19597 }, { "epoch": 2.4930670398168173, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.784076690673828, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8807209730148315, "num_tokens": 744310913.0, "step": 19598 }, { "epoch": 2.493194250095408, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.641449451446533, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8742056488990784, "num_tokens": 744342918.0, "step": 19599 }, { "epoch": 2.4933214603739984, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.780867099761963, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8773511052131653, "num_tokens": 744379000.0, "step": 19600 }, { "epoch": 2.493448670652589, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6248972415924072, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8782639503479004, "num_tokens": 744414641.0, "step": 19601 }, { "epoch": 2.4935758809311794, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8098323345184326, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8731992840766907, "num_tokens": 744444860.0, "step": 19602 }, { "epoch": 2.4937030912097695, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.628100872039795, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.8895499110221863, "num_tokens": 744479657.0, "step": 19603 }, { "epoch": 2.4938303014883605, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.785963296890259, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.880553126335144, "num_tokens": 744520048.0, "step": 19604 }, { "epoch": 2.4939575117669506, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6371610164642334, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8721410036087036, "num_tokens": 744555265.0, "step": 19605 }, { "epoch": 2.494084722045541, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.786386728286743, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8750218152999878, "num_tokens": 744591011.0, "step": 19606 }, { "epoch": 2.4942119323241316, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6269471645355225, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8753429651260376, "num_tokens": 744632302.0, "step": 19607 }, { "epoch": 2.494339142602722, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.784947156906128, "learning_rate": 1e-05, "loss": 0.3845, "mean_token_accuracy": 0.8887591361999512, "num_tokens": 744669827.0, "step": 19608 }, { "epoch": 2.4944663528813127, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6212313175201416, "learning_rate": 1e-05, "loss": 0.3914, "mean_token_accuracy": 0.885828971862793, "num_tokens": 744708438.0, "step": 19609 }, { "epoch": 2.494593563159903, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7972123622894287, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8806920051574707, "num_tokens": 744744218.0, "step": 19610 }, { "epoch": 2.4947207734384937, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.622732639312744, "learning_rate": 1e-05, "loss": 0.4037, "mean_token_accuracy": 0.8838520050048828, "num_tokens": 744782805.0, "step": 19611 }, { "epoch": 2.4948479837170843, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.797698497772217, "learning_rate": 1e-05, "loss": 0.3591, "mean_token_accuracy": 0.8987232446670532, "num_tokens": 744814011.0, "step": 19612 }, { "epoch": 2.494975193995675, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6174216270446777, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8806928396224976, "num_tokens": 744857385.0, "step": 19613 }, { "epoch": 2.4951024042742653, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8094515800476074, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8623248338699341, "num_tokens": 744898212.0, "step": 19614 }, { "epoch": 2.495229614552856, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6368746757507324, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8762457370758057, "num_tokens": 744930665.0, "step": 19615 }, { "epoch": 2.4953568248314464, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8064770698547363, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8696911334991455, "num_tokens": 744968725.0, "step": 19616 }, { "epoch": 2.495484035110037, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.625631093978882, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8678195476531982, "num_tokens": 745008386.0, "step": 19617 }, { "epoch": 2.4956112453886274, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8021960258483887, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8755874037742615, "num_tokens": 745047059.0, "step": 19618 }, { "epoch": 2.495738455667218, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.62648606300354, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8717854619026184, "num_tokens": 745089768.0, "step": 19619 }, { "epoch": 2.4958656659458085, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8067171573638916, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8759629726409912, "num_tokens": 745130848.0, "step": 19620 }, { "epoch": 2.495992876224399, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.638014078140259, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8765833377838135, "num_tokens": 745166715.0, "step": 19621 }, { "epoch": 2.4961200865029896, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8282885551452637, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8714988231658936, "num_tokens": 745200718.0, "step": 19622 }, { "epoch": 2.49624729678158, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6240665912628174, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.8716344833374023, "num_tokens": 745240736.0, "step": 19623 }, { "epoch": 2.4963745070601706, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8059730529785156, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8799876570701599, "num_tokens": 745279040.0, "step": 19624 }, { "epoch": 2.496501717338761, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.637965202331543, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.873857855796814, "num_tokens": 745316936.0, "step": 19625 }, { "epoch": 2.4966289276173517, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8149893283843994, "learning_rate": 1e-05, "loss": 0.4914, "mean_token_accuracy": 0.8587194681167603, "num_tokens": 745355438.0, "step": 19626 }, { "epoch": 2.496756137895942, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.630481004714966, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8839402198791504, "num_tokens": 745394509.0, "step": 19627 }, { "epoch": 2.4968833481745323, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8131585121154785, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8765929937362671, "num_tokens": 745426347.0, "step": 19628 }, { "epoch": 2.4970105584531233, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6436028480529785, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.8866441249847412, "num_tokens": 745457523.0, "step": 19629 }, { "epoch": 2.4971377687317133, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8102803230285645, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.86440509557724, "num_tokens": 745495155.0, "step": 19630 }, { "epoch": 2.497264979010304, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6288859844207764, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8750574588775635, "num_tokens": 745535466.0, "step": 19631 }, { "epoch": 2.4973921892888944, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003509521484375, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8071088790893555, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8769086003303528, "num_tokens": 745569184.0, "step": 19632 }, { "epoch": 2.497519399567485, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6309361457824707, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8724269866943359, "num_tokens": 745610735.0, "step": 19633 }, { "epoch": 2.4976466098460754, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.804788827896118, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8677505254745483, "num_tokens": 745653513.0, "step": 19634 }, { "epoch": 2.497773820124666, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.639303207397461, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8761026859283447, "num_tokens": 745690007.0, "step": 19635 }, { "epoch": 2.4979010304032565, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8084537982940674, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8687341809272766, "num_tokens": 745724886.0, "step": 19636 }, { "epoch": 2.498028240681847, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6323063373565674, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8861581087112427, "num_tokens": 745760782.0, "step": 19637 }, { "epoch": 2.4981554509604376, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.797342538833618, "learning_rate": 1e-05, "loss": 0.3793, "mean_token_accuracy": 0.8936309814453125, "num_tokens": 745794535.0, "step": 19638 }, { "epoch": 2.498282661239028, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6479439735412598, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8790808320045471, "num_tokens": 745831010.0, "step": 19639 }, { "epoch": 2.4984098715176186, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7909653186798096, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8728363513946533, "num_tokens": 745866761.0, "step": 19640 }, { "epoch": 2.498537081796209, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.623220205307007, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8818091154098511, "num_tokens": 745909081.0, "step": 19641 }, { "epoch": 2.4986642920747997, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8162975311279297, "learning_rate": 1e-05, "loss": 0.3817, "mean_token_accuracy": 0.8945268392562866, "num_tokens": 745942907.0, "step": 19642 }, { "epoch": 2.49879150235339, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6284706592559814, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8757028579711914, "num_tokens": 745981558.0, "step": 19643 }, { "epoch": 2.4989187126319807, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8140146732330322, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8799877762794495, "num_tokens": 746015568.0, "step": 19644 }, { "epoch": 2.4990459229105713, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.631432056427002, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.885995090007782, "num_tokens": 746049503.0, "step": 19645 }, { "epoch": 2.499173133189162, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.806077003479004, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8800426721572876, "num_tokens": 746086979.0, "step": 19646 }, { "epoch": 2.4993003434677523, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6311676502227783, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8812326192855835, "num_tokens": 746125073.0, "step": 19647 }, { "epoch": 2.499427553746343, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.806217670440674, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8816905617713928, "num_tokens": 746161621.0, "step": 19648 }, { "epoch": 2.4995547640249334, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6298038959503174, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8746311664581299, "num_tokens": 746203032.0, "step": 19649 }, { "epoch": 2.499681974303524, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.829322099685669, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.879274845123291, "num_tokens": 746233939.0, "step": 19650 }, { "epoch": 2.499809184582114, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.628993511199951, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8894267082214355, "num_tokens": 746269175.0, "step": 19651 }, { "epoch": 2.499936394860705, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8192310333251953, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8743329048156738, "num_tokens": 746310634.0, "step": 19652 }, { "epoch": 2.500063605139295, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.619635820388794, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.883568286895752, "num_tokens": 746353683.0, "step": 19653 }, { "epoch": 2.500190815417886, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8205344676971436, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8763287663459778, "num_tokens": 746392299.0, "step": 19654 }, { "epoch": 2.500318025696476, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6135692596435547, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8821595907211304, "num_tokens": 746435276.0, "step": 19655 }, { "epoch": 2.5004452359750666, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.833930015563965, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8715278506278992, "num_tokens": 746476175.0, "step": 19656 }, { "epoch": 2.500572446253657, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.608879566192627, "learning_rate": 1e-05, "loss": 0.3829, "mean_token_accuracy": 0.8892241716384888, "num_tokens": 746517084.0, "step": 19657 }, { "epoch": 2.5006996565322477, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.836142063140869, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8643599152565002, "num_tokens": 746556073.0, "step": 19658 }, { "epoch": 2.500826866810838, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.630725860595703, "learning_rate": 1e-05, "loss": 0.3831, "mean_token_accuracy": 0.8908384442329407, "num_tokens": 746590188.0, "step": 19659 }, { "epoch": 2.5009540770894287, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.841329336166382, "learning_rate": 1e-05, "loss": 0.5241, "mean_token_accuracy": 0.852638304233551, "num_tokens": 746634667.0, "step": 19660 }, { "epoch": 2.5010812873680193, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6135478019714355, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.8863897919654846, "num_tokens": 746677398.0, "step": 19661 }, { "epoch": 2.50120849764661, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.8484549522399902, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8861650824546814, "num_tokens": 746711607.0, "step": 19662 }, { "epoch": 2.5013357079252003, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6213810443878174, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8864856362342834, "num_tokens": 746751853.0, "step": 19663 }, { "epoch": 2.501462918203791, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.842890739440918, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8672728538513184, "num_tokens": 746792134.0, "step": 19664 }, { "epoch": 2.5015901284823814, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.640226364135742, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.869516909122467, "num_tokens": 746830087.0, "step": 19665 }, { "epoch": 2.501717338760972, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8300743103027344, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8851838111877441, "num_tokens": 746864738.0, "step": 19666 }, { "epoch": 2.5018445490395624, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6428563594818115, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.8631093502044678, "num_tokens": 746904021.0, "step": 19667 }, { "epoch": 2.501971759318153, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8105556964874268, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.8846347332000732, "num_tokens": 746942710.0, "step": 19668 }, { "epoch": 2.5020989695967435, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6242072582244873, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8810133934020996, "num_tokens": 746986694.0, "step": 19669 }, { "epoch": 2.502226179875334, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8040077686309814, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.8911527395248413, "num_tokens": 747028440.0, "step": 19670 }, { "epoch": 2.5023533901539246, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.653881311416626, "learning_rate": 1e-05, "loss": 0.3867, "mean_token_accuracy": 0.8888005614280701, "num_tokens": 747063862.0, "step": 19671 }, { "epoch": 2.502480600432515, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.798443555831909, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8757443428039551, "num_tokens": 747100469.0, "step": 19672 }, { "epoch": 2.5026078107111056, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.649996519088745, "learning_rate": 1e-05, "loss": 0.3769, "mean_token_accuracy": 0.8912097215652466, "num_tokens": 747133074.0, "step": 19673 }, { "epoch": 2.5027350209896957, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.811474084854126, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8669903874397278, "num_tokens": 747167010.0, "step": 19674 }, { "epoch": 2.5028622312682867, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6541669368743896, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8757926821708679, "num_tokens": 747201397.0, "step": 19675 }, { "epoch": 2.5029894415468767, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.777219295501709, "learning_rate": 1e-05, "loss": 0.4014, "mean_token_accuracy": 0.8862773776054382, "num_tokens": 747238430.0, "step": 19676 }, { "epoch": 2.5031166518254677, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.647118091583252, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8787580728530884, "num_tokens": 747277760.0, "step": 19677 }, { "epoch": 2.503243862104058, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.774893283843994, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8782534003257751, "num_tokens": 747317763.0, "step": 19678 }, { "epoch": 2.5033710723826488, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.649583339691162, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8695191740989685, "num_tokens": 747362733.0, "step": 19679 }, { "epoch": 2.503498282661239, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7521603107452393, "learning_rate": 1e-05, "loss": 0.3918, "mean_token_accuracy": 0.8879988193511963, "num_tokens": 747400735.0, "step": 19680 }, { "epoch": 2.5036254929398294, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6461081504821777, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8756740093231201, "num_tokens": 747440263.0, "step": 19681 }, { "epoch": 2.50375270321842, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.755869150161743, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8809165358543396, "num_tokens": 747477918.0, "step": 19682 }, { "epoch": 2.5038799134970104, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6431121826171875, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8782667517662048, "num_tokens": 747513235.0, "step": 19683 }, { "epoch": 2.504007123775601, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7516722679138184, "learning_rate": 1e-05, "loss": 0.3832, "mean_token_accuracy": 0.8966284394264221, "num_tokens": 747553280.0, "step": 19684 }, { "epoch": 2.5041343340541915, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6470465660095215, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8739738464355469, "num_tokens": 747594389.0, "step": 19685 }, { "epoch": 2.504261544332782, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7490274906158447, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8834238648414612, "num_tokens": 747634805.0, "step": 19686 }, { "epoch": 2.5043887546113726, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.658834218978882, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8706198334693909, "num_tokens": 747673963.0, "step": 19687 }, { "epoch": 2.504515964889963, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7499289512634277, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8721402883529663, "num_tokens": 747715067.0, "step": 19688 }, { "epoch": 2.5046431751685536, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.644285202026367, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8847880363464355, "num_tokens": 747752799.0, "step": 19689 }, { "epoch": 2.504770385447144, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.760066509246826, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8706346750259399, "num_tokens": 747789325.0, "step": 19690 }, { "epoch": 2.5048975957257347, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.654390573501587, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.877714991569519, "num_tokens": 747823438.0, "step": 19691 }, { "epoch": 2.505024806004325, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7957277297973633, "learning_rate": 1e-05, "loss": 0.3946, "mean_token_accuracy": 0.887890100479126, "num_tokens": 747862310.0, "step": 19692 }, { "epoch": 2.5051520162829157, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.622363805770874, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8810524940490723, "num_tokens": 747903154.0, "step": 19693 }, { "epoch": 2.5052792265615063, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7899184226989746, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8768665790557861, "num_tokens": 747937849.0, "step": 19694 }, { "epoch": 2.505406436840097, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.625084161758423, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8794198036193848, "num_tokens": 747972387.0, "step": 19695 }, { "epoch": 2.5055336471186873, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8985390663146973, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8655411005020142, "num_tokens": 748007181.0, "step": 19696 }, { "epoch": 2.505660857397278, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.605015993118286, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8848910331726074, "num_tokens": 748044342.0, "step": 19697 }, { "epoch": 2.5057880676758684, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.866594076156616, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8734334707260132, "num_tokens": 748074012.0, "step": 19698 }, { "epoch": 2.5059152779544585, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.591707229614258, "learning_rate": 1e-05, "loss": 0.3908, "mean_token_accuracy": 0.8883472681045532, "num_tokens": 748111055.0, "step": 19699 }, { "epoch": 2.5060424882330494, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.8848116397857666, "learning_rate": 1e-05, "loss": 0.4798, "mean_token_accuracy": 0.8607980012893677, "num_tokens": 748151922.0, "step": 19700 }, { "epoch": 2.5061696985116395, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6102678775787354, "learning_rate": 1e-05, "loss": 0.3918, "mean_token_accuracy": 0.8886438608169556, "num_tokens": 748186090.0, "step": 19701 }, { "epoch": 2.5062969087902305, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.875786542892456, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.8857585787773132, "num_tokens": 748223281.0, "step": 19702 }, { "epoch": 2.5064241190688206, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6393275260925293, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8813824653625488, "num_tokens": 748255657.0, "step": 19703 }, { "epoch": 2.5065513293474115, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8801727294921875, "learning_rate": 1e-05, "loss": 0.4621, "mean_token_accuracy": 0.8704252243041992, "num_tokens": 748296954.0, "step": 19704 }, { "epoch": 2.5066785396260016, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.652273416519165, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8743098378181458, "num_tokens": 748334753.0, "step": 19705 }, { "epoch": 2.506805749904592, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.844461441040039, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8819260597229004, "num_tokens": 748377880.0, "step": 19706 }, { "epoch": 2.5069329601831827, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6496024131774902, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8751755952835083, "num_tokens": 748418230.0, "step": 19707 }, { "epoch": 2.507060170461773, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.828014612197876, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8843253254890442, "num_tokens": 748459320.0, "step": 19708 }, { "epoch": 2.5071873807403637, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.65362286567688, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8623863458633423, "num_tokens": 748503951.0, "step": 19709 }, { "epoch": 2.5073145910189543, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8144073486328125, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8791952133178711, "num_tokens": 748544736.0, "step": 19710 }, { "epoch": 2.507441801297545, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6562280654907227, "learning_rate": 1e-05, "loss": 0.4246, "mean_token_accuracy": 0.8806206583976746, "num_tokens": 748583174.0, "step": 19711 }, { "epoch": 2.5075690115761353, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.80413818359375, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8819212317466736, "num_tokens": 748624580.0, "step": 19712 }, { "epoch": 2.507696221854726, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6651899814605713, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8649338483810425, "num_tokens": 748660343.0, "step": 19713 }, { "epoch": 2.5078234321333164, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7981173992156982, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.867671549320221, "num_tokens": 748695006.0, "step": 19714 }, { "epoch": 2.507950642411907, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6474335193634033, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8855875134468079, "num_tokens": 748732142.0, "step": 19715 }, { "epoch": 2.5080778526904974, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.798882484436035, "learning_rate": 1e-05, "loss": 0.4548, "mean_token_accuracy": 0.8695148229598999, "num_tokens": 748771263.0, "step": 19716 }, { "epoch": 2.508205062969088, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.658399820327759, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8753950595855713, "num_tokens": 748812796.0, "step": 19717 }, { "epoch": 2.5083322732476785, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.77909255027771, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8773548603057861, "num_tokens": 748852701.0, "step": 19718 }, { "epoch": 2.508459483526269, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.666313886642456, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8710914254188538, "num_tokens": 748888799.0, "step": 19719 }, { "epoch": 2.5085866938048595, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.765929698944092, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8814070224761963, "num_tokens": 748927056.0, "step": 19720 }, { "epoch": 2.50871390408345, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6677589416503906, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.8648185729980469, "num_tokens": 748961829.0, "step": 19721 }, { "epoch": 2.5088411143620406, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7518491744995117, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8877464532852173, "num_tokens": 748995281.0, "step": 19722 }, { "epoch": 2.508968324640631, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.659621238708496, "learning_rate": 1e-05, "loss": 0.4008, "mean_token_accuracy": 0.8864354491233826, "num_tokens": 749027574.0, "step": 19723 }, { "epoch": 2.509095534919221, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7409088611602783, "learning_rate": 1e-05, "loss": 0.4044, "mean_token_accuracy": 0.884743332862854, "num_tokens": 749065092.0, "step": 19724 }, { "epoch": 2.509222745197812, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.659006118774414, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8716831207275391, "num_tokens": 749104389.0, "step": 19725 }, { "epoch": 2.5093499554764023, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7428903579711914, "learning_rate": 1e-05, "loss": 0.3753, "mean_token_accuracy": 0.8926130533218384, "num_tokens": 749138433.0, "step": 19726 }, { "epoch": 2.5094771657549932, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6490797996520996, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8864723443984985, "num_tokens": 749178095.0, "step": 19727 }, { "epoch": 2.5096043760335833, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.736048460006714, "learning_rate": 1e-05, "loss": 0.3813, "mean_token_accuracy": 0.8952466249465942, "num_tokens": 749215547.0, "step": 19728 }, { "epoch": 2.5097315863121743, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6600544452667236, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8661558628082275, "num_tokens": 749254935.0, "step": 19729 }, { "epoch": 2.5098587965907644, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7444794178009033, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8907421827316284, "num_tokens": 749290425.0, "step": 19730 }, { "epoch": 2.509986006869355, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.666560173034668, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8845900297164917, "num_tokens": 749319390.0, "step": 19731 }, { "epoch": 2.5101132171479454, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7340853214263916, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8674370050430298, "num_tokens": 749358517.0, "step": 19732 }, { "epoch": 2.510240427426536, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6550474166870117, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8783813714981079, "num_tokens": 749402219.0, "step": 19733 }, { "epoch": 2.5103676377051265, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7614035606384277, "learning_rate": 1e-05, "loss": 0.5022, "mean_token_accuracy": 0.8516572117805481, "num_tokens": 749437514.0, "step": 19734 }, { "epoch": 2.510494847983717, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6366801261901855, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8827837109565735, "num_tokens": 749477148.0, "step": 19735 }, { "epoch": 2.5106220582623076, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7579355239868164, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8834266662597656, "num_tokens": 749511174.0, "step": 19736 }, { "epoch": 2.510749268540898, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.7078676223754883, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.8863934278488159, "num_tokens": 749547117.0, "step": 19737 }, { "epoch": 2.5108764788194886, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7191734313964844, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8880200386047363, "num_tokens": 749583755.0, "step": 19738 }, { "epoch": 2.511003689098079, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6684162616729736, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8833788633346558, "num_tokens": 749621742.0, "step": 19739 }, { "epoch": 2.5111308993766697, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7108185291290283, "learning_rate": 1e-05, "loss": 0.3855, "mean_token_accuracy": 0.8916248083114624, "num_tokens": 749659178.0, "step": 19740 }, { "epoch": 2.51125810965526, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6703145503997803, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.871242880821228, "num_tokens": 749698828.0, "step": 19741 }, { "epoch": 2.5113853199338507, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7423861026763916, "learning_rate": 1e-05, "loss": 0.4624, "mean_token_accuracy": 0.8659330606460571, "num_tokens": 749732609.0, "step": 19742 }, { "epoch": 2.5115125302124413, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.656069278717041, "learning_rate": 1e-05, "loss": 0.3655, "mean_token_accuracy": 0.8982560634613037, "num_tokens": 749768240.0, "step": 19743 }, { "epoch": 2.511639740491032, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.8029141426086426, "learning_rate": 1e-05, "loss": 0.3915, "mean_token_accuracy": 0.8867558240890503, "num_tokens": 749809435.0, "step": 19744 }, { "epoch": 2.5117669507696223, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6248104572296143, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8696320652961731, "num_tokens": 749852790.0, "step": 19745 }, { "epoch": 2.511894161048213, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7687015533447266, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.875253438949585, "num_tokens": 749895111.0, "step": 19746 }, { "epoch": 2.5120213713268034, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.611402988433838, "learning_rate": 1e-05, "loss": 0.3898, "mean_token_accuracy": 0.887836754322052, "num_tokens": 749936142.0, "step": 19747 }, { "epoch": 2.512148581605394, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7876765727996826, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8869512677192688, "num_tokens": 749972106.0, "step": 19748 }, { "epoch": 2.512275791883984, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.613330841064453, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8679158091545105, "num_tokens": 750014090.0, "step": 19749 }, { "epoch": 2.512403002162575, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7960896492004395, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8870970010757446, "num_tokens": 750052542.0, "step": 19750 }, { "epoch": 2.512530212441165, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6252474784851074, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8649653196334839, "num_tokens": 750091846.0, "step": 19751 }, { "epoch": 2.512657422719756, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.811511993408203, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8635875582695007, "num_tokens": 750135721.0, "step": 19752 }, { "epoch": 2.512784632998346, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.618152618408203, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8689086437225342, "num_tokens": 750177786.0, "step": 19753 }, { "epoch": 2.5129118432769366, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.824031114578247, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8761030435562134, "num_tokens": 750209577.0, "step": 19754 }, { "epoch": 2.513039053555527, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.612271308898926, "learning_rate": 1e-05, "loss": 0.3729, "mean_token_accuracy": 0.8954932689666748, "num_tokens": 750243168.0, "step": 19755 }, { "epoch": 2.5131662638341177, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8101963996887207, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8757542967796326, "num_tokens": 750288304.0, "step": 19756 }, { "epoch": 2.513293474112708, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.615427017211914, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8767834901809692, "num_tokens": 750329662.0, "step": 19757 }, { "epoch": 2.5134206843912987, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.828423023223877, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8723205327987671, "num_tokens": 750364581.0, "step": 19758 }, { "epoch": 2.5135478946698893, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.626823902130127, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8831677436828613, "num_tokens": 750403129.0, "step": 19759 }, { "epoch": 2.51367510494848, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8199105262756348, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8820950984954834, "num_tokens": 750439738.0, "step": 19760 }, { "epoch": 2.5138023152270703, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.626225233078003, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8648468255996704, "num_tokens": 750479302.0, "step": 19761 }, { "epoch": 2.513929525505661, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.820779323577881, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8812274932861328, "num_tokens": 750517658.0, "step": 19762 }, { "epoch": 2.5140567357842514, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.633103370666504, "learning_rate": 1e-05, "loss": 0.3773, "mean_token_accuracy": 0.8925670385360718, "num_tokens": 750558192.0, "step": 19763 }, { "epoch": 2.514183946062842, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.809483528137207, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8813214898109436, "num_tokens": 750595209.0, "step": 19764 }, { "epoch": 2.5143111563414324, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.646625280380249, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.873670756816864, "num_tokens": 750634560.0, "step": 19765 }, { "epoch": 2.514438366620023, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.788740873336792, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8821912407875061, "num_tokens": 750675223.0, "step": 19766 }, { "epoch": 2.5145655768986135, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.635908603668213, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8787114024162292, "num_tokens": 750717302.0, "step": 19767 }, { "epoch": 2.514692787177204, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8122124671936035, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8768433928489685, "num_tokens": 750754615.0, "step": 19768 }, { "epoch": 2.5148199974557945, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.639981985092163, "learning_rate": 1e-05, "loss": 0.3787, "mean_token_accuracy": 0.8944085836410522, "num_tokens": 750788745.0, "step": 19769 }, { "epoch": 2.514947207734385, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8056063652038574, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8632661700248718, "num_tokens": 750829727.0, "step": 19770 }, { "epoch": 2.5150744180129756, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6331286430358887, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8669607639312744, "num_tokens": 750873095.0, "step": 19771 }, { "epoch": 2.5152016282915657, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8160998821258545, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8673859238624573, "num_tokens": 750909513.0, "step": 19772 }, { "epoch": 2.5153288385701567, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.653726100921631, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8784639835357666, "num_tokens": 750947234.0, "step": 19773 }, { "epoch": 2.5154560488487467, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7800140380859375, "learning_rate": 1e-05, "loss": 0.3786, "mean_token_accuracy": 0.8902822136878967, "num_tokens": 750984926.0, "step": 19774 }, { "epoch": 2.5155832591273377, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6517791748046875, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8800762891769409, "num_tokens": 751023700.0, "step": 19775 }, { "epoch": 2.515710469405928, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.764598846435547, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8806689977645874, "num_tokens": 751065444.0, "step": 19776 }, { "epoch": 2.5158376796845188, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6631557941436768, "learning_rate": 1e-05, "loss": 0.4645, "mean_token_accuracy": 0.8647414445877075, "num_tokens": 751101627.0, "step": 19777 }, { "epoch": 2.515964889963109, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7517592906951904, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.8854990005493164, "num_tokens": 751138006.0, "step": 19778 }, { "epoch": 2.5160921002416994, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6487715244293213, "learning_rate": 1e-05, "loss": 0.3811, "mean_token_accuracy": 0.8903895020484924, "num_tokens": 751179187.0, "step": 19779 }, { "epoch": 2.51621931052029, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.75103497505188, "learning_rate": 1e-05, "loss": 0.41, "mean_token_accuracy": 0.8832278847694397, "num_tokens": 751217086.0, "step": 19780 }, { "epoch": 2.5163465207988804, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.645202398300171, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8761703968048096, "num_tokens": 751259330.0, "step": 19781 }, { "epoch": 2.516473731077471, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7544617652893066, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8796157836914062, "num_tokens": 751300873.0, "step": 19782 }, { "epoch": 2.5166009413560615, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6621878147125244, "learning_rate": 1e-05, "loss": 0.3746, "mean_token_accuracy": 0.8868818283081055, "num_tokens": 751330783.0, "step": 19783 }, { "epoch": 2.516728151634652, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7629575729370117, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8841177225112915, "num_tokens": 751370032.0, "step": 19784 }, { "epoch": 2.5168553619132426, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.637500524520874, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8836812973022461, "num_tokens": 751412561.0, "step": 19785 }, { "epoch": 2.516982572191833, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7796638011932373, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8786106109619141, "num_tokens": 751442168.0, "step": 19786 }, { "epoch": 2.5171097824704236, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.633991003036499, "learning_rate": 1e-05, "loss": 0.486, "mean_token_accuracy": 0.8585115671157837, "num_tokens": 751488656.0, "step": 19787 }, { "epoch": 2.517236992749014, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.765362501144409, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8778951168060303, "num_tokens": 751533305.0, "step": 19788 }, { "epoch": 2.5173642030276047, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.631685733795166, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8784903883934021, "num_tokens": 751570693.0, "step": 19789 }, { "epoch": 2.517491413306195, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.770871639251709, "learning_rate": 1e-05, "loss": 0.3855, "mean_token_accuracy": 0.8945228457450867, "num_tokens": 751604994.0, "step": 19790 }, { "epoch": 2.5176186235847857, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.641098976135254, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8789474964141846, "num_tokens": 751642552.0, "step": 19791 }, { "epoch": 2.5177458338633762, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7784793376922607, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8862249851226807, "num_tokens": 751677253.0, "step": 19792 }, { "epoch": 2.5178730441419668, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.632704019546509, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8846158981323242, "num_tokens": 751717737.0, "step": 19793 }, { "epoch": 2.5180002544205573, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.779240131378174, "learning_rate": 1e-05, "loss": 0.3662, "mean_token_accuracy": 0.8989872932434082, "num_tokens": 751750453.0, "step": 19794 }, { "epoch": 2.518127464699148, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6219935417175293, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8782126903533936, "num_tokens": 751792179.0, "step": 19795 }, { "epoch": 2.5182546749777384, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7819416522979736, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8815200328826904, "num_tokens": 751833367.0, "step": 19796 }, { "epoch": 2.5183818852563284, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6233456134796143, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8836328983306885, "num_tokens": 751874042.0, "step": 19797 }, { "epoch": 2.5185090955349194, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7982263565063477, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.8561792373657227, "num_tokens": 751914990.0, "step": 19798 }, { "epoch": 2.5186363058135095, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6523518562316895, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8648896217346191, "num_tokens": 751946266.0, "step": 19799 }, { "epoch": 2.5187635160921005, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7663354873657227, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8902493119239807, "num_tokens": 751990359.0, "step": 19800 }, { "epoch": 2.5188907263706906, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6417758464813232, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8772003650665283, "num_tokens": 752026955.0, "step": 19801 }, { "epoch": 2.5190179366492815, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.774083375930786, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8689553737640381, "num_tokens": 752065187.0, "step": 19802 }, { "epoch": 2.5191451469278716, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6421728134155273, "learning_rate": 1e-05, "loss": 0.3721, "mean_token_accuracy": 0.8939266204833984, "num_tokens": 752100855.0, "step": 19803 }, { "epoch": 2.519272357206462, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7756199836730957, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.883081316947937, "num_tokens": 752136672.0, "step": 19804 }, { "epoch": 2.5193995674850527, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6529319286346436, "learning_rate": 1e-05, "loss": 0.3998, "mean_token_accuracy": 0.8862369060516357, "num_tokens": 752167529.0, "step": 19805 }, { "epoch": 2.519526777763643, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.782170057296753, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8659861087799072, "num_tokens": 752201865.0, "step": 19806 }, { "epoch": 2.5196539880422337, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.634168863296509, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8818885684013367, "num_tokens": 752239315.0, "step": 19807 }, { "epoch": 2.5197811983208243, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.769277572631836, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8869274258613586, "num_tokens": 752278505.0, "step": 19808 }, { "epoch": 2.519908408599415, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6310856342315674, "learning_rate": 1e-05, "loss": 0.398, "mean_token_accuracy": 0.8873289823532104, "num_tokens": 752319346.0, "step": 19809 }, { "epoch": 2.5200356188780053, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7725868225097656, "learning_rate": 1e-05, "loss": 0.4598, "mean_token_accuracy": 0.8686886429786682, "num_tokens": 752364210.0, "step": 19810 }, { "epoch": 2.520162829156596, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6362144947052, "learning_rate": 1e-05, "loss": 0.3794, "mean_token_accuracy": 0.8925812244415283, "num_tokens": 752400679.0, "step": 19811 }, { "epoch": 2.5202900394351864, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.779463529586792, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8795821666717529, "num_tokens": 752436003.0, "step": 19812 }, { "epoch": 2.520417249713777, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.630552053451538, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8712932467460632, "num_tokens": 752478690.0, "step": 19813 }, { "epoch": 2.5205444599923674, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.807506561279297, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8652873039245605, "num_tokens": 752513326.0, "step": 19814 }, { "epoch": 2.520671670270958, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.621854782104492, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8702341318130493, "num_tokens": 752556553.0, "step": 19815 }, { "epoch": 2.5207988805495485, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7846429347991943, "learning_rate": 1e-05, "loss": 0.3888, "mean_token_accuracy": 0.8884862661361694, "num_tokens": 752599219.0, "step": 19816 }, { "epoch": 2.520926090828139, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6197030544281006, "learning_rate": 1e-05, "loss": 0.4098, "mean_token_accuracy": 0.8842536807060242, "num_tokens": 752639308.0, "step": 19817 }, { "epoch": 2.5210533011067295, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8051750659942627, "learning_rate": 1e-05, "loss": 0.4751, "mean_token_accuracy": 0.8632066249847412, "num_tokens": 752679384.0, "step": 19818 }, { "epoch": 2.52118051138532, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.619027853012085, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8729335069656372, "num_tokens": 752722395.0, "step": 19819 }, { "epoch": 2.5213077216639106, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.840704917907715, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8789002895355225, "num_tokens": 752754491.0, "step": 19820 }, { "epoch": 2.521434931942501, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.621619701385498, "learning_rate": 1e-05, "loss": 0.3924, "mean_token_accuracy": 0.8878476619720459, "num_tokens": 752792595.0, "step": 19821 }, { "epoch": 2.521562142221091, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.827969789505005, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8770599961280823, "num_tokens": 752830689.0, "step": 19822 }, { "epoch": 2.521689352499682, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.624290943145752, "learning_rate": 1e-05, "loss": 0.3716, "mean_token_accuracy": 0.8933284282684326, "num_tokens": 752862848.0, "step": 19823 }, { "epoch": 2.5218165627782723, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8228769302368164, "learning_rate": 1e-05, "loss": 0.4156, "mean_token_accuracy": 0.8819577693939209, "num_tokens": 752900566.0, "step": 19824 }, { "epoch": 2.5219437730568632, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6220734119415283, "learning_rate": 1e-05, "loss": 0.3948, "mean_token_accuracy": 0.8879073858261108, "num_tokens": 752939201.0, "step": 19825 }, { "epoch": 2.5220709833354533, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8335394859313965, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8873567581176758, "num_tokens": 752973820.0, "step": 19826 }, { "epoch": 2.522198193614044, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6356189250946045, "learning_rate": 1e-05, "loss": 0.4883, "mean_token_accuracy": 0.8591533899307251, "num_tokens": 753018563.0, "step": 19827 }, { "epoch": 2.5223254038926344, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8287763595581055, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8713833093643188, "num_tokens": 753060097.0, "step": 19828 }, { "epoch": 2.522452614171225, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.637035846710205, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8735781908035278, "num_tokens": 753099200.0, "step": 19829 }, { "epoch": 2.5225798244498154, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.829956531524658, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8812515735626221, "num_tokens": 753132421.0, "step": 19830 }, { "epoch": 2.522707034728406, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.649073362350464, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.869340717792511, "num_tokens": 753165380.0, "step": 19831 }, { "epoch": 2.5228342450069965, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.805495500564575, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8800492882728577, "num_tokens": 753203501.0, "step": 19832 }, { "epoch": 2.522961455285587, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.638334274291992, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.869126558303833, "num_tokens": 753247674.0, "step": 19833 }, { "epoch": 2.5230886655641775, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8001129627227783, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8819988369941711, "num_tokens": 753285967.0, "step": 19834 }, { "epoch": 2.523215875842768, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6535580158233643, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8747187852859497, "num_tokens": 753319080.0, "step": 19835 }, { "epoch": 2.5233430861213586, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7998249530792236, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8699706792831421, "num_tokens": 753357924.0, "step": 19836 }, { "epoch": 2.523470296399949, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.638834238052368, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8858500123023987, "num_tokens": 753397922.0, "step": 19837 }, { "epoch": 2.5235975066785397, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.818081855773926, "learning_rate": 1e-05, "loss": 0.4961, "mean_token_accuracy": 0.8557969331741333, "num_tokens": 753431268.0, "step": 19838 }, { "epoch": 2.52372471695713, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6499385833740234, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8735737800598145, "num_tokens": 753465169.0, "step": 19839 }, { "epoch": 2.5238519272357207, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.795656204223633, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8822546601295471, "num_tokens": 753501017.0, "step": 19840 }, { "epoch": 2.5239791375143112, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.633819341659546, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8814498782157898, "num_tokens": 753541039.0, "step": 19841 }, { "epoch": 2.5241063477929018, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.794512987136841, "learning_rate": 1e-05, "loss": 0.5069, "mean_token_accuracy": 0.85218346118927, "num_tokens": 753585793.0, "step": 19842 }, { "epoch": 2.5242335580714923, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.631711483001709, "learning_rate": 1e-05, "loss": 0.3782, "mean_token_accuracy": 0.8908382654190063, "num_tokens": 753623752.0, "step": 19843 }, { "epoch": 2.524360768350083, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.795584201812744, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8725738525390625, "num_tokens": 753662939.0, "step": 19844 }, { "epoch": 2.5244879786286734, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.647794246673584, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8843852281570435, "num_tokens": 753699200.0, "step": 19845 }, { "epoch": 2.524615188907264, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.803154945373535, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8693805932998657, "num_tokens": 753743867.0, "step": 19846 }, { "epoch": 2.524742399185854, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 3.0817368030548096, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8675065636634827, "num_tokens": 753781494.0, "step": 19847 }, { "epoch": 2.524869609464445, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.647555112838745, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8813717365264893, "num_tokens": 753823518.0, "step": 19848 }, { "epoch": 2.524996819743035, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8341991901397705, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8758336305618286, "num_tokens": 753855058.0, "step": 19849 }, { "epoch": 2.525124030021626, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.590604782104492, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8837628960609436, "num_tokens": 753896721.0, "step": 19850 }, { "epoch": 2.525251240300216, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 2.856379985809326, "learning_rate": 1e-05, "loss": 0.3881, "mean_token_accuracy": 0.8924068212509155, "num_tokens": 753933225.0, "step": 19851 }, { "epoch": 2.5253784505788066, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.612203359603882, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8730076551437378, "num_tokens": 753970373.0, "step": 19852 }, { "epoch": 2.525505660857397, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.982948303222656e-05, "grad_norm": 2.864138603210449, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.8870683908462524, "num_tokens": 754007544.0, "step": 19853 }, { "epoch": 2.5256328711359877, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6344428062438965, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8848920464515686, "num_tokens": 754040858.0, "step": 19854 }, { "epoch": 2.525760081414578, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.9114227294921875e-05, "grad_norm": 2.857758045196533, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8619078397750854, "num_tokens": 754083877.0, "step": 19855 }, { "epoch": 2.5258872916931687, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6272785663604736, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.8899762034416199, "num_tokens": 754122792.0, "step": 19856 }, { "epoch": 2.5260145019717593, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 2.8582100868225098, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8801279664039612, "num_tokens": 754157986.0, "step": 19857 }, { "epoch": 2.52614171225035, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.635584831237793, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8785673379898071, "num_tokens": 754198321.0, "step": 19858 }, { "epoch": 2.5262689225289403, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.9114227294921875e-05, "grad_norm": 2.8706395626068115, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8642213344573975, "num_tokens": 754235463.0, "step": 19859 }, { "epoch": 2.526396132807531, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.6328654289245605, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8798603415489197, "num_tokens": 754274319.0, "step": 19860 }, { "epoch": 2.5265233430861214, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003528594970703125, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 2.8536124229431152, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8825917840003967, "num_tokens": 754315206.0, "step": 19861 }, { "epoch": 2.526650553364712, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.640979290008545, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8741455674171448, "num_tokens": 754353300.0, "step": 19862 }, { "epoch": 2.5267777636433024, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.839897155761719e-05, "grad_norm": 2.8468594551086426, "learning_rate": 1e-05, "loss": 0.4652, "mean_token_accuracy": 0.8697490692138672, "num_tokens": 754392692.0, "step": 19863 }, { "epoch": 2.526904973921893, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6471564769744873, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8778712749481201, "num_tokens": 754428688.0, "step": 19864 }, { "epoch": 2.5270321842004835, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8245112895965576, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8800143003463745, "num_tokens": 754467647.0, "step": 19865 }, { "epoch": 2.527159394479074, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6633167266845703, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8680009841918945, "num_tokens": 754501511.0, "step": 19866 }, { "epoch": 2.5272866047576645, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.796738862991333, "learning_rate": 1e-05, "loss": 0.3822, "mean_token_accuracy": 0.8918851613998413, "num_tokens": 754543000.0, "step": 19867 }, { "epoch": 2.527413815036255, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.647268533706665, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8813638687133789, "num_tokens": 754583485.0, "step": 19868 }, { "epoch": 2.5275410253148456, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.799252986907959, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8821697235107422, "num_tokens": 754620472.0, "step": 19869 }, { "epoch": 2.5276682355934357, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.649160623550415, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8830959796905518, "num_tokens": 754659906.0, "step": 19870 }, { "epoch": 2.5277954458720266, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7859745025634766, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8809303045272827, "num_tokens": 754696128.0, "step": 19871 }, { "epoch": 2.5279226561506167, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6586954593658447, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8810195326805115, "num_tokens": 754729553.0, "step": 19872 }, { "epoch": 2.5280498664292077, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7702224254608154, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8647584915161133, "num_tokens": 754773689.0, "step": 19873 }, { "epoch": 2.528177076707798, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.662076950073242, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8748605251312256, "num_tokens": 754809326.0, "step": 19874 }, { "epoch": 2.5283042869863888, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7518301010131836, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8800204992294312, "num_tokens": 754850627.0, "step": 19875 }, { "epoch": 2.528431497264979, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6612942218780518, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8850072622299194, "num_tokens": 754887810.0, "step": 19876 }, { "epoch": 2.5285587075435694, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.761991024017334, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8776651620864868, "num_tokens": 754923778.0, "step": 19877 }, { "epoch": 2.52868591782216, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6519343852996826, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8817988634109497, "num_tokens": 754963486.0, "step": 19878 }, { "epoch": 2.5288131281007504, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.752718210220337, "learning_rate": 1e-05, "loss": 0.3864, "mean_token_accuracy": 0.8907150030136108, "num_tokens": 755000302.0, "step": 19879 }, { "epoch": 2.528940338379341, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6376068592071533, "learning_rate": 1e-05, "loss": 0.3765, "mean_token_accuracy": 0.8923107981681824, "num_tokens": 755040319.0, "step": 19880 }, { "epoch": 2.5290675486579315, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.779090642929077, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8837137222290039, "num_tokens": 755073410.0, "step": 19881 }, { "epoch": 2.529194758936522, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.634202480316162, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8831087350845337, "num_tokens": 755112279.0, "step": 19882 }, { "epoch": 2.5293219692151125, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7972970008850098, "learning_rate": 1e-05, "loss": 0.4801, "mean_token_accuracy": 0.8608327507972717, "num_tokens": 755146904.0, "step": 19883 }, { "epoch": 2.529449179493703, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.64593768119812, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8713400363922119, "num_tokens": 755179881.0, "step": 19884 }, { "epoch": 2.5295763897722936, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7875778675079346, "learning_rate": 1e-05, "loss": 0.3902, "mean_token_accuracy": 0.8911129832267761, "num_tokens": 755214504.0, "step": 19885 }, { "epoch": 2.529703600050884, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.625823736190796, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8782285451889038, "num_tokens": 755254229.0, "step": 19886 }, { "epoch": 2.5298308103294747, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.786151885986328, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8741757869720459, "num_tokens": 755296150.0, "step": 19887 }, { "epoch": 2.529958020608065, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6274685859680176, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8851111531257629, "num_tokens": 755336508.0, "step": 19888 }, { "epoch": 2.5300852308866557, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7766075134277344, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8677124977111816, "num_tokens": 755376863.0, "step": 19889 }, { "epoch": 2.5302124411652462, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.631964921951294, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8874985575675964, "num_tokens": 755414209.0, "step": 19890 }, { "epoch": 2.5303396514438368, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.794224500656128, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8628255724906921, "num_tokens": 755454281.0, "step": 19891 }, { "epoch": 2.5304668617224273, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6194841861724854, "learning_rate": 1e-05, "loss": 0.3667, "mean_token_accuracy": 0.8965370655059814, "num_tokens": 755492441.0, "step": 19892 }, { "epoch": 2.530594072001018, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8078837394714355, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8785598874092102, "num_tokens": 755526264.0, "step": 19893 }, { "epoch": 2.5307212822796084, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.63834285736084, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8839178085327148, "num_tokens": 755560819.0, "step": 19894 }, { "epoch": 2.5308484925581984, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8104496002197266, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8778436779975891, "num_tokens": 755592103.0, "step": 19895 }, { "epoch": 2.5309757028367894, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.618098735809326, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8892739415168762, "num_tokens": 755631864.0, "step": 19896 }, { "epoch": 2.5311029131153795, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8111114501953125, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8734851479530334, "num_tokens": 755672555.0, "step": 19897 }, { "epoch": 2.5312301233939705, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6284615993499756, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.8900820016860962, "num_tokens": 755705606.0, "step": 19898 }, { "epoch": 2.5313573336725606, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.831732988357544, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8750775456428528, "num_tokens": 755738844.0, "step": 19899 }, { "epoch": 2.5314845439511515, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6179957389831543, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8886173963546753, "num_tokens": 755774164.0, "step": 19900 }, { "epoch": 2.5316117542297416, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8273518085479736, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8711473345756531, "num_tokens": 755816848.0, "step": 19901 }, { "epoch": 2.531738964508332, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6225802898406982, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8760106563568115, "num_tokens": 755854281.0, "step": 19902 }, { "epoch": 2.5318661747869227, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8303422927856445, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8639950156211853, "num_tokens": 755893769.0, "step": 19903 }, { "epoch": 2.531993385065513, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6342763900756836, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8762421607971191, "num_tokens": 755931690.0, "step": 19904 }, { "epoch": 2.5321205953441037, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.826530694961548, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8816086649894714, "num_tokens": 755967177.0, "step": 19905 }, { "epoch": 2.5322478056226942, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6350505352020264, "learning_rate": 1e-05, "loss": 0.3846, "mean_token_accuracy": 0.8894811868667603, "num_tokens": 755998440.0, "step": 19906 }, { "epoch": 2.5323750159012848, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.815234422683716, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8761579394340515, "num_tokens": 756038049.0, "step": 19907 }, { "epoch": 2.5325022261798753, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.635488510131836, "learning_rate": 1e-05, "loss": 0.3856, "mean_token_accuracy": 0.8910443782806396, "num_tokens": 756073185.0, "step": 19908 }, { "epoch": 2.532629436458466, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8041937351226807, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8714372515678406, "num_tokens": 756113439.0, "step": 19909 }, { "epoch": 2.5327566467370564, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.634587049484253, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8835150003433228, "num_tokens": 756155516.0, "step": 19910 }, { "epoch": 2.532883857015647, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7998735904693604, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8793916702270508, "num_tokens": 756196685.0, "step": 19911 }, { "epoch": 2.5330110672942374, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6470165252685547, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8752275109291077, "num_tokens": 756230150.0, "step": 19912 }, { "epoch": 2.533138277572828, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7971489429473877, "learning_rate": 1e-05, "loss": 0.3946, "mean_token_accuracy": 0.8905259966850281, "num_tokens": 756270806.0, "step": 19913 }, { "epoch": 2.5332654878514185, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6541101932525635, "learning_rate": 1e-05, "loss": 0.3797, "mean_token_accuracy": 0.8906461596488953, "num_tokens": 756300387.0, "step": 19914 }, { "epoch": 2.533392698130009, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7871506214141846, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8829102516174316, "num_tokens": 756335436.0, "step": 19915 }, { "epoch": 2.5335199084085995, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6575913429260254, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8677306175231934, "num_tokens": 756371942.0, "step": 19916 }, { "epoch": 2.53364711868719, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.759279489517212, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8818172216415405, "num_tokens": 756416102.0, "step": 19917 }, { "epoch": 2.5337743289657806, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.667250633239746, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8671941757202148, "num_tokens": 756457066.0, "step": 19918 }, { "epoch": 2.533901539244371, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.752793312072754, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8805295825004578, "num_tokens": 756493977.0, "step": 19919 }, { "epoch": 2.534028749522961, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6658358573913574, "learning_rate": 1e-05, "loss": 0.4711, "mean_token_accuracy": 0.8593755960464478, "num_tokens": 756535926.0, "step": 19920 }, { "epoch": 2.534155959801552, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.733442544937134, "learning_rate": 1e-05, "loss": 0.3879, "mean_token_accuracy": 0.8939387798309326, "num_tokens": 756578406.0, "step": 19921 }, { "epoch": 2.5342831700801423, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6870639324188232, "learning_rate": 1e-05, "loss": 0.4651, "mean_token_accuracy": 0.8629662990570068, "num_tokens": 756615104.0, "step": 19922 }, { "epoch": 2.5344103803587332, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.72809100151062, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8716921210289001, "num_tokens": 756652059.0, "step": 19923 }, { "epoch": 2.5345375906373233, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6652591228485107, "learning_rate": 1e-05, "loss": 0.3698, "mean_token_accuracy": 0.8971823453903198, "num_tokens": 756688640.0, "step": 19924 }, { "epoch": 2.534664800915914, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7268009185791016, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.8905825614929199, "num_tokens": 756721596.0, "step": 19925 }, { "epoch": 2.5347920111945044, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6810033321380615, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8677551746368408, "num_tokens": 756757663.0, "step": 19926 }, { "epoch": 2.534919221473095, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.704488754272461, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8809900879859924, "num_tokens": 756802021.0, "step": 19927 }, { "epoch": 2.5350464317516854, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6722323894500732, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.869930624961853, "num_tokens": 756842039.0, "step": 19928 }, { "epoch": 2.535173642030276, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7173426151275635, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8803023099899292, "num_tokens": 756875788.0, "step": 19929 }, { "epoch": 2.5353008523088665, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6687417030334473, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8885906934738159, "num_tokens": 756913292.0, "step": 19930 }, { "epoch": 2.535428062587457, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7055821418762207, "learning_rate": 1e-05, "loss": 0.3869, "mean_token_accuracy": 0.8961115479469299, "num_tokens": 756953005.0, "step": 19931 }, { "epoch": 2.5355552728660475, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6823856830596924, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8813040256500244, "num_tokens": 756987656.0, "step": 19932 }, { "epoch": 2.535682483144638, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.714228868484497, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8868295550346375, "num_tokens": 757021040.0, "step": 19933 }, { "epoch": 2.5358096934232286, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.676669120788574, "learning_rate": 1e-05, "loss": 0.3797, "mean_token_accuracy": 0.8924895524978638, "num_tokens": 757053039.0, "step": 19934 }, { "epoch": 2.535936903701819, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7212674617767334, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8734966516494751, "num_tokens": 757090057.0, "step": 19935 }, { "epoch": 2.5360641139804097, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.659824848175049, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8839716911315918, "num_tokens": 757129537.0, "step": 19936 }, { "epoch": 2.536191324259, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7255752086639404, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8892186284065247, "num_tokens": 757164892.0, "step": 19937 }, { "epoch": 2.5363185345375907, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6611440181732178, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8730419874191284, "num_tokens": 757204717.0, "step": 19938 }, { "epoch": 2.5364457448161812, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.743121385574341, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8765897154808044, "num_tokens": 757239725.0, "step": 19939 }, { "epoch": 2.5365729550947718, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6488993167877197, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8623811602592468, "num_tokens": 757280356.0, "step": 19940 }, { "epoch": 2.5367001653733623, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.753307819366455, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8744849562644958, "num_tokens": 757316498.0, "step": 19941 }, { "epoch": 2.536827375651953, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.656095266342163, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.874813437461853, "num_tokens": 757351675.0, "step": 19942 }, { "epoch": 2.5369545859305433, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7377872467041016, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8754650354385376, "num_tokens": 757393310.0, "step": 19943 }, { "epoch": 2.537081796209134, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6631293296813965, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8703767657279968, "num_tokens": 757429800.0, "step": 19944 }, { "epoch": 2.537209006487724, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7455520629882812, "learning_rate": 1e-05, "loss": 0.3807, "mean_token_accuracy": 0.893596351146698, "num_tokens": 757463456.0, "step": 19945 }, { "epoch": 2.537336216766315, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.676835060119629, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8674832582473755, "num_tokens": 757496977.0, "step": 19946 }, { "epoch": 2.537463427044905, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.719302177429199, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8858091831207275, "num_tokens": 757536003.0, "step": 19947 }, { "epoch": 2.537590637323496, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.672785520553589, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8719291090965271, "num_tokens": 757575017.0, "step": 19948 }, { "epoch": 2.537717847602086, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7187161445617676, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8785606622695923, "num_tokens": 757617098.0, "step": 19949 }, { "epoch": 2.5378450578806766, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.683608055114746, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8685946464538574, "num_tokens": 757652343.0, "step": 19950 }, { "epoch": 2.537972268159267, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.71203875541687, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8806751370429993, "num_tokens": 757691021.0, "step": 19951 }, { "epoch": 2.5380994784378577, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6695849895477295, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8845576047897339, "num_tokens": 757727805.0, "step": 19952 }, { "epoch": 2.538226688716448, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.714571952819824, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8718774914741516, "num_tokens": 757767979.0, "step": 19953 }, { "epoch": 2.5383538989950387, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6726815700531006, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8810628652572632, "num_tokens": 757807488.0, "step": 19954 }, { "epoch": 2.5384811092736292, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.736818552017212, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8694124221801758, "num_tokens": 757838298.0, "step": 19955 }, { "epoch": 2.5386083195522198, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6630892753601074, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8743714094161987, "num_tokens": 757875016.0, "step": 19956 }, { "epoch": 2.5387355298308103, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.716890811920166, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8834147453308105, "num_tokens": 757921306.0, "step": 19957 }, { "epoch": 2.538862740109401, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6688077449798584, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8822410106658936, "num_tokens": 757956555.0, "step": 19958 }, { "epoch": 2.5389899503879914, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7258729934692383, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8716999888420105, "num_tokens": 757990433.0, "step": 19959 }, { "epoch": 2.539117160666582, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.655747175216675, "learning_rate": 1e-05, "loss": 0.3915, "mean_token_accuracy": 0.8909229040145874, "num_tokens": 758026606.0, "step": 19960 }, { "epoch": 2.5392443709451724, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7275846004486084, "learning_rate": 1e-05, "loss": 0.377, "mean_token_accuracy": 0.8933273553848267, "num_tokens": 758060995.0, "step": 19961 }, { "epoch": 2.539371581223763, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6546084880828857, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8845891952514648, "num_tokens": 758097479.0, "step": 19962 }, { "epoch": 2.5394987915023535, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7401669025421143, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8799350261688232, "num_tokens": 758134541.0, "step": 19963 }, { "epoch": 2.539626001780944, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.642320394515991, "learning_rate": 1e-05, "loss": 0.3872, "mean_token_accuracy": 0.889008104801178, "num_tokens": 758173885.0, "step": 19964 }, { "epoch": 2.5397532120595345, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.734600782394409, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8696392774581909, "num_tokens": 758219394.0, "step": 19965 }, { "epoch": 2.539880422338125, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6494574546813965, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8765208721160889, "num_tokens": 758261146.0, "step": 19966 }, { "epoch": 2.5400076326167156, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.735349178314209, "learning_rate": 1e-05, "loss": 0.3676, "mean_token_accuracy": 0.8980157375335693, "num_tokens": 758297935.0, "step": 19967 }, { "epoch": 2.5401348428953057, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6676270961761475, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8720408082008362, "num_tokens": 758328579.0, "step": 19968 }, { "epoch": 2.5402620531738966, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.740403890609741, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.876556396484375, "num_tokens": 758361731.0, "step": 19969 }, { "epoch": 2.5403892634524867, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.667351245880127, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8840867877006531, "num_tokens": 758394651.0, "step": 19970 }, { "epoch": 2.5405164737310777, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.722127914428711, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8793185949325562, "num_tokens": 758436455.0, "step": 19971 }, { "epoch": 2.540643684009668, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.656259059906006, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8832293152809143, "num_tokens": 758471880.0, "step": 19972 }, { "epoch": 2.5407708942882588, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.722264528274536, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8802452087402344, "num_tokens": 758514354.0, "step": 19973 }, { "epoch": 2.540898104566849, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.657471179962158, "learning_rate": 1e-05, "loss": 0.3975, "mean_token_accuracy": 0.8852697014808655, "num_tokens": 758554263.0, "step": 19974 }, { "epoch": 2.5410253148454394, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7364509105682373, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8784549236297607, "num_tokens": 758588907.0, "step": 19975 }, { "epoch": 2.54115252512403, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6790666580200195, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.873619556427002, "num_tokens": 758618573.0, "step": 19976 }, { "epoch": 2.5412797354026204, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.720898389816284, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8803755044937134, "num_tokens": 758654315.0, "step": 19977 }, { "epoch": 2.541406945681211, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.677675724029541, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8736869096755981, "num_tokens": 758690670.0, "step": 19978 }, { "epoch": 2.5415341559598015, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.710752487182617, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8811783790588379, "num_tokens": 758725686.0, "step": 19979 }, { "epoch": 2.541661366238392, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.684028148651123, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8836346864700317, "num_tokens": 758757002.0, "step": 19980 }, { "epoch": 2.5417885765169825, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.705465316772461, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8839292526245117, "num_tokens": 758793695.0, "step": 19981 }, { "epoch": 2.541915786795573, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6611969470977783, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.876539409160614, "num_tokens": 758837278.0, "step": 19982 }, { "epoch": 2.5420429970741636, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7183241844177246, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8771113157272339, "num_tokens": 758871252.0, "step": 19983 }, { "epoch": 2.542170207352754, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6625964641571045, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8784363865852356, "num_tokens": 758912250.0, "step": 19984 }, { "epoch": 2.5422974176313446, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.715453624725342, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8799464702606201, "num_tokens": 758951773.0, "step": 19985 }, { "epoch": 2.542424627909935, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.664898633956909, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8888389468193054, "num_tokens": 758989054.0, "step": 19986 }, { "epoch": 2.5425518381885257, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.712449073791504, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.8858870267868042, "num_tokens": 759026917.0, "step": 19987 }, { "epoch": 2.5426790484671162, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.665832996368408, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8779284954071045, "num_tokens": 759065752.0, "step": 19988 }, { "epoch": 2.5428062587457068, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.720526933670044, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8876418471336365, "num_tokens": 759100650.0, "step": 19989 }, { "epoch": 2.5429334690242973, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6618449687957764, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8680732250213623, "num_tokens": 759138274.0, "step": 19990 }, { "epoch": 2.543060679302888, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.727750539779663, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8675121068954468, "num_tokens": 759176060.0, "step": 19991 }, { "epoch": 2.5431878895814783, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6652450561523438, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8878566026687622, "num_tokens": 759208323.0, "step": 19992 }, { "epoch": 2.5433150998600684, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.736660957336426, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8926990628242493, "num_tokens": 759239180.0, "step": 19993 }, { "epoch": 2.5434423101386594, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6446354389190674, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.883746862411499, "num_tokens": 759278489.0, "step": 19994 }, { "epoch": 2.5435695204172495, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7720892429351807, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8788284063339233, "num_tokens": 759315254.0, "step": 19995 }, { "epoch": 2.5436967306958405, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6434006690979004, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8821996450424194, "num_tokens": 759345969.0, "step": 19996 }, { "epoch": 2.5438239409744305, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7538843154907227, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8761591911315918, "num_tokens": 759386530.0, "step": 19997 }, { "epoch": 2.5439511512530215, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.634547472000122, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8758953213691711, "num_tokens": 759428755.0, "step": 19998 }, { "epoch": 2.5440783615316116, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7490365505218506, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8780726194381714, "num_tokens": 759471335.0, "step": 19999 }, { "epoch": 2.544205571810202, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6304068565368652, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8805885314941406, "num_tokens": 759512541.0, "step": 20000 }, { "epoch": 2.5443327820887927, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.763580083847046, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8751407265663147, "num_tokens": 759551015.0, "step": 20001 }, { "epoch": 2.544459992367383, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6336939334869385, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8702928423881531, "num_tokens": 759588698.0, "step": 20002 }, { "epoch": 2.5445872026459737, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7731306552886963, "learning_rate": 1e-05, "loss": 0.4684, "mean_token_accuracy": 0.8639624118804932, "num_tokens": 759630180.0, "step": 20003 }, { "epoch": 2.5447144129245642, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6497788429260254, "learning_rate": 1e-05, "loss": 0.4826, "mean_token_accuracy": 0.8602718114852905, "num_tokens": 759664416.0, "step": 20004 }, { "epoch": 2.5448416232031548, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.751844644546509, "learning_rate": 1e-05, "loss": 0.4508, "mean_token_accuracy": 0.8693997859954834, "num_tokens": 759708868.0, "step": 20005 }, { "epoch": 2.5449688334817453, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.645565986633301, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8813642859458923, "num_tokens": 759742431.0, "step": 20006 }, { "epoch": 2.545096043760336, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.755632162094116, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8845452070236206, "num_tokens": 759776504.0, "step": 20007 }, { "epoch": 2.5452232540389264, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.65281343460083, "learning_rate": 1e-05, "loss": 0.3829, "mean_token_accuracy": 0.8920334577560425, "num_tokens": 759808246.0, "step": 20008 }, { "epoch": 2.545350464317517, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7381675243377686, "learning_rate": 1e-05, "loss": 0.3802, "mean_token_accuracy": 0.8938708305358887, "num_tokens": 759846654.0, "step": 20009 }, { "epoch": 2.5454776745961074, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.666341543197632, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8766813278198242, "num_tokens": 759882594.0, "step": 20010 }, { "epoch": 2.545604884874698, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.733351707458496, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8840758204460144, "num_tokens": 759923557.0, "step": 20011 }, { "epoch": 2.5457320951532885, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.648264169692993, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8735350370407104, "num_tokens": 759961152.0, "step": 20012 }, { "epoch": 2.545859305431879, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7482988834381104, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8803808093070984, "num_tokens": 759995505.0, "step": 20013 }, { "epoch": 2.5459865157104695, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6505401134490967, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8790494799613953, "num_tokens": 760033910.0, "step": 20014 }, { "epoch": 2.54611372598906, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.742213010787964, "learning_rate": 1e-05, "loss": 0.3808, "mean_token_accuracy": 0.8925371170043945, "num_tokens": 760067605.0, "step": 20015 }, { "epoch": 2.5462409362676506, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.6678805351257324, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8642139434814453, "num_tokens": 760101151.0, "step": 20016 }, { "epoch": 2.546368146546241, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7393181324005127, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8692430853843689, "num_tokens": 760140253.0, "step": 20017 }, { "epoch": 2.546495356824831, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.654787302017212, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8797444105148315, "num_tokens": 760175988.0, "step": 20018 }, { "epoch": 2.546622567103422, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7511723041534424, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8762423992156982, "num_tokens": 760211734.0, "step": 20019 }, { "epoch": 2.5467497773820122, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6478688716888428, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8743877410888672, "num_tokens": 760248887.0, "step": 20020 }, { "epoch": 2.546876987660603, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.754645824432373, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8773822784423828, "num_tokens": 760284836.0, "step": 20021 }, { "epoch": 2.5470041979391933, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6478989124298096, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8784337043762207, "num_tokens": 760327328.0, "step": 20022 }, { "epoch": 2.547131408217784, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7573254108428955, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8818364143371582, "num_tokens": 760362876.0, "step": 20023 }, { "epoch": 2.5472586184963744, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6387317180633545, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8849133253097534, "num_tokens": 760399652.0, "step": 20024 }, { "epoch": 2.547385828774965, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7611961364746094, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8867641687393188, "num_tokens": 760436346.0, "step": 20025 }, { "epoch": 2.5475130390535554, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.699521541595459, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8777273297309875, "num_tokens": 760468540.0, "step": 20026 }, { "epoch": 2.547640249332146, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7392849922180176, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8811403512954712, "num_tokens": 760504088.0, "step": 20027 }, { "epoch": 2.5477674596107365, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6642589569091797, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8782236576080322, "num_tokens": 760542543.0, "step": 20028 }, { "epoch": 2.547894669889327, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.731174945831299, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8688827753067017, "num_tokens": 760582472.0, "step": 20029 }, { "epoch": 2.5480218801679175, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6772422790527344, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8729621171951294, "num_tokens": 760618410.0, "step": 20030 }, { "epoch": 2.548149090446508, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.719200372695923, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8860565423965454, "num_tokens": 760656223.0, "step": 20031 }, { "epoch": 2.5482763007250986, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.650984764099121, "learning_rate": 1e-05, "loss": 0.3755, "mean_token_accuracy": 0.8946851491928101, "num_tokens": 760700898.0, "step": 20032 }, { "epoch": 2.548403511003689, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.734302043914795, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8796247839927673, "num_tokens": 760739608.0, "step": 20033 }, { "epoch": 2.5485307212822796, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.664904832839966, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8693420886993408, "num_tokens": 760777808.0, "step": 20034 }, { "epoch": 2.54865793156087, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7274868488311768, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8827108144760132, "num_tokens": 760817497.0, "step": 20035 }, { "epoch": 2.5487851418394607, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.669537305831909, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8757234811782837, "num_tokens": 760855699.0, "step": 20036 }, { "epoch": 2.5489123521180512, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7421987056732178, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8622617125511169, "num_tokens": 760890796.0, "step": 20037 }, { "epoch": 2.5490395623966418, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6535286903381348, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8808834552764893, "num_tokens": 760929056.0, "step": 20038 }, { "epoch": 2.5491667726752323, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7323811054229736, "learning_rate": 1e-05, "loss": 0.3832, "mean_token_accuracy": 0.8910931348800659, "num_tokens": 760963941.0, "step": 20039 }, { "epoch": 2.549293982953823, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.641017436981201, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8784162998199463, "num_tokens": 761009703.0, "step": 20040 }, { "epoch": 2.5494211932324133, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.739534854888916, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8820863962173462, "num_tokens": 761051408.0, "step": 20041 }, { "epoch": 2.549548403511004, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6452677249908447, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8706481456756592, "num_tokens": 761093764.0, "step": 20042 }, { "epoch": 2.549675613789594, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.759983777999878, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8801790475845337, "num_tokens": 761129975.0, "step": 20043 }, { "epoch": 2.549802824068185, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6364338397979736, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8720656633377075, "num_tokens": 761169845.0, "step": 20044 }, { "epoch": 2.549930034346775, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7649600505828857, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8758858442306519, "num_tokens": 761207667.0, "step": 20045 }, { "epoch": 2.550057244625366, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6246988773345947, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.880286693572998, "num_tokens": 761249649.0, "step": 20046 }, { "epoch": 2.550184454903956, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.783665180206299, "learning_rate": 1e-05, "loss": 0.5227, "mean_token_accuracy": 0.8490064740180969, "num_tokens": 761289879.0, "step": 20047 }, { "epoch": 2.5503116651825466, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6339962482452393, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.872974157333374, "num_tokens": 761331398.0, "step": 20048 }, { "epoch": 2.550438875461137, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.785924196243286, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.881184995174408, "num_tokens": 761363928.0, "step": 20049 }, { "epoch": 2.5505660857397277, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.633309841156006, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8625502586364746, "num_tokens": 761405737.0, "step": 20050 }, { "epoch": 2.550693296018318, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.795074939727783, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.8883441686630249, "num_tokens": 761435934.0, "step": 20051 }, { "epoch": 2.5508205062969087, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6395442485809326, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8825673460960388, "num_tokens": 761466700.0, "step": 20052 }, { "epoch": 2.5509477165754992, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7843995094299316, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.880485475063324, "num_tokens": 761505682.0, "step": 20053 }, { "epoch": 2.5510749268540898, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.637831926345825, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8789359331130981, "num_tokens": 761544071.0, "step": 20054 }, { "epoch": 2.5512021371326803, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7692713737487793, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8871844410896301, "num_tokens": 761586595.0, "step": 20055 }, { "epoch": 2.551329347411271, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6447629928588867, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8793845176696777, "num_tokens": 761623924.0, "step": 20056 }, { "epoch": 2.5514565576898613, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7705495357513428, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8762878179550171, "num_tokens": 761660952.0, "step": 20057 }, { "epoch": 2.551583767968452, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.635429620742798, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.8918922543525696, "num_tokens": 761701451.0, "step": 20058 }, { "epoch": 2.5517109782470424, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7844531536102295, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8808790445327759, "num_tokens": 761733638.0, "step": 20059 }, { "epoch": 2.551838188525633, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6561765670776367, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8824620246887207, "num_tokens": 761763978.0, "step": 20060 }, { "epoch": 2.5519653988042235, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.765174627304077, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8861105442047119, "num_tokens": 761802653.0, "step": 20061 }, { "epoch": 2.552092609082814, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6430225372314453, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8808680176734924, "num_tokens": 761840671.0, "step": 20062 }, { "epoch": 2.5522198193614045, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7612226009368896, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.873268723487854, "num_tokens": 761883823.0, "step": 20063 }, { "epoch": 2.552347029639995, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.64479660987854, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8801201581954956, "num_tokens": 761921597.0, "step": 20064 }, { "epoch": 2.5524742399185856, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7443177700042725, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8873685598373413, "num_tokens": 761963400.0, "step": 20065 }, { "epoch": 2.5526014501971757, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.650521755218506, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8789702653884888, "num_tokens": 762004442.0, "step": 20066 }, { "epoch": 2.5527286604757666, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.767489433288574, "learning_rate": 1e-05, "loss": 0.4638, "mean_token_accuracy": 0.8663980960845947, "num_tokens": 762040679.0, "step": 20067 }, { "epoch": 2.5528558707543567, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6262338161468506, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8765446543693542, "num_tokens": 762085249.0, "step": 20068 }, { "epoch": 2.5529830810329477, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.765204906463623, "learning_rate": 1e-05, "loss": 0.3793, "mean_token_accuracy": 0.8943647146224976, "num_tokens": 762125551.0, "step": 20069 }, { "epoch": 2.5531102913115378, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.624276876449585, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8839588165283203, "num_tokens": 762162503.0, "step": 20070 }, { "epoch": 2.5532375015901287, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.784562110900879, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8851069808006287, "num_tokens": 762197302.0, "step": 20071 }, { "epoch": 2.553364711868719, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.627581834793091, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.884326696395874, "num_tokens": 762235069.0, "step": 20072 }, { "epoch": 2.5534919221473094, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.765688180923462, "learning_rate": 1e-05, "loss": 0.3783, "mean_token_accuracy": 0.8914952278137207, "num_tokens": 762272014.0, "step": 20073 }, { "epoch": 2.5536191324259, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.627788543701172, "learning_rate": 1e-05, "loss": 0.3796, "mean_token_accuracy": 0.8914783000946045, "num_tokens": 762312840.0, "step": 20074 }, { "epoch": 2.5537463427044904, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.791240930557251, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.878092348575592, "num_tokens": 762347997.0, "step": 20075 }, { "epoch": 2.553873552983081, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6423587799072266, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8732564449310303, "num_tokens": 762390706.0, "step": 20076 }, { "epoch": 2.5540007632616715, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.788477897644043, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8781663179397583, "num_tokens": 762425894.0, "step": 20077 }, { "epoch": 2.554127973540262, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6404149532318115, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8651273250579834, "num_tokens": 762461962.0, "step": 20078 }, { "epoch": 2.5542551838188525, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7727153301239014, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8771039247512817, "num_tokens": 762500128.0, "step": 20079 }, { "epoch": 2.554382394097443, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6309266090393066, "learning_rate": 1e-05, "loss": 0.3705, "mean_token_accuracy": 0.8971529006958008, "num_tokens": 762536256.0, "step": 20080 }, { "epoch": 2.5545096043760336, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.805039405822754, "learning_rate": 1e-05, "loss": 0.4962, "mean_token_accuracy": 0.8577998876571655, "num_tokens": 762569797.0, "step": 20081 }, { "epoch": 2.554636814654624, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.67755126953125, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8771384954452515, "num_tokens": 762606849.0, "step": 20082 }, { "epoch": 2.5547640249332146, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7556841373443604, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8711230754852295, "num_tokens": 762649526.0, "step": 20083 }, { "epoch": 2.554891235211805, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.645343780517578, "learning_rate": 1e-05, "loss": 0.4047, "mean_token_accuracy": 0.8812938928604126, "num_tokens": 762686240.0, "step": 20084 }, { "epoch": 2.5550184454903957, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.741302251815796, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8695504069328308, "num_tokens": 762727908.0, "step": 20085 }, { "epoch": 2.5551456557689862, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6449265480041504, "learning_rate": 1e-05, "loss": 0.38, "mean_token_accuracy": 0.8917524814605713, "num_tokens": 762765966.0, "step": 20086 }, { "epoch": 2.5552728660475768, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.752443313598633, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8797129392623901, "num_tokens": 762806031.0, "step": 20087 }, { "epoch": 2.5554000763261673, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6334280967712402, "learning_rate": 1e-05, "loss": 0.3896, "mean_token_accuracy": 0.8881816864013672, "num_tokens": 762847531.0, "step": 20088 }, { "epoch": 2.555527286604758, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7762362957000732, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8822500109672546, "num_tokens": 762878409.0, "step": 20089 }, { "epoch": 2.5556544968833483, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6303234100341797, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.878937840461731, "num_tokens": 762918964.0, "step": 20090 }, { "epoch": 2.5557817071619384, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7627689838409424, "learning_rate": 1e-05, "loss": 0.3945, "mean_token_accuracy": 0.8884358406066895, "num_tokens": 762958962.0, "step": 20091 }, { "epoch": 2.5559089174405294, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.655143976211548, "learning_rate": 1e-05, "loss": 0.4655, "mean_token_accuracy": 0.8658914566040039, "num_tokens": 762996062.0, "step": 20092 }, { "epoch": 2.5560361277191195, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7590622901916504, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8808231949806213, "num_tokens": 763033387.0, "step": 20093 }, { "epoch": 2.5561633379977104, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.661001205444336, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8756100535392761, "num_tokens": 763067826.0, "step": 20094 }, { "epoch": 2.5562905482763005, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7594990730285645, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8864955306053162, "num_tokens": 763100442.0, "step": 20095 }, { "epoch": 2.5564177585548915, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.631319284439087, "learning_rate": 1e-05, "loss": 0.3768, "mean_token_accuracy": 0.8925538659095764, "num_tokens": 763137659.0, "step": 20096 }, { "epoch": 2.5565449688334816, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7456696033477783, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8891096711158752, "num_tokens": 763180000.0, "step": 20097 }, { "epoch": 2.556672179112072, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.655637502670288, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8680108785629272, "num_tokens": 763218863.0, "step": 20098 }, { "epoch": 2.5567993893906626, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7478673458099365, "learning_rate": 1e-05, "loss": 0.3866, "mean_token_accuracy": 0.8909038305282593, "num_tokens": 763252973.0, "step": 20099 }, { "epoch": 2.556926599669253, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.669743299484253, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8631227016448975, "num_tokens": 763288781.0, "step": 20100 }, { "epoch": 2.5570538099478437, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7638633251190186, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8743147850036621, "num_tokens": 763321839.0, "step": 20101 }, { "epoch": 2.5571810202264342, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6558499336242676, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8759689331054688, "num_tokens": 763357607.0, "step": 20102 }, { "epoch": 2.5573082305050248, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.745159387588501, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8750585913658142, "num_tokens": 763399489.0, "step": 20103 }, { "epoch": 2.5574354407836153, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6682674884796143, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8630683422088623, "num_tokens": 763441256.0, "step": 20104 }, { "epoch": 2.557562651062206, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7590880393981934, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8708959221839905, "num_tokens": 763475019.0, "step": 20105 }, { "epoch": 2.5576898613407963, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6530609130859375, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8794373273849487, "num_tokens": 763510456.0, "step": 20106 }, { "epoch": 2.557817071619387, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7505626678466797, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.87433922290802, "num_tokens": 763546181.0, "step": 20107 }, { "epoch": 2.5579442818979774, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6714718341827393, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8835992217063904, "num_tokens": 763578763.0, "step": 20108 }, { "epoch": 2.558071492176568, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7399768829345703, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8672986030578613, "num_tokens": 763615615.0, "step": 20109 }, { "epoch": 2.5581987024551585, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.643625497817993, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8905045986175537, "num_tokens": 763655546.0, "step": 20110 }, { "epoch": 2.558325912733749, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7503085136413574, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.873623251914978, "num_tokens": 763695400.0, "step": 20111 }, { "epoch": 2.5584531230123395, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6786859035491943, "learning_rate": 1e-05, "loss": 0.4757, "mean_token_accuracy": 0.8613208532333374, "num_tokens": 763725968.0, "step": 20112 }, { "epoch": 2.55858033329093, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7478933334350586, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8808177709579468, "num_tokens": 763763708.0, "step": 20113 }, { "epoch": 2.5587075435695206, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6636903285980225, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8665597438812256, "num_tokens": 763800836.0, "step": 20114 }, { "epoch": 2.558834753848111, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.734292507171631, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8714514970779419, "num_tokens": 763842119.0, "step": 20115 }, { "epoch": 2.558961964126701, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6675684452056885, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8779023885726929, "num_tokens": 763876797.0, "step": 20116 }, { "epoch": 2.559089174405292, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7381813526153564, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8752173185348511, "num_tokens": 763913131.0, "step": 20117 }, { "epoch": 2.5592163846838822, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.653092622756958, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8800413608551025, "num_tokens": 763952430.0, "step": 20118 }, { "epoch": 2.559343594962473, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.735774517059326, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8797454833984375, "num_tokens": 763988978.0, "step": 20119 }, { "epoch": 2.5594708052410633, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6559205055236816, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8758697509765625, "num_tokens": 764027473.0, "step": 20120 }, { "epoch": 2.559598015519654, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7349350452423096, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8834397792816162, "num_tokens": 764063681.0, "step": 20121 }, { "epoch": 2.5597252257982444, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6531269550323486, "learning_rate": 1e-05, "loss": 0.4212, "mean_token_accuracy": 0.8846117258071899, "num_tokens": 764102463.0, "step": 20122 }, { "epoch": 2.559852436076835, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.728414535522461, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8887493014335632, "num_tokens": 764141912.0, "step": 20123 }, { "epoch": 2.5599796463554254, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6538898944854736, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.889844536781311, "num_tokens": 764176448.0, "step": 20124 }, { "epoch": 2.560106856634016, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.730224132537842, "learning_rate": 1e-05, "loss": 0.4697, "mean_token_accuracy": 0.8640908002853394, "num_tokens": 764217287.0, "step": 20125 }, { "epoch": 2.5602340669126065, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6537299156188965, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8784360885620117, "num_tokens": 764255593.0, "step": 20126 }, { "epoch": 2.560361277191197, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.738564968109131, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8760887384414673, "num_tokens": 764294624.0, "step": 20127 }, { "epoch": 2.5604884874697875, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6566524505615234, "learning_rate": 1e-05, "loss": 0.386, "mean_token_accuracy": 0.8898704051971436, "num_tokens": 764328495.0, "step": 20128 }, { "epoch": 2.560615697748378, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7380974292755127, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8875130414962769, "num_tokens": 764359986.0, "step": 20129 }, { "epoch": 2.5607429080269686, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.663346290588379, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8730895519256592, "num_tokens": 764394039.0, "step": 20130 }, { "epoch": 2.560870118305559, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.736924409866333, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8715112209320068, "num_tokens": 764436106.0, "step": 20131 }, { "epoch": 2.5609973285841496, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.654179573059082, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8830015659332275, "num_tokens": 764470004.0, "step": 20132 }, { "epoch": 2.56112453886274, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7328906059265137, "learning_rate": 1e-05, "loss": 0.3983, "mean_token_accuracy": 0.8867474794387817, "num_tokens": 764507679.0, "step": 20133 }, { "epoch": 2.5612517491413307, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6434147357940674, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.8929349184036255, "num_tokens": 764548818.0, "step": 20134 }, { "epoch": 2.561378959419921, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.744994878768921, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8759938478469849, "num_tokens": 764590289.0, "step": 20135 }, { "epoch": 2.5615061696985117, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.647597312927246, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8858418464660645, "num_tokens": 764624210.0, "step": 20136 }, { "epoch": 2.5616333799771023, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7636170387268066, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8762833476066589, "num_tokens": 764657071.0, "step": 20137 }, { "epoch": 2.561760590255693, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6460390090942383, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8802406787872314, "num_tokens": 764691439.0, "step": 20138 }, { "epoch": 2.5618878005342833, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7530276775360107, "learning_rate": 1e-05, "loss": 0.3735, "mean_token_accuracy": 0.8951127529144287, "num_tokens": 764731551.0, "step": 20139 }, { "epoch": 2.562015010812874, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1484832763671875e-05, "grad_norm": 2.650970935821533, "learning_rate": 1e-05, "loss": 0.3822, "mean_token_accuracy": 0.8960283994674683, "num_tokens": 764763233.0, "step": 20140 }, { "epoch": 2.562142221091464, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.750018358230591, "learning_rate": 1e-05, "loss": 0.3608, "mean_token_accuracy": 0.8995488882064819, "num_tokens": 764796937.0, "step": 20141 }, { "epoch": 2.562269431370055, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.655677556991577, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8832004070281982, "num_tokens": 764831808.0, "step": 20142 }, { "epoch": 2.562396641648645, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.790173053741455, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8802747130393982, "num_tokens": 764863288.0, "step": 20143 }, { "epoch": 2.562523851927236, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6250905990600586, "learning_rate": 1e-05, "loss": 0.3816, "mean_token_accuracy": 0.8902915120124817, "num_tokens": 764905417.0, "step": 20144 }, { "epoch": 2.562651062205826, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7854151725769043, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8817976713180542, "num_tokens": 764942058.0, "step": 20145 }, { "epoch": 2.5627782724844166, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6166155338287354, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8761821389198303, "num_tokens": 764983652.0, "step": 20146 }, { "epoch": 2.562905482763007, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.794388771057129, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.884030282497406, "num_tokens": 765023966.0, "step": 20147 }, { "epoch": 2.5630326930415976, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.634766101837158, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8758307695388794, "num_tokens": 765055847.0, "step": 20148 }, { "epoch": 2.563159903320188, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.802847146987915, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8703655004501343, "num_tokens": 765094387.0, "step": 20149 }, { "epoch": 2.5632871135987787, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6325266361236572, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8706788420677185, "num_tokens": 765128193.0, "step": 20150 }, { "epoch": 2.5634143238773692, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7953639030456543, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8788094520568848, "num_tokens": 765166550.0, "step": 20151 }, { "epoch": 2.5635415341559598, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.638545274734497, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8773238062858582, "num_tokens": 765204228.0, "step": 20152 }, { "epoch": 2.5636687444345503, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.781254529953003, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.883827805519104, "num_tokens": 765247148.0, "step": 20153 }, { "epoch": 2.563795954713141, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6278023719787598, "learning_rate": 1e-05, "loss": 0.368, "mean_token_accuracy": 0.8917852640151978, "num_tokens": 765287902.0, "step": 20154 }, { "epoch": 2.5639231649917313, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.784141778945923, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.8847272992134094, "num_tokens": 765331118.0, "step": 20155 }, { "epoch": 2.564050375270322, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6366019248962402, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8891350030899048, "num_tokens": 765369942.0, "step": 20156 }, { "epoch": 2.5641775855489124, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.786310911178589, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8796491026878357, "num_tokens": 765407664.0, "step": 20157 }, { "epoch": 2.564304795827503, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6213483810424805, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8789600133895874, "num_tokens": 765453303.0, "step": 20158 }, { "epoch": 2.5644320061060935, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.793704032897949, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8814862370491028, "num_tokens": 765492558.0, "step": 20159 }, { "epoch": 2.564559216384684, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6439425945281982, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8571661114692688, "num_tokens": 765529964.0, "step": 20160 }, { "epoch": 2.5646864266632745, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.773857355117798, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8926873803138733, "num_tokens": 765570158.0, "step": 20161 }, { "epoch": 2.564813636941865, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6459743976593018, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8586512207984924, "num_tokens": 765609304.0, "step": 20162 }, { "epoch": 2.5649408472204556, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.763667345046997, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8777248859405518, "num_tokens": 765651124.0, "step": 20163 }, { "epoch": 2.5650680574990457, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.639883518218994, "learning_rate": 1e-05, "loss": 0.3799, "mean_token_accuracy": 0.8925889730453491, "num_tokens": 765685544.0, "step": 20164 }, { "epoch": 2.5651952677776366, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7547287940979004, "learning_rate": 1e-05, "loss": 0.3935, "mean_token_accuracy": 0.8906599283218384, "num_tokens": 765726120.0, "step": 20165 }, { "epoch": 2.5653224780562267, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.650012731552124, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8660789728164673, "num_tokens": 765764145.0, "step": 20166 }, { "epoch": 2.5654496883348177, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7572202682495117, "learning_rate": 1e-05, "loss": 0.3764, "mean_token_accuracy": 0.8916534185409546, "num_tokens": 765795530.0, "step": 20167 }, { "epoch": 2.5655768986134078, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.648577928543091, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8755422830581665, "num_tokens": 765831793.0, "step": 20168 }, { "epoch": 2.5657041088919987, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.736325979232788, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8725615739822388, "num_tokens": 765876724.0, "step": 20169 }, { "epoch": 2.565831319170589, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.643401622772217, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8923836946487427, "num_tokens": 765917604.0, "step": 20170 }, { "epoch": 2.5659585294491793, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.752042531967163, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8811566829681396, "num_tokens": 765953717.0, "step": 20171 }, { "epoch": 2.56608573972777, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.658827304840088, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8764586448669434, "num_tokens": 765989967.0, "step": 20172 }, { "epoch": 2.5662129500063604, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7388381958007812, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8841079473495483, "num_tokens": 766029276.0, "step": 20173 }, { "epoch": 2.566340160284951, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6517608165740967, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.8880383968353271, "num_tokens": 766064819.0, "step": 20174 }, { "epoch": 2.5664673705635415, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.74000883102417, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.875352144241333, "num_tokens": 766103621.0, "step": 20175 }, { "epoch": 2.566594580842132, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6491916179656982, "learning_rate": 1e-05, "loss": 0.3871, "mean_token_accuracy": 0.8883064389228821, "num_tokens": 766138945.0, "step": 20176 }, { "epoch": 2.5667217911207225, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7440428733825684, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8751110434532166, "num_tokens": 766180771.0, "step": 20177 }, { "epoch": 2.566849001399313, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.667278528213501, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8705817461013794, "num_tokens": 766216907.0, "step": 20178 }, { "epoch": 2.5669762116779036, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.734691619873047, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8755531311035156, "num_tokens": 766256302.0, "step": 20179 }, { "epoch": 2.567103421956494, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6697330474853516, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.8818236589431763, "num_tokens": 766291255.0, "step": 20180 }, { "epoch": 2.5672306322350846, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.736997127532959, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.8842110633850098, "num_tokens": 766323895.0, "step": 20181 }, { "epoch": 2.567357842513675, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.653092861175537, "learning_rate": 1e-05, "loss": 0.3956, "mean_token_accuracy": 0.8849606513977051, "num_tokens": 766364518.0, "step": 20182 }, { "epoch": 2.5674850527922657, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.739384174346924, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.86823570728302, "num_tokens": 766402256.0, "step": 20183 }, { "epoch": 2.567612263070856, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.664581298828125, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8792996406555176, "num_tokens": 766435186.0, "step": 20184 }, { "epoch": 2.5677394733494467, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7300312519073486, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8725202083587646, "num_tokens": 766476950.0, "step": 20185 }, { "epoch": 2.5678666836280373, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6489758491516113, "learning_rate": 1e-05, "loss": 0.3844, "mean_token_accuracy": 0.8925273418426514, "num_tokens": 766517908.0, "step": 20186 }, { "epoch": 2.567993893906628, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.747225761413574, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8676924109458923, "num_tokens": 766556251.0, "step": 20187 }, { "epoch": 2.5681211041852183, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6584348678588867, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8683477640151978, "num_tokens": 766593865.0, "step": 20188 }, { "epoch": 2.5682483144638084, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7419772148132324, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8757472038269043, "num_tokens": 766627594.0, "step": 20189 }, { "epoch": 2.5683755247423994, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.641148328781128, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.8858146667480469, "num_tokens": 766669620.0, "step": 20190 }, { "epoch": 2.5685027350209895, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7501204013824463, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8867418766021729, "num_tokens": 766711553.0, "step": 20191 }, { "epoch": 2.5686299452995804, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6381678581237793, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8789836168289185, "num_tokens": 766751717.0, "step": 20192 }, { "epoch": 2.5687571555781705, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7680869102478027, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8824411034584045, "num_tokens": 766782207.0, "step": 20193 }, { "epoch": 2.5688843658567615, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.634185314178467, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8753284215927124, "num_tokens": 766822050.0, "step": 20194 }, { "epoch": 2.5690115761353516, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7635717391967773, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.868889331817627, "num_tokens": 766867713.0, "step": 20195 }, { "epoch": 2.569138786413942, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.622297525405884, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8829668760299683, "num_tokens": 766908191.0, "step": 20196 }, { "epoch": 2.5692659966925326, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.789106607437134, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8687501549720764, "num_tokens": 766945774.0, "step": 20197 }, { "epoch": 2.569393206971123, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6129519939422607, "learning_rate": 1e-05, "loss": 0.3802, "mean_token_accuracy": 0.8938258290290833, "num_tokens": 766986255.0, "step": 20198 }, { "epoch": 2.5695204172497137, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8101816177368164, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.879189133644104, "num_tokens": 767020930.0, "step": 20199 }, { "epoch": 2.5696476275283042, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.625223159790039, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8750203847885132, "num_tokens": 767059863.0, "step": 20200 }, { "epoch": 2.5697748378068948, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.815129280090332, "learning_rate": 1e-05, "loss": 0.505, "mean_token_accuracy": 0.8546476364135742, "num_tokens": 767100566.0, "step": 20201 }, { "epoch": 2.5699020480854853, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.619112014770508, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8714355826377869, "num_tokens": 767137892.0, "step": 20202 }, { "epoch": 2.570029258364076, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8245463371276855, "learning_rate": 1e-05, "loss": 0.493, "mean_token_accuracy": 0.85957932472229, "num_tokens": 767175340.0, "step": 20203 }, { "epoch": 2.5701564686426663, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.639249324798584, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.864021897315979, "num_tokens": 767209601.0, "step": 20204 }, { "epoch": 2.570283678921257, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7974579334259033, "learning_rate": 1e-05, "loss": 0.3676, "mean_token_accuracy": 0.8971145153045654, "num_tokens": 767247500.0, "step": 20205 }, { "epoch": 2.5704108891998474, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.648833751678467, "learning_rate": 1e-05, "loss": 0.4665, "mean_token_accuracy": 0.865463137626648, "num_tokens": 767285251.0, "step": 20206 }, { "epoch": 2.570538099478438, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8051397800445557, "learning_rate": 1e-05, "loss": 0.4033, "mean_token_accuracy": 0.8882121443748474, "num_tokens": 767321493.0, "step": 20207 }, { "epoch": 2.5706653097570284, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.670476198196411, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8876821994781494, "num_tokens": 767351375.0, "step": 20208 }, { "epoch": 2.570792520035619, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7686030864715576, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8778848052024841, "num_tokens": 767392990.0, "step": 20209 }, { "epoch": 2.5709197303142095, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.646779775619507, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8827903270721436, "num_tokens": 767431584.0, "step": 20210 }, { "epoch": 2.5710469405928, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.774980068206787, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8765258193016052, "num_tokens": 767468451.0, "step": 20211 }, { "epoch": 2.5711741508713906, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.645228862762451, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8729179501533508, "num_tokens": 767508133.0, "step": 20212 }, { "epoch": 2.571301361149981, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7815122604370117, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8849770426750183, "num_tokens": 767539427.0, "step": 20213 }, { "epoch": 2.571428571428571, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6543498039245605, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8782268166542053, "num_tokens": 767574842.0, "step": 20214 }, { "epoch": 2.571555781707162, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.781085729598999, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8761172294616699, "num_tokens": 767609378.0, "step": 20215 }, { "epoch": 2.5716829919857522, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6492724418640137, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8729639053344727, "num_tokens": 767642283.0, "step": 20216 }, { "epoch": 2.571810202264343, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.776257276535034, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8854868412017822, "num_tokens": 767680019.0, "step": 20217 }, { "epoch": 2.5719374125429333, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6539511680603027, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8721647262573242, "num_tokens": 767715570.0, "step": 20218 }, { "epoch": 2.572064622821524, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.754647970199585, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8792194128036499, "num_tokens": 767757257.0, "step": 20219 }, { "epoch": 2.5721918331001143, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6431355476379395, "learning_rate": 1e-05, "loss": 0.3591, "mean_token_accuracy": 0.9010542035102844, "num_tokens": 767791684.0, "step": 20220 }, { "epoch": 2.572319043378705, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.76387357711792, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8791965246200562, "num_tokens": 767831687.0, "step": 20221 }, { "epoch": 2.5724462536572954, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.633568286895752, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.8849584460258484, "num_tokens": 767871349.0, "step": 20222 }, { "epoch": 2.572573463935886, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.779545307159424, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8874990940093994, "num_tokens": 767908301.0, "step": 20223 }, { "epoch": 2.5727006742144765, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6422672271728516, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.882304847240448, "num_tokens": 767944809.0, "step": 20224 }, { "epoch": 2.572827884493067, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.773487091064453, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8732874393463135, "num_tokens": 767984629.0, "step": 20225 }, { "epoch": 2.5729550947716575, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6588120460510254, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8670977354049683, "num_tokens": 768017697.0, "step": 20226 }, { "epoch": 2.573082305050248, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7606241703033447, "learning_rate": 1e-05, "loss": 0.4431, "mean_token_accuracy": 0.8744803071022034, "num_tokens": 768060259.0, "step": 20227 }, { "epoch": 2.5732095153288386, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6474688053131104, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.871931791305542, "num_tokens": 768096248.0, "step": 20228 }, { "epoch": 2.573336725607429, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7719621658325195, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8817477822303772, "num_tokens": 768131730.0, "step": 20229 }, { "epoch": 2.5734639358860196, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.646514654159546, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8883610963821411, "num_tokens": 768167551.0, "step": 20230 }, { "epoch": 2.57359114616461, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.787829875946045, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.879104733467102, "num_tokens": 768198212.0, "step": 20231 }, { "epoch": 2.5737183564432007, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.648348331451416, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8776206374168396, "num_tokens": 768233052.0, "step": 20232 }, { "epoch": 2.573845566721791, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7685720920562744, "learning_rate": 1e-05, "loss": 0.3836, "mean_token_accuracy": 0.8908757567405701, "num_tokens": 768272961.0, "step": 20233 }, { "epoch": 2.5739727770003817, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6368234157562256, "learning_rate": 1e-05, "loss": 0.3937, "mean_token_accuracy": 0.8867759704589844, "num_tokens": 768308748.0, "step": 20234 }, { "epoch": 2.5740999872789723, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7881648540496826, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8699250221252441, "num_tokens": 768346592.0, "step": 20235 }, { "epoch": 2.574227197557563, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6207499504089355, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8832793831825256, "num_tokens": 768394367.0, "step": 20236 }, { "epoch": 2.5743544078361533, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7894017696380615, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.873439371585846, "num_tokens": 768437060.0, "step": 20237 }, { "epoch": 2.574481618114744, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6519253253936768, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8786271810531616, "num_tokens": 768470373.0, "step": 20238 }, { "epoch": 2.574608828393334, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7769315242767334, "learning_rate": 1e-05, "loss": 0.3998, "mean_token_accuracy": 0.8879003524780273, "num_tokens": 768507567.0, "step": 20239 }, { "epoch": 2.574736038671925, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.632239818572998, "learning_rate": 1e-05, "loss": 0.3983, "mean_token_accuracy": 0.8884209394454956, "num_tokens": 768547217.0, "step": 20240 }, { "epoch": 2.574863248950515, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7701528072357178, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8858596086502075, "num_tokens": 768586720.0, "step": 20241 }, { "epoch": 2.574990459229106, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.632585048675537, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8821743726730347, "num_tokens": 768627890.0, "step": 20242 }, { "epoch": 2.575117669507696, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7928709983825684, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8736423254013062, "num_tokens": 768663393.0, "step": 20243 }, { "epoch": 2.5752448797862866, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6200711727142334, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8899110555648804, "num_tokens": 768706950.0, "step": 20244 }, { "epoch": 2.575372090064877, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7996225357055664, "learning_rate": 1e-05, "loss": 0.4588, "mean_token_accuracy": 0.8700432777404785, "num_tokens": 768747391.0, "step": 20245 }, { "epoch": 2.5754993003434676, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.620201349258423, "learning_rate": 1e-05, "loss": 0.3579, "mean_token_accuracy": 0.8972682952880859, "num_tokens": 768781170.0, "step": 20246 }, { "epoch": 2.575626510622058, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8011069297790527, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.875933051109314, "num_tokens": 768820505.0, "step": 20247 }, { "epoch": 2.5757537209006487, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.627365827560425, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8778598308563232, "num_tokens": 768859647.0, "step": 20248 }, { "epoch": 2.575880931179239, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7989704608917236, "learning_rate": 1e-05, "loss": 0.376, "mean_token_accuracy": 0.8938966393470764, "num_tokens": 768894883.0, "step": 20249 }, { "epoch": 2.5760081414578297, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.619720697402954, "learning_rate": 1e-05, "loss": 0.3916, "mean_token_accuracy": 0.8915226459503174, "num_tokens": 768934648.0, "step": 20250 }, { "epoch": 2.5761353517364203, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8143630027770996, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8848642110824585, "num_tokens": 768969568.0, "step": 20251 }, { "epoch": 2.576262562015011, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.632781982421875, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8693757057189941, "num_tokens": 769009423.0, "step": 20252 }, { "epoch": 2.5763897722936013, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.819620370864868, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8834904432296753, "num_tokens": 769043812.0, "step": 20253 }, { "epoch": 2.576516982572192, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.622293472290039, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8823717832565308, "num_tokens": 769079959.0, "step": 20254 }, { "epoch": 2.5766441928507824, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8114025592803955, "learning_rate": 1e-05, "loss": 0.3879, "mean_token_accuracy": 0.8908969163894653, "num_tokens": 769119375.0, "step": 20255 }, { "epoch": 2.576771403129373, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6295056343078613, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8875153660774231, "num_tokens": 769157085.0, "step": 20256 }, { "epoch": 2.5768986134079634, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8329074382781982, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8751562833786011, "num_tokens": 769190883.0, "step": 20257 }, { "epoch": 2.577025823686554, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6276614665985107, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8818848133087158, "num_tokens": 769228370.0, "step": 20258 }, { "epoch": 2.5771530339651445, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8114700317382812, "learning_rate": 1e-05, "loss": 0.3491, "mean_token_accuracy": 0.9059921503067017, "num_tokens": 769265551.0, "step": 20259 }, { "epoch": 2.577280244243735, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6437199115753174, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8833513855934143, "num_tokens": 769299482.0, "step": 20260 }, { "epoch": 2.5774074545223256, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8140523433685303, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8800402879714966, "num_tokens": 769339165.0, "step": 20261 }, { "epoch": 2.5775346648009156, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.643956422805786, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8847347497940063, "num_tokens": 769374066.0, "step": 20262 }, { "epoch": 2.5776618750795066, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8033392429351807, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8790252208709717, "num_tokens": 769415536.0, "step": 20263 }, { "epoch": 2.5777890853580967, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6348583698272705, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8775695562362671, "num_tokens": 769454016.0, "step": 20264 }, { "epoch": 2.5779162956366877, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.798823356628418, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8828474879264832, "num_tokens": 769493150.0, "step": 20265 }, { "epoch": 2.5780435059152778, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6566789150238037, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8760292530059814, "num_tokens": 769524247.0, "step": 20266 }, { "epoch": 2.5781707161938687, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7718255519866943, "learning_rate": 1e-05, "loss": 0.3836, "mean_token_accuracy": 0.8919796943664551, "num_tokens": 769566476.0, "step": 20267 }, { "epoch": 2.578297926472459, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6539931297302246, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8772659301757812, "num_tokens": 769603474.0, "step": 20268 }, { "epoch": 2.5784251367510493, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7704572677612305, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8769199252128601, "num_tokens": 769640446.0, "step": 20269 }, { "epoch": 2.57855234702964, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.649263620376587, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8859413862228394, "num_tokens": 769679809.0, "step": 20270 }, { "epoch": 2.5786795573082304, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.764374256134033, "learning_rate": 1e-05, "loss": 0.376, "mean_token_accuracy": 0.8962033987045288, "num_tokens": 769717221.0, "step": 20271 }, { "epoch": 2.578806767586821, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6460585594177246, "learning_rate": 1e-05, "loss": 0.3686, "mean_token_accuracy": 0.8933029174804688, "num_tokens": 769749350.0, "step": 20272 }, { "epoch": 2.5789339778654115, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7739453315734863, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8707646131515503, "num_tokens": 769786604.0, "step": 20273 }, { "epoch": 2.579061188144002, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6274707317352295, "learning_rate": 1e-05, "loss": 0.371, "mean_token_accuracy": 0.8939205408096313, "num_tokens": 769827709.0, "step": 20274 }, { "epoch": 2.5791883984225925, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7837963104248047, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8782322406768799, "num_tokens": 769866713.0, "step": 20275 }, { "epoch": 2.579315608701183, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6366405487060547, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8803974390029907, "num_tokens": 769907365.0, "step": 20276 }, { "epoch": 2.5794428189797736, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7875618934631348, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8796825408935547, "num_tokens": 769945573.0, "step": 20277 }, { "epoch": 2.579570029258364, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.637660503387451, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8757495880126953, "num_tokens": 769980936.0, "step": 20278 }, { "epoch": 2.5796972395369546, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.793416738510132, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8660777807235718, "num_tokens": 770022244.0, "step": 20279 }, { "epoch": 2.579824449815545, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6255836486816406, "learning_rate": 1e-05, "loss": 0.3934, "mean_token_accuracy": 0.8866620063781738, "num_tokens": 770059078.0, "step": 20280 }, { "epoch": 2.5799516600941357, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.79091215133667, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8667205572128296, "num_tokens": 770100816.0, "step": 20281 }, { "epoch": 2.580078870372726, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.626760244369507, "learning_rate": 1e-05, "loss": 0.4014, "mean_token_accuracy": 0.8851854205131531, "num_tokens": 770138912.0, "step": 20282 }, { "epoch": 2.5802060806513167, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7902252674102783, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.886724591255188, "num_tokens": 770178915.0, "step": 20283 }, { "epoch": 2.5803332909299073, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.625946044921875, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8881441354751587, "num_tokens": 770217707.0, "step": 20284 }, { "epoch": 2.580460501208498, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8107640743255615, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.880359411239624, "num_tokens": 770251659.0, "step": 20285 }, { "epoch": 2.5805877114870883, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6207165718078613, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8751826286315918, "num_tokens": 770291612.0, "step": 20286 }, { "epoch": 2.5807149217656784, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7987589836120605, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8727690577507019, "num_tokens": 770333822.0, "step": 20287 }, { "epoch": 2.5808421320442694, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.64432692527771, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8753689527511597, "num_tokens": 770369850.0, "step": 20288 }, { "epoch": 2.5809693423228595, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.811445474624634, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8685125112533569, "num_tokens": 770408825.0, "step": 20289 }, { "epoch": 2.5810965526014504, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6096320152282715, "learning_rate": 1e-05, "loss": 0.3835, "mean_token_accuracy": 0.8928223848342896, "num_tokens": 770451320.0, "step": 20290 }, { "epoch": 2.5812237628800405, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.831937074661255, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8747943639755249, "num_tokens": 770488099.0, "step": 20291 }, { "epoch": 2.5813509731586315, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.616655111312866, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8707467913627625, "num_tokens": 770533689.0, "step": 20292 }, { "epoch": 2.5814781834372216, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8266279697418213, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8815104961395264, "num_tokens": 770570998.0, "step": 20293 }, { "epoch": 2.581605393715812, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6174113750457764, "learning_rate": 1e-05, "loss": 0.3808, "mean_token_accuracy": 0.8926702737808228, "num_tokens": 770610248.0, "step": 20294 }, { "epoch": 2.5817326039944026, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8345024585723877, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8633120656013489, "num_tokens": 770650230.0, "step": 20295 }, { "epoch": 2.581859814272993, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6110644340515137, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.8882628679275513, "num_tokens": 770694740.0, "step": 20296 }, { "epoch": 2.5819870245515837, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8380184173583984, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.886654257774353, "num_tokens": 770734004.0, "step": 20297 }, { "epoch": 2.582114234830174, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.628350257873535, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8769607543945312, "num_tokens": 770770421.0, "step": 20298 }, { "epoch": 2.5822414451087647, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.839951992034912, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8733499050140381, "num_tokens": 770810790.0, "step": 20299 }, { "epoch": 2.5823686553873553, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.63677716255188, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8751314282417297, "num_tokens": 770848614.0, "step": 20300 }, { "epoch": 2.582495865665946, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.820466995239258, "learning_rate": 1e-05, "loss": 0.3683, "mean_token_accuracy": 0.8993092775344849, "num_tokens": 770889888.0, "step": 20301 }, { "epoch": 2.5826230759445363, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6642279624938965, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8561028242111206, "num_tokens": 770925918.0, "step": 20302 }, { "epoch": 2.582750286223127, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.791692018508911, "learning_rate": 1e-05, "loss": 0.3597, "mean_token_accuracy": 0.8979569673538208, "num_tokens": 770963130.0, "step": 20303 }, { "epoch": 2.5828774965017174, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.671851634979248, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8695021867752075, "num_tokens": 770998308.0, "step": 20304 }, { "epoch": 2.583004706780308, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7935688495635986, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8724759817123413, "num_tokens": 771032532.0, "step": 20305 }, { "epoch": 2.5831319170588984, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.654921054840088, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8871196508407593, "num_tokens": 771070859.0, "step": 20306 }, { "epoch": 2.583259127337489, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.778757095336914, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8732154369354248, "num_tokens": 771111216.0, "step": 20307 }, { "epoch": 2.5833863376160795, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.66487979888916, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8805528283119202, "num_tokens": 771146703.0, "step": 20308 }, { "epoch": 2.58351354789467, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.788074254989624, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.883682370185852, "num_tokens": 771178778.0, "step": 20309 }, { "epoch": 2.5836407581732606, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6523005962371826, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8694374561309814, "num_tokens": 771217272.0, "step": 20310 }, { "epoch": 2.583767968451851, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.777235746383667, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8737093806266785, "num_tokens": 771252085.0, "step": 20311 }, { "epoch": 2.583895178730441, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6573617458343506, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8835323452949524, "num_tokens": 771285367.0, "step": 20312 }, { "epoch": 2.584022389009032, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.764418363571167, "learning_rate": 1e-05, "loss": 0.3972, "mean_token_accuracy": 0.8849502205848694, "num_tokens": 771324388.0, "step": 20313 }, { "epoch": 2.5841495992876222, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.642224073410034, "learning_rate": 1e-05, "loss": 0.3937, "mean_token_accuracy": 0.8872512578964233, "num_tokens": 771361307.0, "step": 20314 }, { "epoch": 2.584276809566213, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7785465717315674, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8684031367301941, "num_tokens": 771400383.0, "step": 20315 }, { "epoch": 2.5844040198448033, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6372287273406982, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.8900376558303833, "num_tokens": 771435065.0, "step": 20316 }, { "epoch": 2.584531230123394, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7703840732574463, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8639804124832153, "num_tokens": 771480573.0, "step": 20317 }, { "epoch": 2.5846584404019843, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.639427900314331, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.8863898515701294, "num_tokens": 771517172.0, "step": 20318 }, { "epoch": 2.584785650680575, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7874789237976074, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.86781245470047, "num_tokens": 771554347.0, "step": 20319 }, { "epoch": 2.5849128609591654, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6285629272460938, "learning_rate": 1e-05, "loss": 0.3911, "mean_token_accuracy": 0.8862660527229309, "num_tokens": 771591209.0, "step": 20320 }, { "epoch": 2.585040071237756, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.777068614959717, "learning_rate": 1e-05, "loss": 0.3924, "mean_token_accuracy": 0.8900108337402344, "num_tokens": 771626604.0, "step": 20321 }, { "epoch": 2.5851672815163464, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6434402465820312, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8770323991775513, "num_tokens": 771662525.0, "step": 20322 }, { "epoch": 2.585294491794937, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7702503204345703, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8901557922363281, "num_tokens": 771701616.0, "step": 20323 }, { "epoch": 2.5854217020735275, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6494884490966797, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8723582625389099, "num_tokens": 771736864.0, "step": 20324 }, { "epoch": 2.585548912352118, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.780980110168457, "learning_rate": 1e-05, "loss": 0.3936, "mean_token_accuracy": 0.8886467218399048, "num_tokens": 771774900.0, "step": 20325 }, { "epoch": 2.5856761226307086, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6546401977539062, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8754141330718994, "num_tokens": 771812078.0, "step": 20326 }, { "epoch": 2.585803332909299, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7590928077697754, "learning_rate": 1e-05, "loss": 0.4251, "mean_token_accuracy": 0.8793298006057739, "num_tokens": 771852247.0, "step": 20327 }, { "epoch": 2.5859305431878896, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6503701210021973, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8758505582809448, "num_tokens": 771891264.0, "step": 20328 }, { "epoch": 2.58605775346648, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7709133625030518, "learning_rate": 1e-05, "loss": 0.4587, "mean_token_accuracy": 0.8675729036331177, "num_tokens": 771935356.0, "step": 20329 }, { "epoch": 2.5861849637450707, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6675186157226562, "learning_rate": 1e-05, "loss": 0.4816, "mean_token_accuracy": 0.8576103448867798, "num_tokens": 771967284.0, "step": 20330 }, { "epoch": 2.586312174023661, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7401938438415527, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8854170441627502, "num_tokens": 772013465.0, "step": 20331 }, { "epoch": 2.5864393843022517, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.66084361076355, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8753785490989685, "num_tokens": 772046978.0, "step": 20332 }, { "epoch": 2.5865665945808423, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7369065284729004, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8876500129699707, "num_tokens": 772081630.0, "step": 20333 }, { "epoch": 2.586693804859433, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.665242910385132, "learning_rate": 1e-05, "loss": 0.4568, "mean_token_accuracy": 0.8696094751358032, "num_tokens": 772122197.0, "step": 20334 }, { "epoch": 2.5868210151380233, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7159905433654785, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8819027543067932, "num_tokens": 772162744.0, "step": 20335 }, { "epoch": 2.586948225416614, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.655467987060547, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8745911121368408, "num_tokens": 772207520.0, "step": 20336 }, { "epoch": 2.587075435695204, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.731189727783203, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8782654404640198, "num_tokens": 772244869.0, "step": 20337 }, { "epoch": 2.587202645973795, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6706812381744385, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8798172473907471, "num_tokens": 772279132.0, "step": 20338 }, { "epoch": 2.587329856252385, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.70581316947937, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.8906735777854919, "num_tokens": 772323668.0, "step": 20339 }, { "epoch": 2.587457066530976, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.66141414642334, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8787808418273926, "num_tokens": 772367834.0, "step": 20340 }, { "epoch": 2.587584276809566, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7324957847595215, "learning_rate": 1e-05, "loss": 0.3687, "mean_token_accuracy": 0.8963958024978638, "num_tokens": 772398777.0, "step": 20341 }, { "epoch": 2.5877114870881566, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.669128179550171, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8794443607330322, "num_tokens": 772432330.0, "step": 20342 }, { "epoch": 2.587838697366747, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7218375205993652, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8723255395889282, "num_tokens": 772472412.0, "step": 20343 }, { "epoch": 2.5879659076453376, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6596028804779053, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8793703317642212, "num_tokens": 772509896.0, "step": 20344 }, { "epoch": 2.588093117923928, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.719316005706787, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.8876713514328003, "num_tokens": 772549443.0, "step": 20345 }, { "epoch": 2.5882203282025187, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6588029861450195, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8684433698654175, "num_tokens": 772592086.0, "step": 20346 }, { "epoch": 2.588347538481109, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7223875522613525, "learning_rate": 1e-05, "loss": 0.4664, "mean_token_accuracy": 0.8667630553245544, "num_tokens": 772632805.0, "step": 20347 }, { "epoch": 2.5884747487596997, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.661874532699585, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8681007623672485, "num_tokens": 772673806.0, "step": 20348 }, { "epoch": 2.5886019590382903, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7194833755493164, "learning_rate": 1e-05, "loss": 0.3649, "mean_token_accuracy": 0.9024298191070557, "num_tokens": 772709434.0, "step": 20349 }, { "epoch": 2.588729169316881, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.661100149154663, "learning_rate": 1e-05, "loss": 0.41, "mean_token_accuracy": 0.8816508054733276, "num_tokens": 772748702.0, "step": 20350 }, { "epoch": 2.5888563795954713, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7286972999572754, "learning_rate": 1e-05, "loss": 0.4703, "mean_token_accuracy": 0.8620622754096985, "num_tokens": 772792235.0, "step": 20351 }, { "epoch": 2.588983589874062, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6629555225372314, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8718763589859009, "num_tokens": 772827253.0, "step": 20352 }, { "epoch": 2.5891108001526524, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.707573890686035, "learning_rate": 1e-05, "loss": 0.3852, "mean_token_accuracy": 0.8916484117507935, "num_tokens": 772871951.0, "step": 20353 }, { "epoch": 2.589238010431243, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.668119192123413, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8881452679634094, "num_tokens": 772908585.0, "step": 20354 }, { "epoch": 2.5893652207098334, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7242934703826904, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8736180067062378, "num_tokens": 772946883.0, "step": 20355 }, { "epoch": 2.589492430988424, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6563706398010254, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.8835702538490295, "num_tokens": 772985143.0, "step": 20356 }, { "epoch": 2.5896196412670145, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7140955924987793, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8844362497329712, "num_tokens": 773027004.0, "step": 20357 }, { "epoch": 2.589746851545605, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.658298969268799, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.8680537939071655, "num_tokens": 773067807.0, "step": 20358 }, { "epoch": 2.5898740618241956, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.719522714614868, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8843874931335449, "num_tokens": 773103160.0, "step": 20359 }, { "epoch": 2.5900012721027856, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.662937641143799, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8849910497665405, "num_tokens": 773143036.0, "step": 20360 }, { "epoch": 2.5901284823813766, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7171995639801025, "learning_rate": 1e-05, "loss": 0.4015, "mean_token_accuracy": 0.8848984241485596, "num_tokens": 773180930.0, "step": 20361 }, { "epoch": 2.5902556926599667, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6728477478027344, "learning_rate": 1e-05, "loss": 0.4795, "mean_token_accuracy": 0.8623228073120117, "num_tokens": 773220056.0, "step": 20362 }, { "epoch": 2.5903829029385577, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7152462005615234, "learning_rate": 1e-05, "loss": 0.3658, "mean_token_accuracy": 0.8978267908096313, "num_tokens": 773255890.0, "step": 20363 }, { "epoch": 2.5905101132171477, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.655332565307617, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8839081525802612, "num_tokens": 773298727.0, "step": 20364 }, { "epoch": 2.5906373234957387, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7307398319244385, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8610448837280273, "num_tokens": 773338137.0, "step": 20365 }, { "epoch": 2.590764533774329, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.670362710952759, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.886481761932373, "num_tokens": 773381532.0, "step": 20366 }, { "epoch": 2.5908917440529193, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7442398071289062, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.8686742782592773, "num_tokens": 773416354.0, "step": 20367 }, { "epoch": 2.59101895433151, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6569509506225586, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8794142603874207, "num_tokens": 773456130.0, "step": 20368 }, { "epoch": 2.5911461646101004, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.728107452392578, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.8860599398612976, "num_tokens": 773493812.0, "step": 20369 }, { "epoch": 2.591273374888691, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.657848834991455, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8819559812545776, "num_tokens": 773533000.0, "step": 20370 }, { "epoch": 2.5914005851672814, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7303948402404785, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8756303787231445, "num_tokens": 773572355.0, "step": 20371 }, { "epoch": 2.591527795445872, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6585090160369873, "learning_rate": 1e-05, "loss": 0.3872, "mean_token_accuracy": 0.8903838396072388, "num_tokens": 773609154.0, "step": 20372 }, { "epoch": 2.5916550057244625, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7185206413269043, "learning_rate": 1e-05, "loss": 0.3915, "mean_token_accuracy": 0.8877454996109009, "num_tokens": 773654430.0, "step": 20373 }, { "epoch": 2.591782216003053, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6538777351379395, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8844836950302124, "num_tokens": 773693710.0, "step": 20374 }, { "epoch": 2.5919094262816436, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7521698474884033, "learning_rate": 1e-05, "loss": 0.481, "mean_token_accuracy": 0.8544756770133972, "num_tokens": 773728858.0, "step": 20375 }, { "epoch": 2.592036636560234, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.666005849838257, "learning_rate": 1e-05, "loss": 0.441, "mean_token_accuracy": 0.8730275630950928, "num_tokens": 773761080.0, "step": 20376 }, { "epoch": 2.5921638468388246, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7284674644470215, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8921501636505127, "num_tokens": 773799084.0, "step": 20377 }, { "epoch": 2.592291057117415, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.64337158203125, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.8857706785202026, "num_tokens": 773840378.0, "step": 20378 }, { "epoch": 2.5924182673960057, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7417585849761963, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8832944631576538, "num_tokens": 773878627.0, "step": 20379 }, { "epoch": 2.592545477674596, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6385910511016846, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8710969090461731, "num_tokens": 773921492.0, "step": 20380 }, { "epoch": 2.5926726879531867, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7533905506134033, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.865071177482605, "num_tokens": 773959603.0, "step": 20381 }, { "epoch": 2.5927998982317773, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6437366008758545, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8692169189453125, "num_tokens": 773996550.0, "step": 20382 }, { "epoch": 2.592927108510368, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7852447032928467, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8805468678474426, "num_tokens": 774033815.0, "step": 20383 }, { "epoch": 2.5930543187889583, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6331100463867188, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8794845342636108, "num_tokens": 774068930.0, "step": 20384 }, { "epoch": 2.5931815290675484, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7584619522094727, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8774610757827759, "num_tokens": 774112014.0, "step": 20385 }, { "epoch": 2.5933087393461394, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6373841762542725, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8788257241249084, "num_tokens": 774149676.0, "step": 20386 }, { "epoch": 2.5934359496247295, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7514560222625732, "learning_rate": 1e-05, "loss": 0.3693, "mean_token_accuracy": 0.8984363675117493, "num_tokens": 774190160.0, "step": 20387 }, { "epoch": 2.5935631599033204, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.637181520462036, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8773748874664307, "num_tokens": 774235133.0, "step": 20388 }, { "epoch": 2.5936903701819105, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7479946613311768, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8856797814369202, "num_tokens": 774277962.0, "step": 20389 }, { "epoch": 2.5938175804605015, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6431405544281006, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8817058205604553, "num_tokens": 774316946.0, "step": 20390 }, { "epoch": 2.5939447907390916, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.750790596008301, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8844294548034668, "num_tokens": 774356376.0, "step": 20391 }, { "epoch": 2.594072001017682, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.639726400375366, "learning_rate": 1e-05, "loss": 0.3854, "mean_token_accuracy": 0.8905196189880371, "num_tokens": 774392543.0, "step": 20392 }, { "epoch": 2.5941992112962726, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7632296085357666, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.873375415802002, "num_tokens": 774432114.0, "step": 20393 }, { "epoch": 2.594326421574863, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.643596887588501, "learning_rate": 1e-05, "loss": 0.4423, "mean_token_accuracy": 0.8704550266265869, "num_tokens": 774468758.0, "step": 20394 }, { "epoch": 2.5944536318534537, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7836406230926514, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8675237894058228, "num_tokens": 774503136.0, "step": 20395 }, { "epoch": 2.594580842132044, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.621577262878418, "learning_rate": 1e-05, "loss": 0.3902, "mean_token_accuracy": 0.8879940509796143, "num_tokens": 774545162.0, "step": 20396 }, { "epoch": 2.5947080524106347, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7863752841949463, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8808290958404541, "num_tokens": 774581831.0, "step": 20397 }, { "epoch": 2.5948352626892253, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.62241268157959, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8859758973121643, "num_tokens": 774619258.0, "step": 20398 }, { "epoch": 2.594962472967816, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.817556142807007, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8659815192222595, "num_tokens": 774650244.0, "step": 20399 }, { "epoch": 2.5950896832464063, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6213316917419434, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8657697439193726, "num_tokens": 774690060.0, "step": 20400 }, { "epoch": 2.595216893524997, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8108482360839844, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8759933114051819, "num_tokens": 774733925.0, "step": 20401 }, { "epoch": 2.5953441038035874, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.622831106185913, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8842028975486755, "num_tokens": 774769344.0, "step": 20402 }, { "epoch": 2.595471314082178, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8010129928588867, "learning_rate": 1e-05, "loss": 0.3765, "mean_token_accuracy": 0.8958482146263123, "num_tokens": 774809548.0, "step": 20403 }, { "epoch": 2.5955985243607684, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6269140243530273, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8768680095672607, "num_tokens": 774848042.0, "step": 20404 }, { "epoch": 2.595725734639359, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.821744918823242, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8721984624862671, "num_tokens": 774884226.0, "step": 20405 }, { "epoch": 2.5958529449179495, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.65368914604187, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8769513368606567, "num_tokens": 774913697.0, "step": 20406 }, { "epoch": 2.59598015519654, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8045289516448975, "learning_rate": 1e-05, "loss": 0.3774, "mean_token_accuracy": 0.8953136205673218, "num_tokens": 774948369.0, "step": 20407 }, { "epoch": 2.5961073654751305, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6276443004608154, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8824594616889954, "num_tokens": 774992260.0, "step": 20408 }, { "epoch": 2.596234575753721, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.845726490020752, "learning_rate": 1e-05, "loss": 0.536, "mean_token_accuracy": 0.8398417234420776, "num_tokens": 775023613.0, "step": 20409 }, { "epoch": 2.596361786032311, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.631549119949341, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8754067420959473, "num_tokens": 775063874.0, "step": 20410 }, { "epoch": 2.596488996310902, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.82285213470459, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8745797872543335, "num_tokens": 775104447.0, "step": 20411 }, { "epoch": 2.596616206589492, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.615478515625, "learning_rate": 1e-05, "loss": 0.3988, "mean_token_accuracy": 0.8873812556266785, "num_tokens": 775143472.0, "step": 20412 }, { "epoch": 2.596743416868083, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.845247745513916, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8847665786743164, "num_tokens": 775176274.0, "step": 20413 }, { "epoch": 2.5968706271466733, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.620438814163208, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8721790313720703, "num_tokens": 775216485.0, "step": 20414 }, { "epoch": 2.596997837425264, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.830378770828247, "learning_rate": 1e-05, "loss": 0.393, "mean_token_accuracy": 0.889552116394043, "num_tokens": 775256828.0, "step": 20415 }, { "epoch": 2.5971250477038543, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.637723684310913, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.871126651763916, "num_tokens": 775296577.0, "step": 20416 }, { "epoch": 2.597252257982445, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8419559001922607, "learning_rate": 1e-05, "loss": 0.4327, "mean_token_accuracy": 0.8789068460464478, "num_tokens": 775329567.0, "step": 20417 }, { "epoch": 2.5973794682610354, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.641005516052246, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8834316730499268, "num_tokens": 775363729.0, "step": 20418 }, { "epoch": 2.597506678539626, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.8437960147857666, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8789891600608826, "num_tokens": 775397547.0, "step": 20419 }, { "epoch": 2.5976338888182164, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.625906229019165, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8701253533363342, "num_tokens": 775441033.0, "step": 20420 }, { "epoch": 2.597761099096807, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8435745239257812, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8857867121696472, "num_tokens": 775477516.0, "step": 20421 }, { "epoch": 2.5978883093753975, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.629417896270752, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8852113485336304, "num_tokens": 775515573.0, "step": 20422 }, { "epoch": 2.598015519653988, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.833465337753296, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8840967416763306, "num_tokens": 775552970.0, "step": 20423 }, { "epoch": 2.5981427299325786, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6392221450805664, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8761148452758789, "num_tokens": 775590180.0, "step": 20424 }, { "epoch": 2.598269940211169, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.829366445541382, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8845928907394409, "num_tokens": 775627346.0, "step": 20425 }, { "epoch": 2.5983971504897596, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.637850522994995, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8780955076217651, "num_tokens": 775667173.0, "step": 20426 }, { "epoch": 2.59852436076835, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.833439826965332, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8683198690414429, "num_tokens": 775702710.0, "step": 20427 }, { "epoch": 2.5986515710469407, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6295151710510254, "learning_rate": 1e-05, "loss": 0.4249, "mean_token_accuracy": 0.8794758319854736, "num_tokens": 775743934.0, "step": 20428 }, { "epoch": 2.598778781325531, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8257133960723877, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8837592005729675, "num_tokens": 775783184.0, "step": 20429 }, { "epoch": 2.5989059916041217, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6392812728881836, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8721208572387695, "num_tokens": 775820019.0, "step": 20430 }, { "epoch": 2.5990332018827123, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.822660207748413, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8788787126541138, "num_tokens": 775858740.0, "step": 20431 }, { "epoch": 2.599160412161303, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.635946035385132, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8781957626342773, "num_tokens": 775900615.0, "step": 20432 }, { "epoch": 2.5992876224398933, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.811342716217041, "learning_rate": 1e-05, "loss": 0.3913, "mean_token_accuracy": 0.890172004699707, "num_tokens": 775938953.0, "step": 20433 }, { "epoch": 2.599414832718484, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.652442216873169, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8749823570251465, "num_tokens": 775973203.0, "step": 20434 }, { "epoch": 2.599542042997074, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7976977825164795, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8683573603630066, "num_tokens": 776014984.0, "step": 20435 }, { "epoch": 2.599669253275665, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6496734619140625, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8642653226852417, "num_tokens": 776059883.0, "step": 20436 }, { "epoch": 2.599796463554255, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7974605560302734, "learning_rate": 1e-05, "loss": 0.4622, "mean_token_accuracy": 0.8671102523803711, "num_tokens": 776101025.0, "step": 20437 }, { "epoch": 2.599923673832846, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6489408016204834, "learning_rate": 1e-05, "loss": 0.3958, "mean_token_accuracy": 0.8835139274597168, "num_tokens": 776134356.0, "step": 20438 }, { "epoch": 2.600050884111436, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7718145847320557, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8845006227493286, "num_tokens": 776176158.0, "step": 20439 }, { "epoch": 2.6001780943900266, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.656522512435913, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8800403475761414, "num_tokens": 776209586.0, "step": 20440 }, { "epoch": 2.600305304668617, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7918777465820312, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8712852001190186, "num_tokens": 776240707.0, "step": 20441 }, { "epoch": 2.6004325149472076, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.644650936126709, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.877829909324646, "num_tokens": 776280448.0, "step": 20442 }, { "epoch": 2.600559725225798, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7626659870147705, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.891606867313385, "num_tokens": 776320112.0, "step": 20443 }, { "epoch": 2.6006869355043887, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.646008014678955, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.876704216003418, "num_tokens": 776362720.0, "step": 20444 }, { "epoch": 2.600814145782979, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.768893003463745, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8768487572669983, "num_tokens": 776403006.0, "step": 20445 }, { "epoch": 2.6009413560615697, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6520721912384033, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8763687610626221, "num_tokens": 776436039.0, "step": 20446 }, { "epoch": 2.6010685663401603, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.760303497314453, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8786382079124451, "num_tokens": 776476337.0, "step": 20447 }, { "epoch": 2.601195776618751, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.64900803565979, "learning_rate": 1e-05, "loss": 0.3844, "mean_token_accuracy": 0.889794111251831, "num_tokens": 776512552.0, "step": 20448 }, { "epoch": 2.6013229868973413, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7595794200897217, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8758143782615662, "num_tokens": 776549727.0, "step": 20449 }, { "epoch": 2.601450197175932, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.683090925216675, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8745722770690918, "num_tokens": 776586701.0, "step": 20450 }, { "epoch": 2.6015774074545224, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7934463024139404, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8755653500556946, "num_tokens": 776627386.0, "step": 20451 }, { "epoch": 2.601704617733113, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.637694835662842, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8840450048446655, "num_tokens": 776668389.0, "step": 20452 }, { "epoch": 2.6018318280117034, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7619500160217285, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8753315210342407, "num_tokens": 776708106.0, "step": 20453 }, { "epoch": 2.601959038290294, "ewc_loss": 0.0771484375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.637167453765869, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8798523545265198, "num_tokens": 776744577.0, "step": 20454 }, { "epoch": 2.6020862485688845, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7650296688079834, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8667148351669312, "num_tokens": 776784350.0, "step": 20455 }, { "epoch": 2.602213458847475, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6366679668426514, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8864808082580566, "num_tokens": 776823728.0, "step": 20456 }, { "epoch": 2.6023406691260655, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.79611873626709, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8797862529754639, "num_tokens": 776860097.0, "step": 20457 }, { "epoch": 2.6024678794046556, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6416594982147217, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8827279806137085, "num_tokens": 776895876.0, "step": 20458 }, { "epoch": 2.6025950896832466, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8274338245391846, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8735978603363037, "num_tokens": 776930003.0, "step": 20459 }, { "epoch": 2.6027222999618367, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.618112564086914, "learning_rate": 1e-05, "loss": 0.3596, "mean_token_accuracy": 0.8979251384735107, "num_tokens": 776969699.0, "step": 20460 }, { "epoch": 2.6028495102404277, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8162851333618164, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8819735646247864, "num_tokens": 777006517.0, "step": 20461 }, { "epoch": 2.6029767205190177, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6281814575195312, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8611024618148804, "num_tokens": 777044160.0, "step": 20462 }, { "epoch": 2.6031039307976087, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8235902786254883, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8816797733306885, "num_tokens": 777079890.0, "step": 20463 }, { "epoch": 2.603231141076199, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6101062297821045, "learning_rate": 1e-05, "loss": 0.3827, "mean_token_accuracy": 0.8912228345870972, "num_tokens": 777120928.0, "step": 20464 }, { "epoch": 2.6033583513547893, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8276150226593018, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8669414520263672, "num_tokens": 777166363.0, "step": 20465 }, { "epoch": 2.60348556163338, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6398468017578125, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8795071244239807, "num_tokens": 777201146.0, "step": 20466 }, { "epoch": 2.6036127719119704, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8228096961975098, "learning_rate": 1e-05, "loss": 0.4901, "mean_token_accuracy": 0.8602824211120605, "num_tokens": 777239884.0, "step": 20467 }, { "epoch": 2.603739982190561, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6346075534820557, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8654907941818237, "num_tokens": 777279898.0, "step": 20468 }, { "epoch": 2.6038671924691514, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8172688484191895, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8744152784347534, "num_tokens": 777320614.0, "step": 20469 }, { "epoch": 2.603994402747742, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6240475177764893, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8755662441253662, "num_tokens": 777364782.0, "step": 20470 }, { "epoch": 2.6041216130263325, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.84317684173584, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8719735145568848, "num_tokens": 777398276.0, "step": 20471 }, { "epoch": 2.604248823304923, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.630936622619629, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.885270893573761, "num_tokens": 777434510.0, "step": 20472 }, { "epoch": 2.6043760335835135, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8296210765838623, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.8803524971008301, "num_tokens": 777469266.0, "step": 20473 }, { "epoch": 2.604503243862104, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6411125659942627, "learning_rate": 1e-05, "loss": 0.3738, "mean_token_accuracy": 0.892132043838501, "num_tokens": 777500770.0, "step": 20474 }, { "epoch": 2.6046304541406946, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8188207149505615, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.879744291305542, "num_tokens": 777537977.0, "step": 20475 }, { "epoch": 2.604757664419285, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.637322425842285, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8814905881881714, "num_tokens": 777574065.0, "step": 20476 }, { "epoch": 2.6048848746978757, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8107569217681885, "learning_rate": 1e-05, "loss": 0.4017, "mean_token_accuracy": 0.887742280960083, "num_tokens": 777613366.0, "step": 20477 }, { "epoch": 2.605012084976466, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6517770290374756, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.874592125415802, "num_tokens": 777650954.0, "step": 20478 }, { "epoch": 2.6051392952550567, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7995445728302, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8665347099304199, "num_tokens": 777694370.0, "step": 20479 }, { "epoch": 2.6052665055336472, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6557960510253906, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8768436908721924, "num_tokens": 777734039.0, "step": 20480 }, { "epoch": 2.6053937158122378, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7790799140930176, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8788639307022095, "num_tokens": 777776237.0, "step": 20481 }, { "epoch": 2.6055209260908283, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6499969959259033, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8835135698318481, "num_tokens": 777813767.0, "step": 20482 }, { "epoch": 2.6056481363694184, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.787370204925537, "learning_rate": 1e-05, "loss": 0.4037, "mean_token_accuracy": 0.8822245597839355, "num_tokens": 777848642.0, "step": 20483 }, { "epoch": 2.6057753466480094, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.643347978591919, "learning_rate": 1e-05, "loss": 0.3975, "mean_token_accuracy": 0.8849043846130371, "num_tokens": 777885748.0, "step": 20484 }, { "epoch": 2.6059025569265994, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.785080909729004, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.86723792552948, "num_tokens": 777922366.0, "step": 20485 }, { "epoch": 2.6060297672051904, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.642594337463379, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8781131505966187, "num_tokens": 777965130.0, "step": 20486 }, { "epoch": 2.6061569774837805, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.783949851989746, "learning_rate": 1e-05, "loss": 0.3776, "mean_token_accuracy": 0.8926112651824951, "num_tokens": 777999425.0, "step": 20487 }, { "epoch": 2.6062841877623715, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6399338245391846, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8680518865585327, "num_tokens": 778042378.0, "step": 20488 }, { "epoch": 2.6064113980409616, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7816433906555176, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8802338242530823, "num_tokens": 778083858.0, "step": 20489 }, { "epoch": 2.606538608319552, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.643613338470459, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8860996961593628, "num_tokens": 778122103.0, "step": 20490 }, { "epoch": 2.6066658185981426, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7860405445098877, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.879355788230896, "num_tokens": 778158207.0, "step": 20491 }, { "epoch": 2.606793028876733, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6448984146118164, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8851779103279114, "num_tokens": 778191542.0, "step": 20492 }, { "epoch": 2.6069202391553237, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7710671424865723, "learning_rate": 1e-05, "loss": 0.4264, "mean_token_accuracy": 0.8811687231063843, "num_tokens": 778234929.0, "step": 20493 }, { "epoch": 2.607047449433914, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.625154495239258, "learning_rate": 1e-05, "loss": 0.3616, "mean_token_accuracy": 0.897560715675354, "num_tokens": 778273225.0, "step": 20494 }, { "epoch": 2.6071746597125047, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7893617153167725, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8752444982528687, "num_tokens": 778310998.0, "step": 20495 }, { "epoch": 2.6073018699910953, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6458728313446045, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8728711605072021, "num_tokens": 778348379.0, "step": 20496 }, { "epoch": 2.607429080269686, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7819664478302, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8753504753112793, "num_tokens": 778390742.0, "step": 20497 }, { "epoch": 2.6075562905482763, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6412785053253174, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8842636346817017, "num_tokens": 778429239.0, "step": 20498 }, { "epoch": 2.607683500826867, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7771289348602295, "learning_rate": 1e-05, "loss": 0.3899, "mean_token_accuracy": 0.8888489007949829, "num_tokens": 778467606.0, "step": 20499 }, { "epoch": 2.6078107111054574, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6575467586517334, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8701871633529663, "num_tokens": 778500790.0, "step": 20500 }, { "epoch": 2.607937921384048, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7739768028259277, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8819224834442139, "num_tokens": 778536412.0, "step": 20501 }, { "epoch": 2.6080651316626384, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.639136552810669, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8737050294876099, "num_tokens": 778579812.0, "step": 20502 }, { "epoch": 2.608192341941229, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7529115676879883, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8868874311447144, "num_tokens": 778626135.0, "step": 20503 }, { "epoch": 2.6083195522198195, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.644697427749634, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8853906989097595, "num_tokens": 778666163.0, "step": 20504 }, { "epoch": 2.60844676249841, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7702043056488037, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.881829023361206, "num_tokens": 778703168.0, "step": 20505 }, { "epoch": 2.6085739727770005, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.650228500366211, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.8704802989959717, "num_tokens": 778741696.0, "step": 20506 }, { "epoch": 2.608701183055591, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7669241428375244, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8793744444847107, "num_tokens": 778777450.0, "step": 20507 }, { "epoch": 2.608828393334181, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.654313325881958, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8814396262168884, "num_tokens": 778811448.0, "step": 20508 }, { "epoch": 2.608955603612772, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.755811929702759, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8784075379371643, "num_tokens": 778849333.0, "step": 20509 }, { "epoch": 2.609082813891362, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6529338359832764, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8853790163993835, "num_tokens": 778885264.0, "step": 20510 }, { "epoch": 2.609210024169953, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7606539726257324, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8796014785766602, "num_tokens": 778918160.0, "step": 20511 }, { "epoch": 2.6093372344485433, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.637251377105713, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8838940262794495, "num_tokens": 778955938.0, "step": 20512 }, { "epoch": 2.609464444727134, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7592029571533203, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8753194212913513, "num_tokens": 778994412.0, "step": 20513 }, { "epoch": 2.6095916550057243, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.655106782913208, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8684375286102295, "num_tokens": 779030818.0, "step": 20514 }, { "epoch": 2.609718865284315, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.75071382522583, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.8882100582122803, "num_tokens": 779070217.0, "step": 20515 }, { "epoch": 2.6098460755629054, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6447596549987793, "learning_rate": 1e-05, "loss": 0.3815, "mean_token_accuracy": 0.8924908638000488, "num_tokens": 779105668.0, "step": 20516 }, { "epoch": 2.609973285841496, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.750880479812622, "learning_rate": 1e-05, "loss": 0.3662, "mean_token_accuracy": 0.9000186324119568, "num_tokens": 779140005.0, "step": 20517 }, { "epoch": 2.6101004961200864, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6508896350860596, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8819775581359863, "num_tokens": 779176269.0, "step": 20518 }, { "epoch": 2.610227706398677, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7576563358306885, "learning_rate": 1e-05, "loss": 0.3902, "mean_token_accuracy": 0.8914979100227356, "num_tokens": 779209914.0, "step": 20519 }, { "epoch": 2.6103549166772675, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.651376247406006, "learning_rate": 1e-05, "loss": 0.3943, "mean_token_accuracy": 0.8895220756530762, "num_tokens": 779244994.0, "step": 20520 }, { "epoch": 2.610482126955858, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.752248764038086, "learning_rate": 1e-05, "loss": 0.3988, "mean_token_accuracy": 0.8863917589187622, "num_tokens": 779283700.0, "step": 20521 }, { "epoch": 2.6106093372344485, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.637974977493286, "learning_rate": 1e-05, "loss": 0.3964, "mean_token_accuracy": 0.8900542259216309, "num_tokens": 779324827.0, "step": 20522 }, { "epoch": 2.610736547513039, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7516260147094727, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8703914284706116, "num_tokens": 779363593.0, "step": 20523 }, { "epoch": 2.6108637577916296, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6466732025146484, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.884393572807312, "num_tokens": 779401295.0, "step": 20524 }, { "epoch": 2.61099096807022, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.763978958129883, "learning_rate": 1e-05, "loss": 0.4682, "mean_token_accuracy": 0.8642419576644897, "num_tokens": 779440375.0, "step": 20525 }, { "epoch": 2.6111181783488107, "ewc_loss": 0.078125, "ewc_loss_diag": 0.000354766845703125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6460177898406982, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8763646483421326, "num_tokens": 779475881.0, "step": 20526 }, { "epoch": 2.611245388627401, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7510011196136475, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8699997067451477, "num_tokens": 779520155.0, "step": 20527 }, { "epoch": 2.6113725989059917, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.63440203666687, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8774000406265259, "num_tokens": 779563952.0, "step": 20528 }, { "epoch": 2.6114998091845822, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7594969272613525, "learning_rate": 1e-05, "loss": 0.3968, "mean_token_accuracy": 0.8901343941688538, "num_tokens": 779601618.0, "step": 20529 }, { "epoch": 2.6116270194631728, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.632354259490967, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8735545873641968, "num_tokens": 779645228.0, "step": 20530 }, { "epoch": 2.6117542297417633, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.779984712600708, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.878193736076355, "num_tokens": 779681780.0, "step": 20531 }, { "epoch": 2.611881440020354, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6411993503570557, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8806294798851013, "num_tokens": 779717783.0, "step": 20532 }, { "epoch": 2.612008650298944, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7820441722869873, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8865338563919067, "num_tokens": 779752542.0, "step": 20533 }, { "epoch": 2.612135860577535, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.621997594833374, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8787031769752502, "num_tokens": 779795134.0, "step": 20534 }, { "epoch": 2.612263070856125, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.79665207862854, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.8857078552246094, "num_tokens": 779828589.0, "step": 20535 }, { "epoch": 2.612390281134716, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6441385746002197, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8670920133590698, "num_tokens": 779865904.0, "step": 20536 }, { "epoch": 2.612517491413306, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.773094654083252, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8809664249420166, "num_tokens": 779910204.0, "step": 20537 }, { "epoch": 2.6126447016918966, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6323249340057373, "learning_rate": 1e-05, "loss": 0.3948, "mean_token_accuracy": 0.8875030875205994, "num_tokens": 779951032.0, "step": 20538 }, { "epoch": 2.612771911970487, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7837352752685547, "learning_rate": 1e-05, "loss": 0.3906, "mean_token_accuracy": 0.8909187316894531, "num_tokens": 779983316.0, "step": 20539 }, { "epoch": 2.6128991222490776, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6324055194854736, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8812215328216553, "num_tokens": 780021167.0, "step": 20540 }, { "epoch": 2.613026332527668, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7851033210754395, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.878800094127655, "num_tokens": 780061276.0, "step": 20541 }, { "epoch": 2.6131535428062587, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6421122550964355, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8844008445739746, "num_tokens": 780097204.0, "step": 20542 }, { "epoch": 2.613280753084849, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.781160831451416, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8713559508323669, "num_tokens": 780137208.0, "step": 20543 }, { "epoch": 2.6134079633634397, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6306920051574707, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8728259801864624, "num_tokens": 780177815.0, "step": 20544 }, { "epoch": 2.6135351736420303, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7805402278900146, "learning_rate": 1e-05, "loss": 0.377, "mean_token_accuracy": 0.8944970965385437, "num_tokens": 780215798.0, "step": 20545 }, { "epoch": 2.613662383920621, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.647388458251953, "learning_rate": 1e-05, "loss": 0.4564, "mean_token_accuracy": 0.8642081022262573, "num_tokens": 780252252.0, "step": 20546 }, { "epoch": 2.6137895941992113, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.775580644607544, "learning_rate": 1e-05, "loss": 0.3753, "mean_token_accuracy": 0.8955155611038208, "num_tokens": 780288100.0, "step": 20547 }, { "epoch": 2.613916804477802, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.637274742126465, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8791565895080566, "num_tokens": 780331922.0, "step": 20548 }, { "epoch": 2.6140440147563924, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7730584144592285, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8723885416984558, "num_tokens": 780369930.0, "step": 20549 }, { "epoch": 2.614171225034983, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6469907760620117, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8806508779525757, "num_tokens": 780403893.0, "step": 20550 }, { "epoch": 2.6142984353135734, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7696897983551025, "learning_rate": 1e-05, "loss": 0.3864, "mean_token_accuracy": 0.8910191059112549, "num_tokens": 780438426.0, "step": 20551 }, { "epoch": 2.614425645592164, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.637010097503662, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8832331895828247, "num_tokens": 780483267.0, "step": 20552 }, { "epoch": 2.6145528558707545, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.821978807449341, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8829454183578491, "num_tokens": 780520355.0, "step": 20553 }, { "epoch": 2.614680066149345, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.615724802017212, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8796377778053284, "num_tokens": 780563449.0, "step": 20554 }, { "epoch": 2.6148072764279355, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8124587535858154, "learning_rate": 1e-05, "loss": 0.3826, "mean_token_accuracy": 0.8919605016708374, "num_tokens": 780598607.0, "step": 20555 }, { "epoch": 2.6149344867065256, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.613675594329834, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8814629316329956, "num_tokens": 780638425.0, "step": 20556 }, { "epoch": 2.6150616969851166, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8241217136383057, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8694052696228027, "num_tokens": 780683441.0, "step": 20557 }, { "epoch": 2.6151889072637067, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6210033893585205, "learning_rate": 1e-05, "loss": 0.376, "mean_token_accuracy": 0.894437313079834, "num_tokens": 780720281.0, "step": 20558 }, { "epoch": 2.6153161175422976, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8251941204071045, "learning_rate": 1e-05, "loss": 0.3926, "mean_token_accuracy": 0.8904643654823303, "num_tokens": 780761038.0, "step": 20559 }, { "epoch": 2.6154433278208877, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6303746700286865, "learning_rate": 1e-05, "loss": 0.4881, "mean_token_accuracy": 0.8561381101608276, "num_tokens": 780802358.0, "step": 20560 }, { "epoch": 2.6155705380994787, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.833632230758667, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8836140632629395, "num_tokens": 780839820.0, "step": 20561 }, { "epoch": 2.615697748378069, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.624083995819092, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.879034161567688, "num_tokens": 780879969.0, "step": 20562 }, { "epoch": 2.6158249586566593, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.8399245738983154, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8804852962493896, "num_tokens": 780916119.0, "step": 20563 }, { "epoch": 2.61595216893525, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.630905866622925, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8787593245506287, "num_tokens": 780957311.0, "step": 20564 }, { "epoch": 2.6160793792138404, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8291585445404053, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8738433718681335, "num_tokens": 780998322.0, "step": 20565 }, { "epoch": 2.616206589492431, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6428990364074707, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8664394021034241, "num_tokens": 781036249.0, "step": 20566 }, { "epoch": 2.6163337997710214, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8208963871002197, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.881806492805481, "num_tokens": 781075228.0, "step": 20567 }, { "epoch": 2.616461010049612, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.642174005508423, "learning_rate": 1e-05, "loss": 0.3877, "mean_token_accuracy": 0.8869348764419556, "num_tokens": 781110907.0, "step": 20568 }, { "epoch": 2.6165882203282025, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.821061134338379, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8707172274589539, "num_tokens": 781148615.0, "step": 20569 }, { "epoch": 2.616715430606793, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6588971614837646, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8860335946083069, "num_tokens": 781181159.0, "step": 20570 }, { "epoch": 2.6168426408853835, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8124072551727295, "learning_rate": 1e-05, "loss": 0.4956, "mean_token_accuracy": 0.857736349105835, "num_tokens": 781221563.0, "step": 20571 }, { "epoch": 2.616969851163974, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.651679277420044, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8672854900360107, "num_tokens": 781257606.0, "step": 20572 }, { "epoch": 2.6170970614425646, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7979013919830322, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8767866492271423, "num_tokens": 781296653.0, "step": 20573 }, { "epoch": 2.617224271721155, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6452746391296387, "learning_rate": 1e-05, "loss": 0.372, "mean_token_accuracy": 0.8943414688110352, "num_tokens": 781331596.0, "step": 20574 }, { "epoch": 2.6173514819997457, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.785555362701416, "learning_rate": 1e-05, "loss": 0.3559, "mean_token_accuracy": 0.9027878046035767, "num_tokens": 781365599.0, "step": 20575 }, { "epoch": 2.617478692278336, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.647501230239868, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8774677515029907, "num_tokens": 781406169.0, "step": 20576 }, { "epoch": 2.6176059025569267, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7856829166412354, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8701614141464233, "num_tokens": 781444570.0, "step": 20577 }, { "epoch": 2.6177331128355172, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.649566411972046, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.868830144405365, "num_tokens": 781482752.0, "step": 20578 }, { "epoch": 2.6178603231141078, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.780271530151367, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8682760000228882, "num_tokens": 781524549.0, "step": 20579 }, { "epoch": 2.6179875333926983, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.666452169418335, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.8670884370803833, "num_tokens": 781560152.0, "step": 20580 }, { "epoch": 2.6181147436712884, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7475948333740234, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8785951137542725, "num_tokens": 781606399.0, "step": 20581 }, { "epoch": 2.6182419539498794, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.655548334121704, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8808863162994385, "num_tokens": 781646386.0, "step": 20582 }, { "epoch": 2.6183691642284694, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.762769937515259, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8736875653266907, "num_tokens": 781685629.0, "step": 20583 }, { "epoch": 2.6184963745070604, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6504387855529785, "learning_rate": 1e-05, "loss": 0.3603, "mean_token_accuracy": 0.8957419395446777, "num_tokens": 781717198.0, "step": 20584 }, { "epoch": 2.6186235847856505, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.793246269226074, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8670046925544739, "num_tokens": 781749281.0, "step": 20585 }, { "epoch": 2.6187507950642415, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.634603261947632, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8829220533370972, "num_tokens": 781791544.0, "step": 20586 }, { "epoch": 2.6188780053428315, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7760567665100098, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8729742765426636, "num_tokens": 781832864.0, "step": 20587 }, { "epoch": 2.619005215621422, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6374905109405518, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8804917335510254, "num_tokens": 781875757.0, "step": 20588 }, { "epoch": 2.6191324259000126, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.783823251724243, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8827997446060181, "num_tokens": 781911200.0, "step": 20589 }, { "epoch": 2.619259636178603, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6312429904937744, "learning_rate": 1e-05, "loss": 0.3855, "mean_token_accuracy": 0.8922327756881714, "num_tokens": 781948372.0, "step": 20590 }, { "epoch": 2.6193868464571937, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.799095392227173, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8777759671211243, "num_tokens": 781985360.0, "step": 20591 }, { "epoch": 2.619514056735784, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6316442489624023, "learning_rate": 1e-05, "loss": 0.3911, "mean_token_accuracy": 0.8842273950576782, "num_tokens": 782019439.0, "step": 20592 }, { "epoch": 2.6196412670143747, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.777512788772583, "learning_rate": 1e-05, "loss": 0.343, "mean_token_accuracy": 0.9052380323410034, "num_tokens": 782055805.0, "step": 20593 }, { "epoch": 2.6197684772929652, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6430866718292236, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8759194612503052, "num_tokens": 782092879.0, "step": 20594 }, { "epoch": 2.6198956875715558, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7806396484375, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8768903613090515, "num_tokens": 782133123.0, "step": 20595 }, { "epoch": 2.6200228978501463, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6490204334259033, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.8659992218017578, "num_tokens": 782169999.0, "step": 20596 }, { "epoch": 2.620150108128737, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.774498224258423, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8835951089859009, "num_tokens": 782208269.0, "step": 20597 }, { "epoch": 2.6202773184073274, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.652400255203247, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8773871660232544, "num_tokens": 782242927.0, "step": 20598 }, { "epoch": 2.620404528685918, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7748892307281494, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8753031492233276, "num_tokens": 782281018.0, "step": 20599 }, { "epoch": 2.6205317389645084, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.687021017074585, "learning_rate": 1e-05, "loss": 0.512, "mean_token_accuracy": 0.8543362617492676, "num_tokens": 782315336.0, "step": 20600 }, { "epoch": 2.620658949243099, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7450084686279297, "learning_rate": 1e-05, "loss": 0.3895, "mean_token_accuracy": 0.8932258486747742, "num_tokens": 782351163.0, "step": 20601 }, { "epoch": 2.6207861595216895, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6736221313476562, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8624600172042847, "num_tokens": 782396955.0, "step": 20602 }, { "epoch": 2.62091336980028, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7387232780456543, "learning_rate": 1e-05, "loss": 0.3851, "mean_token_accuracy": 0.8937753438949585, "num_tokens": 782432005.0, "step": 20603 }, { "epoch": 2.6210405800788705, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.677995443344116, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8837408423423767, "num_tokens": 782467659.0, "step": 20604 }, { "epoch": 2.621167790357461, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7440738677978516, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8784801363945007, "num_tokens": 782502621.0, "step": 20605 }, { "epoch": 2.621295000636051, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.672880172729492, "learning_rate": 1e-05, "loss": 0.4008, "mean_token_accuracy": 0.8876954317092896, "num_tokens": 782535803.0, "step": 20606 }, { "epoch": 2.621422210914642, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.720531463623047, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8785949349403381, "num_tokens": 782574720.0, "step": 20607 }, { "epoch": 2.621549421193232, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6562368869781494, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8842872381210327, "num_tokens": 782615477.0, "step": 20608 }, { "epoch": 2.621676631471823, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7328238487243652, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8714604377746582, "num_tokens": 782653153.0, "step": 20609 }, { "epoch": 2.6218038417504133, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7033097743988037, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8695343732833862, "num_tokens": 782691024.0, "step": 20610 }, { "epoch": 2.621931052029004, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7052180767059326, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8809401392936707, "num_tokens": 782732031.0, "step": 20611 }, { "epoch": 2.6220582623075943, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6778154373168945, "learning_rate": 1e-05, "loss": 0.4301, "mean_token_accuracy": 0.8771687746047974, "num_tokens": 782773430.0, "step": 20612 }, { "epoch": 2.622185472586185, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.703122138977051, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8737744092941284, "num_tokens": 782815852.0, "step": 20613 }, { "epoch": 2.6223126828647754, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.681145429611206, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.887249231338501, "num_tokens": 782851327.0, "step": 20614 }, { "epoch": 2.622439893143366, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7195823192596436, "learning_rate": 1e-05, "loss": 0.451, "mean_token_accuracy": 0.8694143295288086, "num_tokens": 782887382.0, "step": 20615 }, { "epoch": 2.6225671034219564, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6854147911071777, "learning_rate": 1e-05, "loss": 0.389, "mean_token_accuracy": 0.890110969543457, "num_tokens": 782923667.0, "step": 20616 }, { "epoch": 2.622694313700547, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.711916446685791, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.8860505819320679, "num_tokens": 782961200.0, "step": 20617 }, { "epoch": 2.6228215239791375, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6937201023101807, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8789422512054443, "num_tokens": 782995832.0, "step": 20618 }, { "epoch": 2.622948734257728, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.685206890106201, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8799324035644531, "num_tokens": 783040880.0, "step": 20619 }, { "epoch": 2.6230759445363185, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.705264091491699, "learning_rate": 1e-05, "loss": 0.3919, "mean_token_accuracy": 0.8863101601600647, "num_tokens": 783075068.0, "step": 20620 }, { "epoch": 2.623203154814909, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6905078887939453, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.882504940032959, "num_tokens": 783109321.0, "step": 20621 }, { "epoch": 2.6233303650934996, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7048120498657227, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8838156461715698, "num_tokens": 783145222.0, "step": 20622 }, { "epoch": 2.62345757537209, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6886892318725586, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8717293739318848, "num_tokens": 783185017.0, "step": 20623 }, { "epoch": 2.6235847856506807, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.700958251953125, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.873763918876648, "num_tokens": 783228137.0, "step": 20624 }, { "epoch": 2.623711995929271, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6948134899139404, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.876663088798523, "num_tokens": 783266525.0, "step": 20625 }, { "epoch": 2.6238392062078617, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.701296091079712, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8755449652671814, "num_tokens": 783301090.0, "step": 20626 }, { "epoch": 2.6239664164864522, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6884942054748535, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8815334439277649, "num_tokens": 783336816.0, "step": 20627 }, { "epoch": 2.6240936267650428, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.691925525665283, "learning_rate": 1e-05, "loss": 0.3915, "mean_token_accuracy": 0.8913019895553589, "num_tokens": 783376766.0, "step": 20628 }, { "epoch": 2.6242208370436333, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.683267831802368, "learning_rate": 1e-05, "loss": 0.385, "mean_token_accuracy": 0.8908380270004272, "num_tokens": 783416780.0, "step": 20629 }, { "epoch": 2.624348047322224, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.693819999694824, "learning_rate": 1e-05, "loss": 0.3876, "mean_token_accuracy": 0.8913509845733643, "num_tokens": 783449957.0, "step": 20630 }, { "epoch": 2.624475257600814, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7022228240966797, "learning_rate": 1e-05, "loss": 0.365, "mean_token_accuracy": 0.8968029022216797, "num_tokens": 783480488.0, "step": 20631 }, { "epoch": 2.624602467879405, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6810474395751953, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8770017623901367, "num_tokens": 783521296.0, "step": 20632 }, { "epoch": 2.624729678157995, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6958112716674805, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8708174228668213, "num_tokens": 783561937.0, "step": 20633 }, { "epoch": 2.624856888436586, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6816329956054688, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8870365619659424, "num_tokens": 783598928.0, "step": 20634 }, { "epoch": 2.624984098715176, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.69160795211792, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.8856629729270935, "num_tokens": 783635538.0, "step": 20635 }, { "epoch": 2.6251113089937665, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6773905754089355, "learning_rate": 1e-05, "loss": 0.3833, "mean_token_accuracy": 0.8916640281677246, "num_tokens": 783675712.0, "step": 20636 }, { "epoch": 2.625238519272357, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.694974422454834, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8807754516601562, "num_tokens": 783715091.0, "step": 20637 }, { "epoch": 2.6253657295509476, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.679936408996582, "learning_rate": 1e-05, "loss": 0.3915, "mean_token_accuracy": 0.8889157772064209, "num_tokens": 783753263.0, "step": 20638 }, { "epoch": 2.625492939829538, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7199535369873047, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8739112615585327, "num_tokens": 783786251.0, "step": 20639 }, { "epoch": 2.6256201501081287, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.663254976272583, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8813165426254272, "num_tokens": 783829751.0, "step": 20640 }, { "epoch": 2.625747360386719, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7279927730560303, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8711402416229248, "num_tokens": 783867786.0, "step": 20641 }, { "epoch": 2.6258745706653097, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.678957939147949, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8723931908607483, "num_tokens": 783904245.0, "step": 20642 }, { "epoch": 2.6260017809439002, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.696925640106201, "learning_rate": 1e-05, "loss": 0.3818, "mean_token_accuracy": 0.8924240469932556, "num_tokens": 783946892.0, "step": 20643 }, { "epoch": 2.6261289912224908, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.687126398086548, "learning_rate": 1e-05, "loss": 0.4686, "mean_token_accuracy": 0.867881178855896, "num_tokens": 783988132.0, "step": 20644 }, { "epoch": 2.6262562015010813, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.720255136489868, "learning_rate": 1e-05, "loss": 0.4716, "mean_token_accuracy": 0.8701380491256714, "num_tokens": 784024321.0, "step": 20645 }, { "epoch": 2.626383411779672, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6827616691589355, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8735324144363403, "num_tokens": 784061266.0, "step": 20646 }, { "epoch": 2.6265106220582624, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7017178535461426, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8756073713302612, "num_tokens": 784101991.0, "step": 20647 }, { "epoch": 2.626637832336853, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6785759925842285, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8799667358398438, "num_tokens": 784137726.0, "step": 20648 }, { "epoch": 2.6267650426154434, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6978490352630615, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8761316537857056, "num_tokens": 784177773.0, "step": 20649 }, { "epoch": 2.626892252894034, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.681607484817505, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.877923846244812, "num_tokens": 784217429.0, "step": 20650 }, { "epoch": 2.6270194631726245, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7003467082977295, "learning_rate": 1e-05, "loss": 0.385, "mean_token_accuracy": 0.8930863738059998, "num_tokens": 784252289.0, "step": 20651 }, { "epoch": 2.627146673451215, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6829943656921387, "learning_rate": 1e-05, "loss": 0.4992, "mean_token_accuracy": 0.8563235998153687, "num_tokens": 784293472.0, "step": 20652 }, { "epoch": 2.6272738837298055, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6909821033477783, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8762431144714355, "num_tokens": 784339423.0, "step": 20653 }, { "epoch": 2.6274010940083956, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7058393955230713, "learning_rate": 1e-05, "loss": 0.3722, "mean_token_accuracy": 0.8910698890686035, "num_tokens": 784367026.0, "step": 20654 }, { "epoch": 2.6275283042869866, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.694413185119629, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8799829483032227, "num_tokens": 784404004.0, "step": 20655 }, { "epoch": 2.6276555145655767, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6909384727478027, "learning_rate": 1e-05, "loss": 0.3819, "mean_token_accuracy": 0.8917104005813599, "num_tokens": 784439335.0, "step": 20656 }, { "epoch": 2.6277827248441676, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.682163715362549, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.873191237449646, "num_tokens": 784480141.0, "step": 20657 }, { "epoch": 2.6279099351227577, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7088968753814697, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8830289840698242, "num_tokens": 784514785.0, "step": 20658 }, { "epoch": 2.6280371454013487, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.691539764404297, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8703179359436035, "num_tokens": 784549776.0, "step": 20659 }, { "epoch": 2.628164355679939, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.743204116821289, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8656088709831238, "num_tokens": 784578558.0, "step": 20660 }, { "epoch": 2.6282915659585293, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6674444675445557, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8836239576339722, "num_tokens": 784614056.0, "step": 20661 }, { "epoch": 2.62841877623712, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7177011966705322, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.8821326494216919, "num_tokens": 784651663.0, "step": 20662 }, { "epoch": 2.6285459865157104, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.662973642349243, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8716132640838623, "num_tokens": 784690343.0, "step": 20663 }, { "epoch": 2.628673196794301, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7207584381103516, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8813958168029785, "num_tokens": 784728569.0, "step": 20664 }, { "epoch": 2.6288004070728914, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6644093990325928, "learning_rate": 1e-05, "loss": 0.3901, "mean_token_accuracy": 0.8875436186790466, "num_tokens": 784762376.0, "step": 20665 }, { "epoch": 2.628927617351482, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.738292932510376, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8793197274208069, "num_tokens": 784793154.0, "step": 20666 }, { "epoch": 2.6290548276300725, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6665685176849365, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8814929723739624, "num_tokens": 784829640.0, "step": 20667 }, { "epoch": 2.629182037908663, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.724001169204712, "learning_rate": 1e-05, "loss": 0.3937, "mean_token_accuracy": 0.8901721239089966, "num_tokens": 784867045.0, "step": 20668 }, { "epoch": 2.6293092481872535, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6555817127227783, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8816614151000977, "num_tokens": 784907157.0, "step": 20669 }, { "epoch": 2.629436458465844, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.73710560798645, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8820450305938721, "num_tokens": 784942247.0, "step": 20670 }, { "epoch": 2.6295636687444346, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6715102195739746, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8777788281440735, "num_tokens": 784976910.0, "step": 20671 }, { "epoch": 2.629690879023025, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7350428104400635, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8617086410522461, "num_tokens": 785017969.0, "step": 20672 }, { "epoch": 2.6298180893016156, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.659161329269409, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8830364942550659, "num_tokens": 785056182.0, "step": 20673 }, { "epoch": 2.629945299580206, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.736452102661133, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8885098695755005, "num_tokens": 785091996.0, "step": 20674 }, { "epoch": 2.6300725098587967, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.648768424987793, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8857229948043823, "num_tokens": 785129268.0, "step": 20675 }, { "epoch": 2.6301997201373872, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7309346199035645, "learning_rate": 1e-05, "loss": 0.3867, "mean_token_accuracy": 0.8906532526016235, "num_tokens": 785168418.0, "step": 20676 }, { "epoch": 2.6303269304159778, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6437203884124756, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8853906393051147, "num_tokens": 785209981.0, "step": 20677 }, { "epoch": 2.6304541406945683, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7396719455718994, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8827007412910461, "num_tokens": 785250354.0, "step": 20678 }, { "epoch": 2.6305813509731584, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6505982875823975, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8789808750152588, "num_tokens": 785288594.0, "step": 20679 }, { "epoch": 2.6307085612517493, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7428135871887207, "learning_rate": 1e-05, "loss": 0.3777, "mean_token_accuracy": 0.8924550414085388, "num_tokens": 785323885.0, "step": 20680 }, { "epoch": 2.6308357715303394, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6410746574401855, "learning_rate": 1e-05, "loss": 0.3821, "mean_token_accuracy": 0.8934010863304138, "num_tokens": 785361950.0, "step": 20681 }, { "epoch": 2.6309629818089304, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.757664442062378, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8733291029930115, "num_tokens": 785398650.0, "step": 20682 }, { "epoch": 2.6310901920875205, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.623434066772461, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8721702098846436, "num_tokens": 785451273.0, "step": 20683 }, { "epoch": 2.6312174023661115, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.764575481414795, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8759263753890991, "num_tokens": 785491244.0, "step": 20684 }, { "epoch": 2.6313446126447015, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6149861812591553, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.881000280380249, "num_tokens": 785536558.0, "step": 20685 }, { "epoch": 2.631471822923292, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8009109497070312, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8751009106636047, "num_tokens": 785572635.0, "step": 20686 }, { "epoch": 2.6315990332018826, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6214637756347656, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8768547177314758, "num_tokens": 785615765.0, "step": 20687 }, { "epoch": 2.631726243480473, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.80061936378479, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8812072277069092, "num_tokens": 785652934.0, "step": 20688 }, { "epoch": 2.6318534537590637, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6285557746887207, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8755057454109192, "num_tokens": 785689018.0, "step": 20689 }, { "epoch": 2.631980664037654, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7977089881896973, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8731586933135986, "num_tokens": 785728257.0, "step": 20690 }, { "epoch": 2.6321078743162447, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6263577938079834, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8719422817230225, "num_tokens": 785768381.0, "step": 20691 }, { "epoch": 2.6322350845948352, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.809130907058716, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.877274751663208, "num_tokens": 785806700.0, "step": 20692 }, { "epoch": 2.6323622948734258, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.623307466506958, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.864190399646759, "num_tokens": 785848602.0, "step": 20693 }, { "epoch": 2.6324895051520163, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.815465211868286, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8820961713790894, "num_tokens": 785888714.0, "step": 20694 }, { "epoch": 2.632616715430607, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.623168468475342, "learning_rate": 1e-05, "loss": 0.385, "mean_token_accuracy": 0.8917044997215271, "num_tokens": 785925163.0, "step": 20695 }, { "epoch": 2.6327439257091974, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.829745054244995, "learning_rate": 1e-05, "loss": 0.4676, "mean_token_accuracy": 0.8627384901046753, "num_tokens": 785963566.0, "step": 20696 }, { "epoch": 2.632871135987788, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6381850242614746, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8693938255310059, "num_tokens": 785997654.0, "step": 20697 }, { "epoch": 2.6329983462663784, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8199353218078613, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8744802474975586, "num_tokens": 786038761.0, "step": 20698 }, { "epoch": 2.633125556544969, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6447842121124268, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8837331533432007, "num_tokens": 786072380.0, "step": 20699 }, { "epoch": 2.6332527668235595, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8014750480651855, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8779846429824829, "num_tokens": 786112751.0, "step": 20700 }, { "epoch": 2.63337997710215, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6351022720336914, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8796091675758362, "num_tokens": 786156841.0, "step": 20701 }, { "epoch": 2.6335071873807405, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.820439338684082, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8736591339111328, "num_tokens": 786191015.0, "step": 20702 }, { "epoch": 2.633634397659331, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6275577545166016, "learning_rate": 1e-05, "loss": 0.4729, "mean_token_accuracy": 0.8615568280220032, "num_tokens": 786231420.0, "step": 20703 }, { "epoch": 2.633761607937921, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8196332454681396, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8868618607521057, "num_tokens": 786266586.0, "step": 20704 }, { "epoch": 2.633888818216512, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6268465518951416, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8778952360153198, "num_tokens": 786307489.0, "step": 20705 }, { "epoch": 2.634016028495102, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.830571174621582, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8845492005348206, "num_tokens": 786340960.0, "step": 20706 }, { "epoch": 2.634143238773693, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6259143352508545, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8761289119720459, "num_tokens": 786383086.0, "step": 20707 }, { "epoch": 2.6342704490522832, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.819359064102173, "learning_rate": 1e-05, "loss": 0.3986, "mean_token_accuracy": 0.8859168291091919, "num_tokens": 786421093.0, "step": 20708 }, { "epoch": 2.6343976593308738, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6306538581848145, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8779576420783997, "num_tokens": 786462712.0, "step": 20709 }, { "epoch": 2.6345248696094643, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003566741943359375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.82270884513855, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8770262002944946, "num_tokens": 786501607.0, "step": 20710 }, { "epoch": 2.634652079888055, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.640754461288452, "learning_rate": 1e-05, "loss": 0.4121, "mean_token_accuracy": 0.8819190263748169, "num_tokens": 786537163.0, "step": 20711 }, { "epoch": 2.6347792901666454, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8214919567108154, "learning_rate": 1e-05, "loss": 0.3909, "mean_token_accuracy": 0.8890440464019775, "num_tokens": 786572880.0, "step": 20712 }, { "epoch": 2.634906500445236, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.642634391784668, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8820145130157471, "num_tokens": 786610082.0, "step": 20713 }, { "epoch": 2.6350337107238264, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8164851665496826, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8776125311851501, "num_tokens": 786648875.0, "step": 20714 }, { "epoch": 2.635160921002417, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.642993688583374, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8749868273735046, "num_tokens": 786688076.0, "step": 20715 }, { "epoch": 2.6352881312810075, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8243050575256348, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8766113519668579, "num_tokens": 786723969.0, "step": 20716 }, { "epoch": 2.635415341559598, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.641911268234253, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.8734325170516968, "num_tokens": 786764018.0, "step": 20717 }, { "epoch": 2.6355425518381885, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.814558267593384, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8856685161590576, "num_tokens": 786796586.0, "step": 20718 }, { "epoch": 2.635669762116779, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.656592607498169, "learning_rate": 1e-05, "loss": 0.4607, "mean_token_accuracy": 0.8669299483299255, "num_tokens": 786833797.0, "step": 20719 }, { "epoch": 2.6357969723953696, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.809234142303467, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8691667318344116, "num_tokens": 786869971.0, "step": 20720 }, { "epoch": 2.63592418267396, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.656754970550537, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8826141357421875, "num_tokens": 786905854.0, "step": 20721 }, { "epoch": 2.6360513929525506, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.802079677581787, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8794823884963989, "num_tokens": 786940535.0, "step": 20722 }, { "epoch": 2.636178603231141, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.659677028656006, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8686086535453796, "num_tokens": 786974976.0, "step": 20723 }, { "epoch": 2.6363058135097317, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.792682647705078, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8770564794540405, "num_tokens": 787013389.0, "step": 20724 }, { "epoch": 2.6364330237883222, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6545865535736084, "learning_rate": 1e-05, "loss": 0.3761, "mean_token_accuracy": 0.8932209014892578, "num_tokens": 787044463.0, "step": 20725 }, { "epoch": 2.6365602340669128, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8024139404296875, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8825359344482422, "num_tokens": 787075123.0, "step": 20726 }, { "epoch": 2.636687444345503, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6454896926879883, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8746029138565063, "num_tokens": 787114790.0, "step": 20727 }, { "epoch": 2.636814654624094, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.783888339996338, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8899239301681519, "num_tokens": 787153263.0, "step": 20728 }, { "epoch": 2.636941864902684, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6414310932159424, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.887306809425354, "num_tokens": 787191955.0, "step": 20729 }, { "epoch": 2.637069075181275, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7948102951049805, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8814492225646973, "num_tokens": 787228330.0, "step": 20730 }, { "epoch": 2.637196285459865, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6473278999328613, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8585293292999268, "num_tokens": 787271927.0, "step": 20731 }, { "epoch": 2.637323495738456, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7924137115478516, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8906111717224121, "num_tokens": 787310515.0, "step": 20732 }, { "epoch": 2.637450706017046, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6479833126068115, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8779425621032715, "num_tokens": 787346720.0, "step": 20733 }, { "epoch": 2.6375779162956365, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.798687219619751, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8758214712142944, "num_tokens": 787387591.0, "step": 20734 }, { "epoch": 2.637705126574227, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6401796340942383, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8776630759239197, "num_tokens": 787427762.0, "step": 20735 }, { "epoch": 2.6378323368528176, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.816894769668579, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8689741492271423, "num_tokens": 787463047.0, "step": 20736 }, { "epoch": 2.637959547131408, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.628835678100586, "learning_rate": 1e-05, "loss": 0.381, "mean_token_accuracy": 0.8935642242431641, "num_tokens": 787500862.0, "step": 20737 }, { "epoch": 2.6380867574099987, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8260841369628906, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8895125389099121, "num_tokens": 787529625.0, "step": 20738 }, { "epoch": 2.638213967688589, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6295039653778076, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8829176425933838, "num_tokens": 787567248.0, "step": 20739 }, { "epoch": 2.6383411779671797, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8270297050476074, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8756684064865112, "num_tokens": 787606763.0, "step": 20740 }, { "epoch": 2.6384683882457702, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6379590034484863, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8773884773254395, "num_tokens": 787638795.0, "step": 20741 }, { "epoch": 2.6385955985243608, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8160011768341064, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8737573623657227, "num_tokens": 787679495.0, "step": 20742 }, { "epoch": 2.6387228088029513, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6302108764648438, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8812366724014282, "num_tokens": 787718583.0, "step": 20743 }, { "epoch": 2.638850019081542, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8110909461975098, "learning_rate": 1e-05, "loss": 0.372, "mean_token_accuracy": 0.8959261178970337, "num_tokens": 787756558.0, "step": 20744 }, { "epoch": 2.6389772293601323, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6498167514801025, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8809384107589722, "num_tokens": 787789796.0, "step": 20745 }, { "epoch": 2.639104439638723, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8076694011688232, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8689942359924316, "num_tokens": 787832054.0, "step": 20746 }, { "epoch": 2.6392316499173134, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6365818977355957, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8788756132125854, "num_tokens": 787876723.0, "step": 20747 }, { "epoch": 2.639358860195904, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.818699836730957, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8684109449386597, "num_tokens": 787912183.0, "step": 20748 }, { "epoch": 2.6394860704744945, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6576852798461914, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8683754205703735, "num_tokens": 787946865.0, "step": 20749 }, { "epoch": 2.639613280753085, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.7997896671295166, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8882584571838379, "num_tokens": 787984286.0, "step": 20750 }, { "epoch": 2.6397404910316755, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6458823680877686, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8916556239128113, "num_tokens": 788020565.0, "step": 20751 }, { "epoch": 2.6398677013102656, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.802419900894165, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8884866237640381, "num_tokens": 788051833.0, "step": 20752 }, { "epoch": 2.6399949115888566, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6343936920166016, "learning_rate": 1e-05, "loss": 0.3788, "mean_token_accuracy": 0.891944169998169, "num_tokens": 788092780.0, "step": 20753 }, { "epoch": 2.6401221218674467, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7937090396881104, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8729802370071411, "num_tokens": 788140088.0, "step": 20754 }, { "epoch": 2.6402493321460376, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.637911319732666, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8835883140563965, "num_tokens": 788177951.0, "step": 20755 }, { "epoch": 2.6403765424246277, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.814542055130005, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8855386972427368, "num_tokens": 788212124.0, "step": 20756 }, { "epoch": 2.6405037527032187, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.625520706176758, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8775069117546082, "num_tokens": 788255860.0, "step": 20757 }, { "epoch": 2.6406309629818088, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8041772842407227, "learning_rate": 1e-05, "loss": 0.3902, "mean_token_accuracy": 0.88869309425354, "num_tokens": 788296481.0, "step": 20758 }, { "epoch": 2.6407581732603993, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.645260810852051, "learning_rate": 1e-05, "loss": 0.3853, "mean_token_accuracy": 0.8897445201873779, "num_tokens": 788328818.0, "step": 20759 }, { "epoch": 2.64088538353899, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7904629707336426, "learning_rate": 1e-05, "loss": 0.3705, "mean_token_accuracy": 0.8970566391944885, "num_tokens": 788370359.0, "step": 20760 }, { "epoch": 2.6410125938175804, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.635538101196289, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8796675205230713, "num_tokens": 788412425.0, "step": 20761 }, { "epoch": 2.641139804096171, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.819762945175171, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8650545477867126, "num_tokens": 788445137.0, "step": 20762 }, { "epoch": 2.6412670143747614, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.642369031906128, "learning_rate": 1e-05, "loss": 0.3916, "mean_token_accuracy": 0.8870500326156616, "num_tokens": 788478185.0, "step": 20763 }, { "epoch": 2.641394224653352, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8097777366638184, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8778629302978516, "num_tokens": 788514329.0, "step": 20764 }, { "epoch": 2.6415214349319425, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.631528615951538, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.883560061454773, "num_tokens": 788554234.0, "step": 20765 }, { "epoch": 2.641648645210533, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8044772148132324, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8834158778190613, "num_tokens": 788591684.0, "step": 20766 }, { "epoch": 2.6417758554891235, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.63289475440979, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.8825160264968872, "num_tokens": 788626802.0, "step": 20767 }, { "epoch": 2.641903065767714, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.795969247817993, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8837994337081909, "num_tokens": 788673826.0, "step": 20768 }, { "epoch": 2.6420302760463046, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.635624647140503, "learning_rate": 1e-05, "loss": 0.3881, "mean_token_accuracy": 0.8891661167144775, "num_tokens": 788708393.0, "step": 20769 }, { "epoch": 2.642157486324895, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7944064140319824, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8880372047424316, "num_tokens": 788746480.0, "step": 20770 }, { "epoch": 2.6422846966034856, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.640559673309326, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8796960115432739, "num_tokens": 788785295.0, "step": 20771 }, { "epoch": 2.642411906882076, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7966954708099365, "learning_rate": 1e-05, "loss": 0.4043, "mean_token_accuracy": 0.884719967842102, "num_tokens": 788818803.0, "step": 20772 }, { "epoch": 2.6425391171606667, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6440815925598145, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8768530488014221, "num_tokens": 788857266.0, "step": 20773 }, { "epoch": 2.6426663274392572, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7929527759552, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8620124459266663, "num_tokens": 788898433.0, "step": 20774 }, { "epoch": 2.6427935377178478, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6505162715911865, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8741284608840942, "num_tokens": 788942259.0, "step": 20775 }, { "epoch": 2.6429207479964383, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.785694122314453, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8705583214759827, "num_tokens": 788981484.0, "step": 20776 }, { "epoch": 2.6430479582750284, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.644040107727051, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.881801426410675, "num_tokens": 789019863.0, "step": 20777 }, { "epoch": 2.6431751685536193, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7778329849243164, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8812772035598755, "num_tokens": 789059139.0, "step": 20778 }, { "epoch": 2.6433023788322094, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.634422540664673, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8842456340789795, "num_tokens": 789100429.0, "step": 20779 }, { "epoch": 2.6434295891108004, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7856438159942627, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8839630484580994, "num_tokens": 789136219.0, "step": 20780 }, { "epoch": 2.6435567993893905, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.630706787109375, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8830727338790894, "num_tokens": 789177541.0, "step": 20781 }, { "epoch": 2.6436840096679814, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.777867317199707, "learning_rate": 1e-05, "loss": 0.3729, "mean_token_accuracy": 0.8965786695480347, "num_tokens": 789217165.0, "step": 20782 }, { "epoch": 2.6438112199465715, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6340742111206055, "learning_rate": 1e-05, "loss": 0.4017, "mean_token_accuracy": 0.8838201761245728, "num_tokens": 789255384.0, "step": 20783 }, { "epoch": 2.643938430225162, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7876956462860107, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8833625316619873, "num_tokens": 789294153.0, "step": 20784 }, { "epoch": 2.6440656405037526, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6417582035064697, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8621337413787842, "num_tokens": 789334933.0, "step": 20785 }, { "epoch": 2.644192850782343, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.798635244369507, "learning_rate": 1e-05, "loss": 0.4796, "mean_token_accuracy": 0.8598949313163757, "num_tokens": 789370068.0, "step": 20786 }, { "epoch": 2.6443200610609336, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.620131492614746, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.8849924802780151, "num_tokens": 789413887.0, "step": 20787 }, { "epoch": 2.644447271339524, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.799462080001831, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8904991149902344, "num_tokens": 789452319.0, "step": 20788 }, { "epoch": 2.6445744816181147, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.6298654079437256, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.8884338736534119, "num_tokens": 789488988.0, "step": 20789 }, { "epoch": 2.6447016918967052, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.798804759979248, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8736860752105713, "num_tokens": 789529849.0, "step": 20790 }, { "epoch": 2.6448289021752958, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6294631958007812, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8774849772453308, "num_tokens": 789572877.0, "step": 20791 }, { "epoch": 2.6449561124538863, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8287887573242188, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8690282106399536, "num_tokens": 789605189.0, "step": 20792 }, { "epoch": 2.645083322732477, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6203863620758057, "learning_rate": 1e-05, "loss": 0.399, "mean_token_accuracy": 0.8872497081756592, "num_tokens": 789641589.0, "step": 20793 }, { "epoch": 2.6452105330110673, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.858480930328369, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8699193000793457, "num_tokens": 789673869.0, "step": 20794 }, { "epoch": 2.645337743289658, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6216697692871094, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8790152668952942, "num_tokens": 789710605.0, "step": 20795 }, { "epoch": 2.6454649535682484, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.836921215057373, "learning_rate": 1e-05, "loss": 0.3914, "mean_token_accuracy": 0.8882993459701538, "num_tokens": 789747641.0, "step": 20796 }, { "epoch": 2.645592163846839, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6226463317871094, "learning_rate": 1e-05, "loss": 0.3717, "mean_token_accuracy": 0.8946014046669006, "num_tokens": 789781345.0, "step": 20797 }, { "epoch": 2.6457193741254295, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.85353946685791, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8770015835762024, "num_tokens": 789819582.0, "step": 20798 }, { "epoch": 2.64584658440402, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6254467964172363, "learning_rate": 1e-05, "loss": 0.3801, "mean_token_accuracy": 0.8902261257171631, "num_tokens": 789856281.0, "step": 20799 }, { "epoch": 2.6459737946826105, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.84602689743042, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8895061612129211, "num_tokens": 789895711.0, "step": 20800 }, { "epoch": 2.646101004961201, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.641035318374634, "learning_rate": 1e-05, "loss": 0.3889, "mean_token_accuracy": 0.8936988115310669, "num_tokens": 789931751.0, "step": 20801 }, { "epoch": 2.646228215239791, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8537471294403076, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8753284215927124, "num_tokens": 789969207.0, "step": 20802 }, { "epoch": 2.646355425518382, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.642901659011841, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8775637149810791, "num_tokens": 790005706.0, "step": 20803 }, { "epoch": 2.646482635796972, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.8462722301483154, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8781195878982544, "num_tokens": 790047389.0, "step": 20804 }, { "epoch": 2.646609846075563, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6527156829833984, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8825408220291138, "num_tokens": 790085430.0, "step": 20805 }, { "epoch": 2.6467370563541532, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.827251672744751, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8761883974075317, "num_tokens": 790126620.0, "step": 20806 }, { "epoch": 2.6468642666327438, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6392416954040527, "learning_rate": 1e-05, "loss": 0.4015, "mean_token_accuracy": 0.8840625286102295, "num_tokens": 790166033.0, "step": 20807 }, { "epoch": 2.6469914769113343, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8269965648651123, "learning_rate": 1e-05, "loss": 0.3904, "mean_token_accuracy": 0.8894025087356567, "num_tokens": 790202463.0, "step": 20808 }, { "epoch": 2.647118687189925, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6425058841705322, "learning_rate": 1e-05, "loss": 0.4061, "mean_token_accuracy": 0.8838738203048706, "num_tokens": 790241600.0, "step": 20809 }, { "epoch": 2.6472458974685154, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8444790840148926, "learning_rate": 1e-05, "loss": 0.4924, "mean_token_accuracy": 0.8574638366699219, "num_tokens": 790277669.0, "step": 20810 }, { "epoch": 2.647373107747106, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6458961963653564, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8724924325942993, "num_tokens": 790317453.0, "step": 20811 }, { "epoch": 2.6475003180256964, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.826803207397461, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8797091841697693, "num_tokens": 790354460.0, "step": 20812 }, { "epoch": 2.647627528304287, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6389217376708984, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8805186152458191, "num_tokens": 790393933.0, "step": 20813 }, { "epoch": 2.6477547385828775, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8396503925323486, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8812016844749451, "num_tokens": 790425477.0, "step": 20814 }, { "epoch": 2.647881948861468, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6355509757995605, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8599015474319458, "num_tokens": 790470501.0, "step": 20815 }, { "epoch": 2.6480091591400585, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8380184173583984, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.875624418258667, "num_tokens": 790508869.0, "step": 20816 }, { "epoch": 2.648136369418649, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6661384105682373, "learning_rate": 1e-05, "loss": 0.3716, "mean_token_accuracy": 0.8959431648254395, "num_tokens": 790545675.0, "step": 20817 }, { "epoch": 2.6482635796972396, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8145925998687744, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8837148547172546, "num_tokens": 790586832.0, "step": 20818 }, { "epoch": 2.64839078997583, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6397483348846436, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8768736124038696, "num_tokens": 790633587.0, "step": 20819 }, { "epoch": 2.6485180002544206, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.813136100769043, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8709132075309753, "num_tokens": 790668768.0, "step": 20820 }, { "epoch": 2.648645210533011, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6423912048339844, "learning_rate": 1e-05, "loss": 0.393, "mean_token_accuracy": 0.8888790011405945, "num_tokens": 790708393.0, "step": 20821 }, { "epoch": 2.6487724208116017, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.818969249725342, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8684984445571899, "num_tokens": 790742632.0, "step": 20822 }, { "epoch": 2.648899631090192, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6580612659454346, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8674592971801758, "num_tokens": 790779740.0, "step": 20823 }, { "epoch": 2.6490268413687827, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7907440662384033, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.8883696794509888, "num_tokens": 790819040.0, "step": 20824 }, { "epoch": 2.649154051647373, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.649461507797241, "learning_rate": 1e-05, "loss": 0.3803, "mean_token_accuracy": 0.894088089466095, "num_tokens": 790853382.0, "step": 20825 }, { "epoch": 2.649281261925964, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7970662117004395, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8696015477180481, "num_tokens": 790890606.0, "step": 20826 }, { "epoch": 2.649408472204554, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.650228261947632, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8868191838264465, "num_tokens": 790929469.0, "step": 20827 }, { "epoch": 2.649535682483145, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7799010276794434, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8789205551147461, "num_tokens": 790970745.0, "step": 20828 }, { "epoch": 2.649662892761735, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.669609308242798, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8705136775970459, "num_tokens": 791010365.0, "step": 20829 }, { "epoch": 2.649790103040326, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7695159912109375, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8819448947906494, "num_tokens": 791047012.0, "step": 20830 }, { "epoch": 2.649917313318916, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6438663005828857, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.8839563727378845, "num_tokens": 791092836.0, "step": 20831 }, { "epoch": 2.6500445235975065, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.779787540435791, "learning_rate": 1e-05, "loss": 0.3954, "mean_token_accuracy": 0.8898128867149353, "num_tokens": 791127725.0, "step": 20832 }, { "epoch": 2.650171733876097, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6522104740142822, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8873422145843506, "num_tokens": 791164530.0, "step": 20833 }, { "epoch": 2.6502989441546876, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7685654163360596, "learning_rate": 1e-05, "loss": 0.3731, "mean_token_accuracy": 0.8947114944458008, "num_tokens": 791202384.0, "step": 20834 }, { "epoch": 2.650426154433278, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6420373916625977, "learning_rate": 1e-05, "loss": 0.3736, "mean_token_accuracy": 0.8947545886039734, "num_tokens": 791237998.0, "step": 20835 }, { "epoch": 2.6505533647118686, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.78855037689209, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8592743873596191, "num_tokens": 791277755.0, "step": 20836 }, { "epoch": 2.650680574990459, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6468074321746826, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8691275119781494, "num_tokens": 791315779.0, "step": 20837 }, { "epoch": 2.6508077852690497, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7734060287475586, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8762555122375488, "num_tokens": 791359212.0, "step": 20838 }, { "epoch": 2.6509349955476402, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.651254892349243, "learning_rate": 1e-05, "loss": 0.3641, "mean_token_accuracy": 0.8976033926010132, "num_tokens": 791388308.0, "step": 20839 }, { "epoch": 2.6510622058262308, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.780101776123047, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8781303763389587, "num_tokens": 791425645.0, "step": 20840 }, { "epoch": 2.6511894161048213, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6421828269958496, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8795292973518372, "num_tokens": 791462632.0, "step": 20841 }, { "epoch": 2.651316626383412, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.786057233810425, "learning_rate": 1e-05, "loss": 0.3666, "mean_token_accuracy": 0.8974114656448364, "num_tokens": 791502834.0, "step": 20842 }, { "epoch": 2.6514438366620023, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.651841640472412, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8772629499435425, "num_tokens": 791538973.0, "step": 20843 }, { "epoch": 2.651571046940593, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.774336338043213, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8817927837371826, "num_tokens": 791577804.0, "step": 20844 }, { "epoch": 2.6516982572191834, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6315701007843018, "learning_rate": 1e-05, "loss": 0.3732, "mean_token_accuracy": 0.8921399116516113, "num_tokens": 791618100.0, "step": 20845 }, { "epoch": 2.651825467497774, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7823293209075928, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8769386410713196, "num_tokens": 791660112.0, "step": 20846 }, { "epoch": 2.6519526777763645, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.64123272895813, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8784914016723633, "num_tokens": 791698133.0, "step": 20847 }, { "epoch": 2.652079888054955, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7843246459960938, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.886880099773407, "num_tokens": 791737810.0, "step": 20848 }, { "epoch": 2.6522070983335455, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.644042730331421, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8753834366798401, "num_tokens": 791775187.0, "step": 20849 }, { "epoch": 2.6523343086121356, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.787442445755005, "learning_rate": 1e-05, "loss": 0.3987, "mean_token_accuracy": 0.8878056406974792, "num_tokens": 791810673.0, "step": 20850 }, { "epoch": 2.6524615188907266, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6249871253967285, "learning_rate": 1e-05, "loss": 0.3534, "mean_token_accuracy": 0.9017210602760315, "num_tokens": 791848144.0, "step": 20851 }, { "epoch": 2.6525887291693167, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8013551235198975, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8797168731689453, "num_tokens": 791885025.0, "step": 20852 }, { "epoch": 2.6527159394479076, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.639613151550293, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8779758214950562, "num_tokens": 791922885.0, "step": 20853 }, { "epoch": 2.6528431497264977, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8058903217315674, "learning_rate": 1e-05, "loss": 0.4807, "mean_token_accuracy": 0.8613712787628174, "num_tokens": 791959791.0, "step": 20854 }, { "epoch": 2.6529703600050887, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.625976324081421, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8820018768310547, "num_tokens": 791998559.0, "step": 20855 }, { "epoch": 2.6530975702836788, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8123927116394043, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8800493478775024, "num_tokens": 792038428.0, "step": 20856 }, { "epoch": 2.6532247805622693, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.644139051437378, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8687475323677063, "num_tokens": 792074405.0, "step": 20857 }, { "epoch": 2.65335199084086, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.811814785003662, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8697429895401001, "num_tokens": 792112831.0, "step": 20858 }, { "epoch": 2.6534792011194503, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.633561134338379, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8753361701965332, "num_tokens": 792154122.0, "step": 20859 }, { "epoch": 2.653606411398041, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.823819637298584, "learning_rate": 1e-05, "loss": 0.4733, "mean_token_accuracy": 0.8627632856369019, "num_tokens": 792191818.0, "step": 20860 }, { "epoch": 2.6537336216766314, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6320552825927734, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8796291351318359, "num_tokens": 792226753.0, "step": 20861 }, { "epoch": 2.653860831955222, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.830049753189087, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8787350654602051, "num_tokens": 792263941.0, "step": 20862 }, { "epoch": 2.6539880422338125, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.62581467628479, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8833594918251038, "num_tokens": 792301433.0, "step": 20863 }, { "epoch": 2.654115252512403, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8212568759918213, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.8802056312561035, "num_tokens": 792340188.0, "step": 20864 }, { "epoch": 2.6542424627909935, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6326615810394287, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.873971700668335, "num_tokens": 792383498.0, "step": 20865 }, { "epoch": 2.654369673069584, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8248274326324463, "learning_rate": 1e-05, "loss": 0.3911, "mean_token_accuracy": 0.8900147676467896, "num_tokens": 792417160.0, "step": 20866 }, { "epoch": 2.6544968833481746, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.643766164779663, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8768289089202881, "num_tokens": 792451609.0, "step": 20867 }, { "epoch": 2.654624093626765, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.804687738418579, "learning_rate": 1e-05, "loss": 0.4677, "mean_token_accuracy": 0.8676089644432068, "num_tokens": 792498309.0, "step": 20868 }, { "epoch": 2.6547513039053556, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.644329786300659, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.870627760887146, "num_tokens": 792538954.0, "step": 20869 }, { "epoch": 2.654878514183946, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.811495542526245, "learning_rate": 1e-05, "loss": 0.4663, "mean_token_accuracy": 0.865481972694397, "num_tokens": 792577037.0, "step": 20870 }, { "epoch": 2.6550057244625367, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6646530628204346, "learning_rate": 1e-05, "loss": 0.4504, "mean_token_accuracy": 0.86883145570755, "num_tokens": 792609538.0, "step": 20871 }, { "epoch": 2.655132934741127, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.793123483657837, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8805661201477051, "num_tokens": 792647793.0, "step": 20872 }, { "epoch": 2.6552601450197177, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.651359796524048, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8835704326629639, "num_tokens": 792686746.0, "step": 20873 }, { "epoch": 2.6553873552983083, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.798612594604492, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8820605278015137, "num_tokens": 792716324.0, "step": 20874 }, { "epoch": 2.6555145655768984, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.635676860809326, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8842654824256897, "num_tokens": 792757242.0, "step": 20875 }, { "epoch": 2.6556417758554893, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.797124147415161, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8742995262145996, "num_tokens": 792795545.0, "step": 20876 }, { "epoch": 2.6557689861340794, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6365487575531006, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8769079446792603, "num_tokens": 792835772.0, "step": 20877 }, { "epoch": 2.6558961964126704, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.791635274887085, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.869672417640686, "num_tokens": 792877901.0, "step": 20878 }, { "epoch": 2.6560234066912605, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6358723640441895, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8787073493003845, "num_tokens": 792918528.0, "step": 20879 }, { "epoch": 2.6561506169698514, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7923083305358887, "learning_rate": 1e-05, "loss": 0.3779, "mean_token_accuracy": 0.892888069152832, "num_tokens": 792954251.0, "step": 20880 }, { "epoch": 2.6562778272484415, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.647709846496582, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.883345365524292, "num_tokens": 792991452.0, "step": 20881 }, { "epoch": 2.656405037527032, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8062899112701416, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8742038011550903, "num_tokens": 793024488.0, "step": 20882 }, { "epoch": 2.6565322478056226, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.638345241546631, "learning_rate": 1e-05, "loss": 0.4061, "mean_token_accuracy": 0.8869651556015015, "num_tokens": 793064070.0, "step": 20883 }, { "epoch": 2.656659458084213, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8017327785491943, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8696563839912415, "num_tokens": 793103088.0, "step": 20884 }, { "epoch": 2.6567866683628036, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.643540143966675, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8667301535606384, "num_tokens": 793146415.0, "step": 20885 }, { "epoch": 2.656913878641394, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8058018684387207, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8818390965461731, "num_tokens": 793181588.0, "step": 20886 }, { "epoch": 2.6570410889199847, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6471333503723145, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8632038831710815, "num_tokens": 793217764.0, "step": 20887 }, { "epoch": 2.6571682991985752, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.790477991104126, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8858566284179688, "num_tokens": 793255945.0, "step": 20888 }, { "epoch": 2.6572955094771658, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.648946523666382, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8772399425506592, "num_tokens": 793288981.0, "step": 20889 }, { "epoch": 2.6574227197557563, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7839853763580322, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8825278282165527, "num_tokens": 793327635.0, "step": 20890 }, { "epoch": 2.657549930034347, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.676449775695801, "learning_rate": 1e-05, "loss": 0.4146, "mean_token_accuracy": 0.8796623349189758, "num_tokens": 793357077.0, "step": 20891 }, { "epoch": 2.6576771403129373, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7608494758605957, "learning_rate": 1e-05, "loss": 0.3758, "mean_token_accuracy": 0.8969258069992065, "num_tokens": 793389664.0, "step": 20892 }, { "epoch": 2.657804350591528, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6495261192321777, "learning_rate": 1e-05, "loss": 0.3838, "mean_token_accuracy": 0.8932898044586182, "num_tokens": 793429951.0, "step": 20893 }, { "epoch": 2.6579315608701184, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7576444149017334, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8681639432907104, "num_tokens": 793471404.0, "step": 20894 }, { "epoch": 2.658058771148709, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.657780647277832, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8783812522888184, "num_tokens": 793510140.0, "step": 20895 }, { "epoch": 2.6581859814272994, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7546892166137695, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.881086528301239, "num_tokens": 793553927.0, "step": 20896 }, { "epoch": 2.65831319170589, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6542882919311523, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.8854939341545105, "num_tokens": 793590413.0, "step": 20897 }, { "epoch": 2.6584404019844805, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.8051559925079346, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8602920174598694, "num_tokens": 793626320.0, "step": 20898 }, { "epoch": 2.658567612263071, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.623098611831665, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8816128373146057, "num_tokens": 793665733.0, "step": 20899 }, { "epoch": 2.658694822541661, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7951505184173584, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8789547681808472, "num_tokens": 793705593.0, "step": 20900 }, { "epoch": 2.658822032820252, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.622405767440796, "learning_rate": 1e-05, "loss": 0.4008, "mean_token_accuracy": 0.8841366767883301, "num_tokens": 793744874.0, "step": 20901 }, { "epoch": 2.658949243098842, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8074750900268555, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8657185435295105, "num_tokens": 793786598.0, "step": 20902 }, { "epoch": 2.659076453377433, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.626995801925659, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8780884742736816, "num_tokens": 793826548.0, "step": 20903 }, { "epoch": 2.6592036636560232, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8146398067474365, "learning_rate": 1e-05, "loss": 0.4724, "mean_token_accuracy": 0.8661946654319763, "num_tokens": 793868411.0, "step": 20904 }, { "epoch": 2.6593308739346138, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6398284435272217, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8736556172370911, "num_tokens": 793905931.0, "step": 20905 }, { "epoch": 2.6594580842132043, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8191630840301514, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8759480714797974, "num_tokens": 793939711.0, "step": 20906 }, { "epoch": 2.659585294491795, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.635141134262085, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8816307187080383, "num_tokens": 793976746.0, "step": 20907 }, { "epoch": 2.6597125047703853, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.825967788696289, "learning_rate": 1e-05, "loss": 0.4763, "mean_token_accuracy": 0.8623622059822083, "num_tokens": 794014132.0, "step": 20908 }, { "epoch": 2.659839715048976, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6483139991760254, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.864712119102478, "num_tokens": 794051958.0, "step": 20909 }, { "epoch": 2.6599669253275664, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8057498931884766, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.882887601852417, "num_tokens": 794087494.0, "step": 20910 }, { "epoch": 2.660094135606157, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6559994220733643, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8666476011276245, "num_tokens": 794123156.0, "step": 20911 }, { "epoch": 2.6602213458847475, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.77600359916687, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8672672510147095, "num_tokens": 794169023.0, "step": 20912 }, { "epoch": 2.660348556163338, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6531715393066406, "learning_rate": 1e-05, "loss": 0.3809, "mean_token_accuracy": 0.8911111354827881, "num_tokens": 794202145.0, "step": 20913 }, { "epoch": 2.6604757664419285, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7678442001342773, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8931450843811035, "num_tokens": 794239791.0, "step": 20914 }, { "epoch": 2.660602976720519, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.642198085784912, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8860304355621338, "num_tokens": 794285184.0, "step": 20915 }, { "epoch": 2.6607301869991096, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.771881103515625, "learning_rate": 1e-05, "loss": 0.3861, "mean_token_accuracy": 0.8937068581581116, "num_tokens": 794324705.0, "step": 20916 }, { "epoch": 2.6608573972777, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6606433391571045, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8739001154899597, "num_tokens": 794360712.0, "step": 20917 }, { "epoch": 2.6609846075562906, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.754709243774414, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8847254514694214, "num_tokens": 794401634.0, "step": 20918 }, { "epoch": 2.661111817834881, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6727964878082275, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8828935623168945, "num_tokens": 794433460.0, "step": 20919 }, { "epoch": 2.6612390281134717, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.750188112258911, "learning_rate": 1e-05, "loss": 0.3696, "mean_token_accuracy": 0.8951804041862488, "num_tokens": 794468258.0, "step": 20920 }, { "epoch": 2.661366238392062, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.676823854446411, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8763425350189209, "num_tokens": 794505493.0, "step": 20921 }, { "epoch": 2.6614934486706527, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7359676361083984, "learning_rate": 1e-05, "loss": 0.3508, "mean_token_accuracy": 0.9031678438186646, "num_tokens": 794539370.0, "step": 20922 }, { "epoch": 2.661620658949243, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.662696361541748, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8769631385803223, "num_tokens": 794580820.0, "step": 20923 }, { "epoch": 2.661747869227834, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7336251735687256, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8873997926712036, "num_tokens": 794617503.0, "step": 20924 }, { "epoch": 2.661875079506424, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.676692247390747, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8799242973327637, "num_tokens": 794653701.0, "step": 20925 }, { "epoch": 2.662002289785015, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.734712839126587, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8770761489868164, "num_tokens": 794689938.0, "step": 20926 }, { "epoch": 2.662129500063605, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6750664710998535, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8703879117965698, "num_tokens": 794730460.0, "step": 20927 }, { "epoch": 2.662256710342196, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.723630428314209, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8864437937736511, "num_tokens": 794770686.0, "step": 20928 }, { "epoch": 2.662383920620786, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6776182651519775, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8666544556617737, "num_tokens": 794814396.0, "step": 20929 }, { "epoch": 2.6625111308993765, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.712689161300659, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8833490610122681, "num_tokens": 794857425.0, "step": 20930 }, { "epoch": 2.662638341177967, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.680663824081421, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8854732513427734, "num_tokens": 794894115.0, "step": 20931 }, { "epoch": 2.6627655514565576, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.718134641647339, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8765279054641724, "num_tokens": 794929423.0, "step": 20932 }, { "epoch": 2.662892761735148, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6790480613708496, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8802663087844849, "num_tokens": 794966019.0, "step": 20933 }, { "epoch": 2.6630199720137386, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7123215198516846, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8768656253814697, "num_tokens": 795006501.0, "step": 20934 }, { "epoch": 2.663147182292329, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6710190773010254, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8806336522102356, "num_tokens": 795048068.0, "step": 20935 }, { "epoch": 2.6632743925709197, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7300808429718018, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8678390979766846, "num_tokens": 795084918.0, "step": 20936 }, { "epoch": 2.66340160284951, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.665884256362915, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.8848778009414673, "num_tokens": 795119261.0, "step": 20937 }, { "epoch": 2.6635288131281007, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7052812576293945, "learning_rate": 1e-05, "loss": 0.3945, "mean_token_accuracy": 0.8879622220993042, "num_tokens": 795162559.0, "step": 20938 }, { "epoch": 2.6636560234066913, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6688947677612305, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8847231864929199, "num_tokens": 795200853.0, "step": 20939 }, { "epoch": 2.663783233685282, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.722352981567383, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8877527713775635, "num_tokens": 795238405.0, "step": 20940 }, { "epoch": 2.6639104439638723, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6732563972473145, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8753517866134644, "num_tokens": 795278341.0, "step": 20941 }, { "epoch": 2.664037654242463, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.716665744781494, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8737674951553345, "num_tokens": 795319108.0, "step": 20942 }, { "epoch": 2.6641648645210534, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6783370971679688, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.878000795841217, "num_tokens": 795354641.0, "step": 20943 }, { "epoch": 2.664292074799644, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7079622745513916, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8650734424591064, "num_tokens": 795399098.0, "step": 20944 }, { "epoch": 2.6644192850782344, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.660383939743042, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8849732875823975, "num_tokens": 795444191.0, "step": 20945 }, { "epoch": 2.664546495356825, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7381751537323, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8789966702461243, "num_tokens": 795476450.0, "step": 20946 }, { "epoch": 2.6646737056354155, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6543259620666504, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8772498369216919, "num_tokens": 795518740.0, "step": 20947 }, { "epoch": 2.6648009159140056, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7423343658447266, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8750167489051819, "num_tokens": 795558944.0, "step": 20948 }, { "epoch": 2.6649281261925966, "ewc_loss": 0.07763671875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.172325134277344e-05, "grad_norm": 2.6611878871917725, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.879534900188446, "num_tokens": 795592685.0, "step": 20949 }, { "epoch": 2.6650553364711866, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7559571266174316, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8736714124679565, "num_tokens": 795627878.0, "step": 20950 }, { "epoch": 2.6651825467497776, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.655480146408081, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8680866956710815, "num_tokens": 795666784.0, "step": 20951 }, { "epoch": 2.6653097570283677, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7774603366851807, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8807940483093262, "num_tokens": 795701921.0, "step": 20952 }, { "epoch": 2.6654369673069587, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6297340393066406, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8731064200401306, "num_tokens": 795744497.0, "step": 20953 }, { "epoch": 2.6655641775855488, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.769808053970337, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.866196870803833, "num_tokens": 795786871.0, "step": 20954 }, { "epoch": 2.6656913878641393, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.624192953109741, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8884962201118469, "num_tokens": 795829403.0, "step": 20955 }, { "epoch": 2.66581859814273, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7942371368408203, "learning_rate": 1e-05, "loss": 0.4779, "mean_token_accuracy": 0.8638479709625244, "num_tokens": 795867352.0, "step": 20956 }, { "epoch": 2.6659458084213203, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.634610176086426, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.8717361688613892, "num_tokens": 795907209.0, "step": 20957 }, { "epoch": 2.666073018699911, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7983901500701904, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8756667375564575, "num_tokens": 795946515.0, "step": 20958 }, { "epoch": 2.6662002289785014, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6302332878112793, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8877071142196655, "num_tokens": 795983328.0, "step": 20959 }, { "epoch": 2.666327439257092, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.809579372406006, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8625348806381226, "num_tokens": 796018609.0, "step": 20960 }, { "epoch": 2.6664546495356825, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6262130737304688, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8724705576896667, "num_tokens": 796059676.0, "step": 20961 }, { "epoch": 2.666581859814273, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8093860149383545, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8758939504623413, "num_tokens": 796100896.0, "step": 20962 }, { "epoch": 2.6667090700928635, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.613555908203125, "learning_rate": 1e-05, "loss": 0.3807, "mean_token_accuracy": 0.8940256834030151, "num_tokens": 796140513.0, "step": 20963 }, { "epoch": 2.666836280371454, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.82305645942688, "learning_rate": 1e-05, "loss": 0.3922, "mean_token_accuracy": 0.8903295397758484, "num_tokens": 796174339.0, "step": 20964 }, { "epoch": 2.6669634906500446, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.61208438873291, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8821885585784912, "num_tokens": 796216624.0, "step": 20965 }, { "epoch": 2.667090700928635, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.847761869430542, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8664466142654419, "num_tokens": 796250681.0, "step": 20966 }, { "epoch": 2.6672179112072256, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6324801445007324, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8813379406929016, "num_tokens": 796284860.0, "step": 20967 }, { "epoch": 2.667345121485816, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.824394702911377, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.8919960260391235, "num_tokens": 796325833.0, "step": 20968 }, { "epoch": 2.6674723317644067, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6540637016296387, "learning_rate": 1e-05, "loss": 0.3907, "mean_token_accuracy": 0.8882980942726135, "num_tokens": 796354955.0, "step": 20969 }, { "epoch": 2.667599542042997, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.823655366897583, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8784003853797913, "num_tokens": 796390505.0, "step": 20970 }, { "epoch": 2.6677267523215877, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6423087120056152, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.8873493671417236, "num_tokens": 796428003.0, "step": 20971 }, { "epoch": 2.6678539626001783, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8106796741485596, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.8881552815437317, "num_tokens": 796467396.0, "step": 20972 }, { "epoch": 2.6679811728787683, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.647343873977661, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.876379668712616, "num_tokens": 796505320.0, "step": 20973 }, { "epoch": 2.6681083831573593, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8058154582977295, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8883452415466309, "num_tokens": 796542109.0, "step": 20974 }, { "epoch": 2.6682355934359494, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6269431114196777, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8772584199905396, "num_tokens": 796588523.0, "step": 20975 }, { "epoch": 2.6683628037145404, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.82712984085083, "learning_rate": 1e-05, "loss": 0.4827, "mean_token_accuracy": 0.8631401658058167, "num_tokens": 796624493.0, "step": 20976 }, { "epoch": 2.6684900139931305, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.627373218536377, "learning_rate": 1e-05, "loss": 0.3883, "mean_token_accuracy": 0.8916630148887634, "num_tokens": 796662147.0, "step": 20977 }, { "epoch": 2.668617224271721, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8244681358337402, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8766093254089355, "num_tokens": 796700024.0, "step": 20978 }, { "epoch": 2.6687444345503115, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6383235454559326, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8845149278640747, "num_tokens": 796735058.0, "step": 20979 }, { "epoch": 2.668871644828902, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.819512367248535, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8715019226074219, "num_tokens": 796778419.0, "step": 20980 }, { "epoch": 2.6689988551074926, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6278696060180664, "learning_rate": 1e-05, "loss": 0.3994, "mean_token_accuracy": 0.8853992223739624, "num_tokens": 796817495.0, "step": 20981 }, { "epoch": 2.669126065386083, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8402321338653564, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8768638968467712, "num_tokens": 796850432.0, "step": 20982 }, { "epoch": 2.6692532756646736, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6370410919189453, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8741213083267212, "num_tokens": 796889584.0, "step": 20983 }, { "epoch": 2.669380485943264, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.8245768547058105, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8796178698539734, "num_tokens": 796924052.0, "step": 20984 }, { "epoch": 2.6695076962218547, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6419129371643066, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8701438307762146, "num_tokens": 796963795.0, "step": 20985 }, { "epoch": 2.669634906500445, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8140499591827393, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8884921073913574, "num_tokens": 797002621.0, "step": 20986 }, { "epoch": 2.6697621167790357, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.64462947845459, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8572198748588562, "num_tokens": 797047173.0, "step": 20987 }, { "epoch": 2.6698893270576263, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8045711517333984, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.8782590627670288, "num_tokens": 797088394.0, "step": 20988 }, { "epoch": 2.670016537336217, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.644300699234009, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8868578672409058, "num_tokens": 797126108.0, "step": 20989 }, { "epoch": 2.6701437476148073, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8051645755767822, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8851738572120667, "num_tokens": 797162718.0, "step": 20990 }, { "epoch": 2.670270957893398, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6606545448303223, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8692658543586731, "num_tokens": 797197552.0, "step": 20991 }, { "epoch": 2.6703981681719884, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8191211223602295, "learning_rate": 1e-05, "loss": 0.4855, "mean_token_accuracy": 0.8581311702728271, "num_tokens": 797232523.0, "step": 20992 }, { "epoch": 2.670525378450579, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.65464186668396, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8827020525932312, "num_tokens": 797266620.0, "step": 20993 }, { "epoch": 2.6706525887291694, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7863266468048096, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8857836127281189, "num_tokens": 797305378.0, "step": 20994 }, { "epoch": 2.67077979900776, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.648686647415161, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8848135471343994, "num_tokens": 797343269.0, "step": 20995 }, { "epoch": 2.6709070092863505, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7913975715637207, "learning_rate": 1e-05, "loss": 0.3496, "mean_token_accuracy": 0.9038357138633728, "num_tokens": 797375034.0, "step": 20996 }, { "epoch": 2.671034219564941, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6574289798736572, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8746819496154785, "num_tokens": 797410869.0, "step": 20997 }, { "epoch": 2.671161429843531, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7868692874908447, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8889906406402588, "num_tokens": 797449189.0, "step": 20998 }, { "epoch": 2.671288640122122, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.646144151687622, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8845332264900208, "num_tokens": 797492182.0, "step": 20999 }, { "epoch": 2.671415850400712, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.778606653213501, "learning_rate": 1e-05, "loss": 0.367, "mean_token_accuracy": 0.8965718746185303, "num_tokens": 797529043.0, "step": 21000 }, { "epoch": 2.671543060679303, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6763532161712646, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8732171058654785, "num_tokens": 797560630.0, "step": 21001 }, { "epoch": 2.6716702709578932, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7717244625091553, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8733263611793518, "num_tokens": 797598667.0, "step": 21002 }, { "epoch": 2.6717974812364838, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6526730060577393, "learning_rate": 1e-05, "loss": 0.3946, "mean_token_accuracy": 0.8893594741821289, "num_tokens": 797637420.0, "step": 21003 }, { "epoch": 2.6719246915150743, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7827813625335693, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8867088556289673, "num_tokens": 797670969.0, "step": 21004 }, { "epoch": 2.672051901793665, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.641791820526123, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8811354637145996, "num_tokens": 797711915.0, "step": 21005 }, { "epoch": 2.6721791120722553, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.781595230102539, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.878894567489624, "num_tokens": 797751036.0, "step": 21006 }, { "epoch": 2.672306322350846, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.639364719390869, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8729088306427002, "num_tokens": 797793999.0, "step": 21007 }, { "epoch": 2.6724335326294364, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.776461601257324, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.877698540687561, "num_tokens": 797834836.0, "step": 21008 }, { "epoch": 2.672560742908027, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6563971042633057, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8689005970954895, "num_tokens": 797872855.0, "step": 21009 }, { "epoch": 2.6726879531866174, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.779390811920166, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8757162690162659, "num_tokens": 797910925.0, "step": 21010 }, { "epoch": 2.672815163465208, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6469664573669434, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8719079494476318, "num_tokens": 797952201.0, "step": 21011 }, { "epoch": 2.6729423737437985, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7958216667175293, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.8602050542831421, "num_tokens": 797988706.0, "step": 21012 }, { "epoch": 2.673069584022389, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6469757556915283, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8732326030731201, "num_tokens": 798028438.0, "step": 21013 }, { "epoch": 2.6731967943009796, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7852771282196045, "learning_rate": 1e-05, "loss": 0.3927, "mean_token_accuracy": 0.8896054029464722, "num_tokens": 798062864.0, "step": 21014 }, { "epoch": 2.67332400457957, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6459853649139404, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8842187523841858, "num_tokens": 798098271.0, "step": 21015 }, { "epoch": 2.6734512148581606, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.783043146133423, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8850799798965454, "num_tokens": 798131286.0, "step": 21016 }, { "epoch": 2.673578425136751, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.654053211212158, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8797296285629272, "num_tokens": 798164711.0, "step": 21017 }, { "epoch": 2.6737056354153417, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.77583646774292, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.879776120185852, "num_tokens": 798205280.0, "step": 21018 }, { "epoch": 2.673832845693932, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6463589668273926, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.8809686303138733, "num_tokens": 798246561.0, "step": 21019 }, { "epoch": 2.6739600559725227, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.774268627166748, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.881074070930481, "num_tokens": 798287494.0, "step": 21020 }, { "epoch": 2.674087266251113, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.650563955307007, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8811991214752197, "num_tokens": 798325305.0, "step": 21021 }, { "epoch": 2.674214476529704, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.771047592163086, "learning_rate": 1e-05, "loss": 0.3891, "mean_token_accuracy": 0.8893394470214844, "num_tokens": 798363083.0, "step": 21022 }, { "epoch": 2.674341686808294, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6579501628875732, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8605906963348389, "num_tokens": 798403324.0, "step": 21023 }, { "epoch": 2.674468897086885, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7640795707702637, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8859628438949585, "num_tokens": 798441548.0, "step": 21024 }, { "epoch": 2.674596107365475, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6380128860473633, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8743996024131775, "num_tokens": 798486038.0, "step": 21025 }, { "epoch": 2.674723317644066, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7566051483154297, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8878904581069946, "num_tokens": 798530657.0, "step": 21026 }, { "epoch": 2.674850527922656, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.642359733581543, "learning_rate": 1e-05, "loss": 0.4131, "mean_token_accuracy": 0.8811163902282715, "num_tokens": 798571027.0, "step": 21027 }, { "epoch": 2.6749777382012465, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7589168548583984, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8876211047172546, "num_tokens": 798609874.0, "step": 21028 }, { "epoch": 2.675104948479837, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6523854732513428, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8713804483413696, "num_tokens": 798645191.0, "step": 21029 }, { "epoch": 2.6752321587584276, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.747157335281372, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.8866914510726929, "num_tokens": 798688445.0, "step": 21030 }, { "epoch": 2.675359369037018, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6464457511901855, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8817793130874634, "num_tokens": 798729582.0, "step": 21031 }, { "epoch": 2.6754865793156086, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.761370897293091, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8869817852973938, "num_tokens": 798765178.0, "step": 21032 }, { "epoch": 2.675613789594199, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.638990640640259, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8794872760772705, "num_tokens": 798806048.0, "step": 21033 }, { "epoch": 2.6757409998727897, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7645885944366455, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8854472637176514, "num_tokens": 798845554.0, "step": 21034 }, { "epoch": 2.67586821015138, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6366963386535645, "learning_rate": 1e-05, "loss": 0.3738, "mean_token_accuracy": 0.8960138559341431, "num_tokens": 798883124.0, "step": 21035 }, { "epoch": 2.6759954204299707, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7735583782196045, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.8922088146209717, "num_tokens": 798916040.0, "step": 21036 }, { "epoch": 2.6761226307085613, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6407861709594727, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8782510757446289, "num_tokens": 798952562.0, "step": 21037 }, { "epoch": 2.676249840987152, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7800869941711426, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8742082118988037, "num_tokens": 798990689.0, "step": 21038 }, { "epoch": 2.6763770512657423, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6334919929504395, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8846117258071899, "num_tokens": 799029398.0, "step": 21039 }, { "epoch": 2.676504261544333, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7741048336029053, "learning_rate": 1e-05, "loss": 0.3859, "mean_token_accuracy": 0.8930021524429321, "num_tokens": 799073200.0, "step": 21040 }, { "epoch": 2.6766314718229234, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6225154399871826, "learning_rate": 1e-05, "loss": 0.3961, "mean_token_accuracy": 0.8868415355682373, "num_tokens": 799115020.0, "step": 21041 }, { "epoch": 2.676758682101514, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7957077026367188, "learning_rate": 1e-05, "loss": 0.3874, "mean_token_accuracy": 0.891174852848053, "num_tokens": 799152348.0, "step": 21042 }, { "epoch": 2.6768858923801044, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6353256702423096, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8767226934432983, "num_tokens": 799191120.0, "step": 21043 }, { "epoch": 2.677013102658695, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7976691722869873, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8807756900787354, "num_tokens": 799229044.0, "step": 21044 }, { "epoch": 2.6771403129372855, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6267244815826416, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8868770003318787, "num_tokens": 799269148.0, "step": 21045 }, { "epoch": 2.6772675232158756, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.804407835006714, "learning_rate": 1e-05, "loss": 0.3887, "mean_token_accuracy": 0.8883494138717651, "num_tokens": 799305188.0, "step": 21046 }, { "epoch": 2.6773947334944665, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.624929666519165, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8723114132881165, "num_tokens": 799344156.0, "step": 21047 }, { "epoch": 2.6775219437730566, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.837851047515869, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8682933449745178, "num_tokens": 799374792.0, "step": 21048 }, { "epoch": 2.6776491540516476, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6318061351776123, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8809123635292053, "num_tokens": 799408425.0, "step": 21049 }, { "epoch": 2.6777763643302377, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.812178373336792, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8844977617263794, "num_tokens": 799448362.0, "step": 21050 }, { "epoch": 2.6779035746088287, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6306915283203125, "learning_rate": 1e-05, "loss": 0.3975, "mean_token_accuracy": 0.8857940435409546, "num_tokens": 799482225.0, "step": 21051 }, { "epoch": 2.6780307848874187, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.816965103149414, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.865912914276123, "num_tokens": 799523053.0, "step": 21052 }, { "epoch": 2.6781579951660093, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6310670375823975, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8804664611816406, "num_tokens": 799563070.0, "step": 21053 }, { "epoch": 2.6782852054446, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8285505771636963, "learning_rate": 1e-05, "loss": 0.3942, "mean_token_accuracy": 0.8880561590194702, "num_tokens": 799595465.0, "step": 21054 }, { "epoch": 2.6784124157231903, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.629209041595459, "learning_rate": 1e-05, "loss": 0.3934, "mean_token_accuracy": 0.8887453079223633, "num_tokens": 799632308.0, "step": 21055 }, { "epoch": 2.678539626001781, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8196334838867188, "learning_rate": 1e-05, "loss": 0.385, "mean_token_accuracy": 0.893257200717926, "num_tokens": 799669563.0, "step": 21056 }, { "epoch": 2.6786668362803714, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.630376100540161, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8776427507400513, "num_tokens": 799714575.0, "step": 21057 }, { "epoch": 2.678794046558962, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.832200050354004, "learning_rate": 1e-05, "loss": 0.4214, "mean_token_accuracy": 0.8833357095718384, "num_tokens": 799752730.0, "step": 21058 }, { "epoch": 2.6789212568375524, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6435494422912598, "learning_rate": 1e-05, "loss": 0.4555, "mean_token_accuracy": 0.8683449029922485, "num_tokens": 799788110.0, "step": 21059 }, { "epoch": 2.679048467116143, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8198697566986084, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8886291980743408, "num_tokens": 799825561.0, "step": 21060 }, { "epoch": 2.6791756773947335, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6333839893341064, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.8875616788864136, "num_tokens": 799865412.0, "step": 21061 }, { "epoch": 2.679302887673324, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8395495414733887, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8686679601669312, "num_tokens": 799900238.0, "step": 21062 }, { "epoch": 2.6794300979519146, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6465468406677246, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8742472529411316, "num_tokens": 799936724.0, "step": 21063 }, { "epoch": 2.679557308230505, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8125641345977783, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8759971857070923, "num_tokens": 799978347.0, "step": 21064 }, { "epoch": 2.6796845185090956, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6496469974517822, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8746909499168396, "num_tokens": 800015419.0, "step": 21065 }, { "epoch": 2.679811728787686, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8066189289093018, "learning_rate": 1e-05, "loss": 0.3658, "mean_token_accuracy": 0.9000828266143799, "num_tokens": 800048435.0, "step": 21066 }, { "epoch": 2.6799389390662767, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6538939476013184, "learning_rate": 1e-05, "loss": 0.4121, "mean_token_accuracy": 0.8815197944641113, "num_tokens": 800083317.0, "step": 21067 }, { "epoch": 2.680066149344867, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7946693897247314, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8770729303359985, "num_tokens": 800123303.0, "step": 21068 }, { "epoch": 2.6801933596234577, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6367902755737305, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8831398487091064, "num_tokens": 800165157.0, "step": 21069 }, { "epoch": 2.6803205699020483, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8002686500549316, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.874427855014801, "num_tokens": 800203657.0, "step": 21070 }, { "epoch": 2.6804477801806383, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6395883560180664, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8852807283401489, "num_tokens": 800245008.0, "step": 21071 }, { "epoch": 2.6805749904592293, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8048229217529297, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8808374404907227, "num_tokens": 800285438.0, "step": 21072 }, { "epoch": 2.6807022007378194, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.656728506088257, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8756129741668701, "num_tokens": 800321345.0, "step": 21073 }, { "epoch": 2.6808294110164104, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.790944814682007, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.872597873210907, "num_tokens": 800361657.0, "step": 21074 }, { "epoch": 2.6809566212950005, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6730921268463135, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8608601093292236, "num_tokens": 800393809.0, "step": 21075 }, { "epoch": 2.681083831573591, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7774758338928223, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8813788890838623, "num_tokens": 800432541.0, "step": 21076 }, { "epoch": 2.6812110418521815, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.66435170173645, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8777275681495667, "num_tokens": 800468552.0, "step": 21077 }, { "epoch": 2.681338252130772, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7771177291870117, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8859567642211914, "num_tokens": 800503911.0, "step": 21078 }, { "epoch": 2.6814654624093626, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.639716625213623, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8792188167572021, "num_tokens": 800549755.0, "step": 21079 }, { "epoch": 2.681592672687953, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7789571285247803, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.8867011666297913, "num_tokens": 800583240.0, "step": 21080 }, { "epoch": 2.6817198829665436, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6556191444396973, "learning_rate": 1e-05, "loss": 0.4771, "mean_token_accuracy": 0.8639228343963623, "num_tokens": 800624016.0, "step": 21081 }, { "epoch": 2.681847093245134, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.782322645187378, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8635408878326416, "num_tokens": 800661052.0, "step": 21082 }, { "epoch": 2.6819743035237247, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.647019624710083, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8826249241828918, "num_tokens": 800701549.0, "step": 21083 }, { "epoch": 2.682101513802315, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7725391387939453, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.876388669013977, "num_tokens": 800738650.0, "step": 21084 }, { "epoch": 2.6822287240809057, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.649705410003662, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8828117847442627, "num_tokens": 800773687.0, "step": 21085 }, { "epoch": 2.6823559343594963, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7699835300445557, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.886913001537323, "num_tokens": 800812523.0, "step": 21086 }, { "epoch": 2.682483144638087, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.656780958175659, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8807710409164429, "num_tokens": 800850838.0, "step": 21087 }, { "epoch": 2.6826103549166773, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7720842361450195, "learning_rate": 1e-05, "loss": 0.3864, "mean_token_accuracy": 0.8886740803718567, "num_tokens": 800886734.0, "step": 21088 }, { "epoch": 2.682737565195268, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.648473024368286, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8754603266716003, "num_tokens": 800930816.0, "step": 21089 }, { "epoch": 2.6828647754738584, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.791130781173706, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.878571629524231, "num_tokens": 800962982.0, "step": 21090 }, { "epoch": 2.682991985752449, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.648857593536377, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8677095174789429, "num_tokens": 801003608.0, "step": 21091 }, { "epoch": 2.6831191960310394, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8101823329925537, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8760479688644409, "num_tokens": 801032889.0, "step": 21092 }, { "epoch": 2.68324640630963, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6543214321136475, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8704365491867065, "num_tokens": 801068717.0, "step": 21093 }, { "epoch": 2.6833736165882205, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8002190589904785, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8793681859970093, "num_tokens": 801105120.0, "step": 21094 }, { "epoch": 2.683500826866811, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6574151515960693, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8739058971405029, "num_tokens": 801140301.0, "step": 21095 }, { "epoch": 2.683628037145401, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7881851196289062, "learning_rate": 1e-05, "loss": 0.3602, "mean_token_accuracy": 0.8997299075126648, "num_tokens": 801176314.0, "step": 21096 }, { "epoch": 2.683755247423992, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6466124057769775, "learning_rate": 1e-05, "loss": 0.4181, "mean_token_accuracy": 0.8842319250106812, "num_tokens": 801212906.0, "step": 21097 }, { "epoch": 2.683882457702582, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8155243396759033, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8686493635177612, "num_tokens": 801243084.0, "step": 21098 }, { "epoch": 2.684009667981173, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.640932559967041, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8782473802566528, "num_tokens": 801282701.0, "step": 21099 }, { "epoch": 2.684136878259763, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8160593509674072, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8872232437133789, "num_tokens": 801315363.0, "step": 21100 }, { "epoch": 2.6842640885383537, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6361260414123535, "learning_rate": 1e-05, "loss": 0.4401, "mean_token_accuracy": 0.8761494755744934, "num_tokens": 801354103.0, "step": 21101 }, { "epoch": 2.6843912988169443, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8130528926849365, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8784549832344055, "num_tokens": 801393351.0, "step": 21102 }, { "epoch": 2.684518509095535, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.632758855819702, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8750597834587097, "num_tokens": 801432829.0, "step": 21103 }, { "epoch": 2.6846457193741253, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.822068929672241, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.876742959022522, "num_tokens": 801468735.0, "step": 21104 }, { "epoch": 2.684772929652716, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.628505229949951, "learning_rate": 1e-05, "loss": 0.3874, "mean_token_accuracy": 0.8903666138648987, "num_tokens": 801507238.0, "step": 21105 }, { "epoch": 2.6849001399313064, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8219821453094482, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.886396050453186, "num_tokens": 801543040.0, "step": 21106 }, { "epoch": 2.685027350209897, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.628769874572754, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8795998096466064, "num_tokens": 801584466.0, "step": 21107 }, { "epoch": 2.6851545604884874, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8448967933654785, "learning_rate": 1e-05, "loss": 0.4017, "mean_token_accuracy": 0.8876898288726807, "num_tokens": 801617466.0, "step": 21108 }, { "epoch": 2.685281770767078, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6518566608428955, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8706845045089722, "num_tokens": 801653408.0, "step": 21109 }, { "epoch": 2.6854089810456685, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8350565433502197, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.878504753112793, "num_tokens": 801687200.0, "step": 21110 }, { "epoch": 2.685536191324259, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6449952125549316, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8770954608917236, "num_tokens": 801728547.0, "step": 21111 }, { "epoch": 2.6856634016028496, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8242812156677246, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8729177713394165, "num_tokens": 801765425.0, "step": 21112 }, { "epoch": 2.68579061188144, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.642341136932373, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8820723295211792, "num_tokens": 801800694.0, "step": 21113 }, { "epoch": 2.6859178221600306, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8200716972351074, "learning_rate": 1e-05, "loss": 0.3862, "mean_token_accuracy": 0.8913360834121704, "num_tokens": 801835748.0, "step": 21114 }, { "epoch": 2.686045032438621, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6458587646484375, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8737683296203613, "num_tokens": 801873527.0, "step": 21115 }, { "epoch": 2.6861722427172117, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8205583095550537, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8883602023124695, "num_tokens": 801910630.0, "step": 21116 }, { "epoch": 2.686299452995802, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6441328525543213, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8844605684280396, "num_tokens": 801946136.0, "step": 21117 }, { "epoch": 2.6864266632743927, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.824504852294922, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8771718740463257, "num_tokens": 801982283.0, "step": 21118 }, { "epoch": 2.686553873552983, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.643143892288208, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8756372928619385, "num_tokens": 802022263.0, "step": 21119 }, { "epoch": 2.686681083831574, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8370954990386963, "learning_rate": 1e-05, "loss": 0.5326, "mean_token_accuracy": 0.8419249653816223, "num_tokens": 802061837.0, "step": 21120 }, { "epoch": 2.686808294110164, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6462810039520264, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.8870261311531067, "num_tokens": 802094528.0, "step": 21121 }, { "epoch": 2.686935504388755, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8205554485321045, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8760654926300049, "num_tokens": 802135157.0, "step": 21122 }, { "epoch": 2.687062714667345, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6388132572174072, "learning_rate": 1e-05, "loss": 0.3811, "mean_token_accuracy": 0.8904697895050049, "num_tokens": 802168818.0, "step": 21123 }, { "epoch": 2.687189924945936, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8132174015045166, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8769176006317139, "num_tokens": 802209486.0, "step": 21124 }, { "epoch": 2.687317135224526, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.650858163833618, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8672114610671997, "num_tokens": 802248534.0, "step": 21125 }, { "epoch": 2.6874443455031165, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8059065341949463, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8797365427017212, "num_tokens": 802289330.0, "step": 21126 }, { "epoch": 2.687571555781707, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.657829761505127, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.87970370054245, "num_tokens": 802322065.0, "step": 21127 }, { "epoch": 2.6876987660602976, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7916903495788574, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.876562774181366, "num_tokens": 802361820.0, "step": 21128 }, { "epoch": 2.687825976338888, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6586079597473145, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8815720081329346, "num_tokens": 802397748.0, "step": 21129 }, { "epoch": 2.6879531866174786, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7925329208374023, "learning_rate": 1e-05, "loss": 0.4589, "mean_token_accuracy": 0.8649851083755493, "num_tokens": 802432152.0, "step": 21130 }, { "epoch": 2.688080396896069, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.643850564956665, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8849433064460754, "num_tokens": 802471464.0, "step": 21131 }, { "epoch": 2.6882076071746597, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8062233924865723, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.879705548286438, "num_tokens": 802503096.0, "step": 21132 }, { "epoch": 2.68833481745325, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6758832931518555, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8698102235794067, "num_tokens": 802540036.0, "step": 21133 }, { "epoch": 2.6884620277318407, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.771829128265381, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8791841864585876, "num_tokens": 802580798.0, "step": 21134 }, { "epoch": 2.6885892380104313, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6588447093963623, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.874028205871582, "num_tokens": 802619472.0, "step": 21135 }, { "epoch": 2.688716448289022, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.765857696533203, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.8810228109359741, "num_tokens": 802656834.0, "step": 21136 }, { "epoch": 2.6888436585676123, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.649599552154541, "learning_rate": 1e-05, "loss": 0.3572, "mean_token_accuracy": 0.8990398645401001, "num_tokens": 802690269.0, "step": 21137 }, { "epoch": 2.688970868846203, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7658183574676514, "learning_rate": 1e-05, "loss": 0.3923, "mean_token_accuracy": 0.8907663226127625, "num_tokens": 802726822.0, "step": 21138 }, { "epoch": 2.6890980791247934, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6480181217193604, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.888532817363739, "num_tokens": 802764479.0, "step": 21139 }, { "epoch": 2.689225289403384, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7677178382873535, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8836076855659485, "num_tokens": 802800938.0, "step": 21140 }, { "epoch": 2.6893524996819744, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6465227603912354, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8835687041282654, "num_tokens": 802841593.0, "step": 21141 }, { "epoch": 2.689479709960565, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.77990984916687, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8783440589904785, "num_tokens": 802875847.0, "step": 21142 }, { "epoch": 2.6896069202391555, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6453452110290527, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8846728205680847, "num_tokens": 802915786.0, "step": 21143 }, { "epoch": 2.6897341305177456, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.773066282272339, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8760659694671631, "num_tokens": 802958774.0, "step": 21144 }, { "epoch": 2.6898613407963365, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6376383304595947, "learning_rate": 1e-05, "loss": 0.3843, "mean_token_accuracy": 0.8944496512413025, "num_tokens": 802991360.0, "step": 21145 }, { "epoch": 2.6899885510749266, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.777061700820923, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8876301050186157, "num_tokens": 803030852.0, "step": 21146 }, { "epoch": 2.6901157613535176, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6412549018859863, "learning_rate": 1e-05, "loss": 0.4077, "mean_token_accuracy": 0.8842068910598755, "num_tokens": 803068965.0, "step": 21147 }, { "epoch": 2.6902429716321077, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7866642475128174, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8751667737960815, "num_tokens": 803103331.0, "step": 21148 }, { "epoch": 2.6903701819106987, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6376407146453857, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.8832442760467529, "num_tokens": 803144242.0, "step": 21149 }, { "epoch": 2.6904973921892887, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8119850158691406, "learning_rate": 1e-05, "loss": 0.4699, "mean_token_accuracy": 0.8632422685623169, "num_tokens": 803178355.0, "step": 21150 }, { "epoch": 2.6906246024678793, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6274328231811523, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.877405047416687, "num_tokens": 803218783.0, "step": 21151 }, { "epoch": 2.69075181274647, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8827250003814697, "learning_rate": 1e-05, "loss": 0.4923, "mean_token_accuracy": 0.8598818778991699, "num_tokens": 803256391.0, "step": 21152 }, { "epoch": 2.6908790230250603, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.633927345275879, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.879328191280365, "num_tokens": 803285708.0, "step": 21153 }, { "epoch": 2.691006233303651, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8509459495544434, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8810811042785645, "num_tokens": 803319711.0, "step": 21154 }, { "epoch": 2.6911334435822414, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.640038251876831, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8643261790275574, "num_tokens": 803353916.0, "step": 21155 }, { "epoch": 2.691260653860832, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8366565704345703, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8905593752861023, "num_tokens": 803388579.0, "step": 21156 }, { "epoch": 2.6913878641394224, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6172032356262207, "learning_rate": 1e-05, "loss": 0.3862, "mean_token_accuracy": 0.8896008729934692, "num_tokens": 803427715.0, "step": 21157 }, { "epoch": 2.691515074418013, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.837146520614624, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8697057366371155, "num_tokens": 803470363.0, "step": 21158 }, { "epoch": 2.6916422846966035, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.634432792663574, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8756040334701538, "num_tokens": 803515009.0, "step": 21159 }, { "epoch": 2.691769494975194, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.839897155761719e-05, "grad_norm": 2.839019298553467, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8830576539039612, "num_tokens": 803554773.0, "step": 21160 }, { "epoch": 2.6918967052537845, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6489980220794678, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8852669596672058, "num_tokens": 803590404.0, "step": 21161 }, { "epoch": 2.692023915532375, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.8202552795410156, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.888769805431366, "num_tokens": 803634388.0, "step": 21162 }, { "epoch": 2.6921511258109656, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6503288745880127, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8806959986686707, "num_tokens": 803673164.0, "step": 21163 }, { "epoch": 2.692278336089556, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8279919624328613, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.878867506980896, "num_tokens": 803715271.0, "step": 21164 }, { "epoch": 2.6924055463681467, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6638169288635254, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8857260942459106, "num_tokens": 803748295.0, "step": 21165 }, { "epoch": 2.692532756646737, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8035590648651123, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8705694675445557, "num_tokens": 803787591.0, "step": 21166 }, { "epoch": 2.6926599669253277, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6768813133239746, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8804208636283875, "num_tokens": 803817640.0, "step": 21167 }, { "epoch": 2.6927871772039182, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.802266836166382, "learning_rate": 1e-05, "loss": 0.4485, "mean_token_accuracy": 0.8710809946060181, "num_tokens": 803847932.0, "step": 21168 }, { "epoch": 2.6929143874825083, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6537325382232666, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8712467551231384, "num_tokens": 803884206.0, "step": 21169 }, { "epoch": 2.6930415977610993, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.782785654067993, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8833839893341064, "num_tokens": 803923225.0, "step": 21170 }, { "epoch": 2.6931688080396894, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.672549247741699, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.864668607711792, "num_tokens": 803961880.0, "step": 21171 }, { "epoch": 2.6932960183182804, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7563812732696533, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8828580975532532, "num_tokens": 804004676.0, "step": 21172 }, { "epoch": 2.6934232285968704, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.657721519470215, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8823492527008057, "num_tokens": 804047875.0, "step": 21173 }, { "epoch": 2.693550438875461, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.798011064529419, "learning_rate": 1e-05, "loss": 0.3616, "mean_token_accuracy": 0.8985797166824341, "num_tokens": 804083467.0, "step": 21174 }, { "epoch": 2.6936776491540515, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6440317630767822, "learning_rate": 1e-05, "loss": 0.3938, "mean_token_accuracy": 0.8884083032608032, "num_tokens": 804124653.0, "step": 21175 }, { "epoch": 2.693804859432642, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7861225605010986, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8873995542526245, "num_tokens": 804159081.0, "step": 21176 }, { "epoch": 2.6939320697112326, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6334285736083984, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8698853254318237, "num_tokens": 804200687.0, "step": 21177 }, { "epoch": 2.694059279989823, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7965962886810303, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8769713044166565, "num_tokens": 804237192.0, "step": 21178 }, { "epoch": 2.6941864902684136, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.635380268096924, "learning_rate": 1e-05, "loss": 0.4021, "mean_token_accuracy": 0.8847401142120361, "num_tokens": 804275078.0, "step": 21179 }, { "epoch": 2.694313700547004, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8061695098876953, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8742825984954834, "num_tokens": 804308551.0, "step": 21180 }, { "epoch": 2.6944409108255947, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.624279499053955, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8785755038261414, "num_tokens": 804347530.0, "step": 21181 }, { "epoch": 2.694568121104185, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8137340545654297, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8725869655609131, "num_tokens": 804388663.0, "step": 21182 }, { "epoch": 2.6946953313827757, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.629918336868286, "learning_rate": 1e-05, "loss": 0.4061, "mean_token_accuracy": 0.8850231170654297, "num_tokens": 804423956.0, "step": 21183 }, { "epoch": 2.6948225416613663, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.808194875717163, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8774275183677673, "num_tokens": 804466736.0, "step": 21184 }, { "epoch": 2.694949751939957, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.653722047805786, "learning_rate": 1e-05, "loss": 0.5093, "mean_token_accuracy": 0.8551566004753113, "num_tokens": 804505101.0, "step": 21185 }, { "epoch": 2.6950769622185473, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.782297134399414, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8862337470054626, "num_tokens": 804542520.0, "step": 21186 }, { "epoch": 2.695204172497138, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.656656503677368, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.8873243927955627, "num_tokens": 804577749.0, "step": 21187 }, { "epoch": 2.6953313827757284, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.780958890914917, "learning_rate": 1e-05, "loss": 0.3707, "mean_token_accuracy": 0.8960917592048645, "num_tokens": 804611370.0, "step": 21188 }, { "epoch": 2.695458593054319, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6637327671051025, "learning_rate": 1e-05, "loss": 0.3818, "mean_token_accuracy": 0.8904986381530762, "num_tokens": 804643696.0, "step": 21189 }, { "epoch": 2.6955858033329094, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7694575786590576, "learning_rate": 1e-05, "loss": 0.3756, "mean_token_accuracy": 0.8969746232032776, "num_tokens": 804678343.0, "step": 21190 }, { "epoch": 2.6957130136115, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.656867265701294, "learning_rate": 1e-05, "loss": 0.3762, "mean_token_accuracy": 0.8925855755805969, "num_tokens": 804711303.0, "step": 21191 }, { "epoch": 2.6958402238900905, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.785943031311035, "learning_rate": 1e-05, "loss": 0.4891, "mean_token_accuracy": 0.8567192554473877, "num_tokens": 804749016.0, "step": 21192 }, { "epoch": 2.695967434168681, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.645658493041992, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.872969388961792, "num_tokens": 804789600.0, "step": 21193 }, { "epoch": 2.696094644447271, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7786731719970703, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8853313326835632, "num_tokens": 804829033.0, "step": 21194 }, { "epoch": 2.696221854725862, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6596529483795166, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8703680038452148, "num_tokens": 804865114.0, "step": 21195 }, { "epoch": 2.696349065004452, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.778688669204712, "learning_rate": 1e-05, "loss": 0.4717, "mean_token_accuracy": 0.8653506636619568, "num_tokens": 804907604.0, "step": 21196 }, { "epoch": 2.696476275283043, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.662825107574463, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8724086880683899, "num_tokens": 804942361.0, "step": 21197 }, { "epoch": 2.696603485561633, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7619471549987793, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8803756237030029, "num_tokens": 804980452.0, "step": 21198 }, { "epoch": 2.6967306958402237, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.653533935546875, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8848080039024353, "num_tokens": 805017969.0, "step": 21199 }, { "epoch": 2.6968579061188143, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7590491771698, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8866354823112488, "num_tokens": 805057053.0, "step": 21200 }, { "epoch": 2.696985116397405, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.650418519973755, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8820260167121887, "num_tokens": 805098928.0, "step": 21201 }, { "epoch": 2.6971123266759953, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.760934829711914, "learning_rate": 1e-05, "loss": 0.4182, "mean_token_accuracy": 0.8811463117599487, "num_tokens": 805140729.0, "step": 21202 }, { "epoch": 2.697239536954586, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.662128448486328, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8828418254852295, "num_tokens": 805177150.0, "step": 21203 }, { "epoch": 2.6973667472331764, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7506978511810303, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.880009651184082, "num_tokens": 805221233.0, "step": 21204 }, { "epoch": 2.697493957511767, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.672452211380005, "learning_rate": 1e-05, "loss": 0.3871, "mean_token_accuracy": 0.892611026763916, "num_tokens": 805255075.0, "step": 21205 }, { "epoch": 2.6976211677903574, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7615714073181152, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.8743508458137512, "num_tokens": 805292468.0, "step": 21206 }, { "epoch": 2.697748378068948, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.653627395629883, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8809911012649536, "num_tokens": 805330648.0, "step": 21207 }, { "epoch": 2.6978755883475385, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.757845878601074, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8814252614974976, "num_tokens": 805368137.0, "step": 21208 }, { "epoch": 2.698002798626129, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6740458011627197, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8808145523071289, "num_tokens": 805399898.0, "step": 21209 }, { "epoch": 2.6981300089047195, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7483863830566406, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8869460225105286, "num_tokens": 805439292.0, "step": 21210 }, { "epoch": 2.69825721918331, "ewc_loss": 0.078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6659960746765137, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8660593032836914, "num_tokens": 805479464.0, "step": 21211 }, { "epoch": 2.6983844294619006, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.730877637863159, "learning_rate": 1e-05, "loss": 0.3478, "mean_token_accuracy": 0.9034733772277832, "num_tokens": 805518844.0, "step": 21212 }, { "epoch": 2.698511639740491, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6790518760681152, "learning_rate": 1e-05, "loss": 0.3866, "mean_token_accuracy": 0.8897484540939331, "num_tokens": 805552192.0, "step": 21213 }, { "epoch": 2.6986388500190817, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7393620014190674, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8843725919723511, "num_tokens": 805587343.0, "step": 21214 }, { "epoch": 2.698766060297672, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6683731079101562, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8820462226867676, "num_tokens": 805623256.0, "step": 21215 }, { "epoch": 2.6988932705762627, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7442569732666016, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8796274662017822, "num_tokens": 805664627.0, "step": 21216 }, { "epoch": 2.699020480854853, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6815590858459473, "learning_rate": 1e-05, "loss": 0.4524, "mean_token_accuracy": 0.8727014660835266, "num_tokens": 805698808.0, "step": 21217 }, { "epoch": 2.6991476911334438, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7257516384124756, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8822190761566162, "num_tokens": 805738468.0, "step": 21218 }, { "epoch": 2.699274901412034, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.685110569000244, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8676862716674805, "num_tokens": 805776233.0, "step": 21219 }, { "epoch": 2.699402111690625, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7076950073242188, "learning_rate": 1e-05, "loss": 0.3778, "mean_token_accuracy": 0.8969433307647705, "num_tokens": 805818459.0, "step": 21220 }, { "epoch": 2.699529321969215, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.694685935974121, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8879175186157227, "num_tokens": 805858340.0, "step": 21221 }, { "epoch": 2.699656532247806, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7078778743743896, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8765951991081238, "num_tokens": 805895920.0, "step": 21222 }, { "epoch": 2.699783742526396, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.680964708328247, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8842095136642456, "num_tokens": 805937692.0, "step": 21223 }, { "epoch": 2.6999109528049865, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.721845865249634, "learning_rate": 1e-05, "loss": 0.4404, "mean_token_accuracy": 0.87474125623703, "num_tokens": 805974573.0, "step": 21224 }, { "epoch": 2.700038163083577, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.681400775909424, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8816409111022949, "num_tokens": 806013643.0, "step": 21225 }, { "epoch": 2.7001653733621676, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7196333408355713, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8829163908958435, "num_tokens": 806048027.0, "step": 21226 }, { "epoch": 2.700292583640758, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.688054084777832, "learning_rate": 1e-05, "loss": 0.4464, "mean_token_accuracy": 0.8704379200935364, "num_tokens": 806087269.0, "step": 21227 }, { "epoch": 2.7004197939193486, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.00035858154296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7208962440490723, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.8731847405433655, "num_tokens": 806126924.0, "step": 21228 }, { "epoch": 2.700547004197939, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.684459924697876, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8800133466720581, "num_tokens": 806162510.0, "step": 21229 }, { "epoch": 2.7006742144765297, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.71773624420166, "learning_rate": 1e-05, "loss": 0.3926, "mean_token_accuracy": 0.888565182685852, "num_tokens": 806196519.0, "step": 21230 }, { "epoch": 2.70080142475512, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.690396785736084, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.866335391998291, "num_tokens": 806233995.0, "step": 21231 }, { "epoch": 2.7009286350337107, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.706296682357788, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8821395635604858, "num_tokens": 806268660.0, "step": 21232 }, { "epoch": 2.7010558453123013, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6923093795776367, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.87116938829422, "num_tokens": 806304187.0, "step": 21233 }, { "epoch": 2.701183055590892, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.697495222091675, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8767716884613037, "num_tokens": 806341524.0, "step": 21234 }, { "epoch": 2.7013102658694823, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6892826557159424, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8703160881996155, "num_tokens": 806384983.0, "step": 21235 }, { "epoch": 2.701437476148073, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.689727544784546, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8832627534866333, "num_tokens": 806429923.0, "step": 21236 }, { "epoch": 2.7015646864266634, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6939446926116943, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8873574733734131, "num_tokens": 806467562.0, "step": 21237 }, { "epoch": 2.701691896705254, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6946794986724854, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8789836168289185, "num_tokens": 806508634.0, "step": 21238 }, { "epoch": 2.7018191069838444, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7044670581817627, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8707963228225708, "num_tokens": 806547972.0, "step": 21239 }, { "epoch": 2.701946317262435, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.688225746154785, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8717190623283386, "num_tokens": 806591461.0, "step": 21240 }, { "epoch": 2.7020735275410255, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.708462715148926, "learning_rate": 1e-05, "loss": 0.3825, "mean_token_accuracy": 0.8926083445549011, "num_tokens": 806623276.0, "step": 21241 }, { "epoch": 2.7022007378196156, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.690479040145874, "learning_rate": 1e-05, "loss": 0.4033, "mean_token_accuracy": 0.8843525052070618, "num_tokens": 806657732.0, "step": 21242 }, { "epoch": 2.7023279480982065, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.704336166381836, "learning_rate": 1e-05, "loss": 0.387, "mean_token_accuracy": 0.8882516622543335, "num_tokens": 806695018.0, "step": 21243 }, { "epoch": 2.7024551583767966, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.666874647140503, "learning_rate": 1e-05, "loss": 0.3679, "mean_token_accuracy": 0.8986167311668396, "num_tokens": 806735344.0, "step": 21244 }, { "epoch": 2.7025823686553876, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7097907066345215, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8781505227088928, "num_tokens": 806776854.0, "step": 21245 }, { "epoch": 2.7027095789339777, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6862070560455322, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8864592909812927, "num_tokens": 806811594.0, "step": 21246 }, { "epoch": 2.7028367892125686, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7263853549957275, "learning_rate": 1e-05, "loss": 0.4542, "mean_token_accuracy": 0.8731189370155334, "num_tokens": 806845249.0, "step": 21247 }, { "epoch": 2.7029639994911587, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6866073608398438, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8740344047546387, "num_tokens": 806885110.0, "step": 21248 }, { "epoch": 2.7030912097697493, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7053916454315186, "learning_rate": 1e-05, "loss": 0.4098, "mean_token_accuracy": 0.8804038763046265, "num_tokens": 806924809.0, "step": 21249 }, { "epoch": 2.70321842004834, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6679816246032715, "learning_rate": 1e-05, "loss": 0.3718, "mean_token_accuracy": 0.895867645740509, "num_tokens": 806962386.0, "step": 21250 }, { "epoch": 2.7033456303269303, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.721604108810425, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8763347864151001, "num_tokens": 806998754.0, "step": 21251 }, { "epoch": 2.703472840605521, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6834418773651123, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8812427520751953, "num_tokens": 807036140.0, "step": 21252 }, { "epoch": 2.7036000508841114, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7157838344573975, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.8749097585678101, "num_tokens": 807077144.0, "step": 21253 }, { "epoch": 2.703727261162702, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.679692506790161, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8771008253097534, "num_tokens": 807114928.0, "step": 21254 }, { "epoch": 2.7038544714412924, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7194602489471436, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8837865591049194, "num_tokens": 807151067.0, "step": 21255 }, { "epoch": 2.703981681719883, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6892991065979004, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8715652227401733, "num_tokens": 807185992.0, "step": 21256 }, { "epoch": 2.7041088919984735, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6991140842437744, "learning_rate": 1e-05, "loss": 0.3894, "mean_token_accuracy": 0.8892984390258789, "num_tokens": 807229122.0, "step": 21257 }, { "epoch": 2.704236102277064, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.7004990577697754, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8749850392341614, "num_tokens": 807264118.0, "step": 21258 }, { "epoch": 2.7043633125556545, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.698460102081299, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8843386769294739, "num_tokens": 807298087.0, "step": 21259 }, { "epoch": 2.704490522834245, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6921637058258057, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8819143772125244, "num_tokens": 807338367.0, "step": 21260 }, { "epoch": 2.7046177331128356, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6902782917022705, "learning_rate": 1e-05, "loss": 0.3898, "mean_token_accuracy": 0.8897659182548523, "num_tokens": 807376858.0, "step": 21261 }, { "epoch": 2.704744943391426, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.713717460632324, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8706569671630859, "num_tokens": 807410840.0, "step": 21262 }, { "epoch": 2.7048721536700167, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6831257343292236, "learning_rate": 1e-05, "loss": 0.3705, "mean_token_accuracy": 0.8942084312438965, "num_tokens": 807445128.0, "step": 21263 }, { "epoch": 2.704999363948607, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.720306396484375, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8724640607833862, "num_tokens": 807478884.0, "step": 21264 }, { "epoch": 2.7051265742271977, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6698546409606934, "learning_rate": 1e-05, "loss": 0.3921, "mean_token_accuracy": 0.8887841701507568, "num_tokens": 807517020.0, "step": 21265 }, { "epoch": 2.7052537845057882, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7330989837646484, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8672881722450256, "num_tokens": 807557037.0, "step": 21266 }, { "epoch": 2.7053809947843783, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.649085283279419, "learning_rate": 1e-05, "loss": 0.3812, "mean_token_accuracy": 0.8931636214256287, "num_tokens": 807596876.0, "step": 21267 }, { "epoch": 2.7055082050629693, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.738539457321167, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8724991083145142, "num_tokens": 807638107.0, "step": 21268 }, { "epoch": 2.7056354153415594, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.693920135498047, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8666720986366272, "num_tokens": 807678751.0, "step": 21269 }, { "epoch": 2.7057626256201504, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.752842903137207, "learning_rate": 1e-05, "loss": 0.4865, "mean_token_accuracy": 0.8578046560287476, "num_tokens": 807711447.0, "step": 21270 }, { "epoch": 2.7058898358987404, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6580212116241455, "learning_rate": 1e-05, "loss": 0.4004, "mean_token_accuracy": 0.8842902779579163, "num_tokens": 807747157.0, "step": 21271 }, { "epoch": 2.706017046177331, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.75075101852417, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8738059997558594, "num_tokens": 807782695.0, "step": 21272 }, { "epoch": 2.7061442564559215, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.656507730484009, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8817969560623169, "num_tokens": 807816850.0, "step": 21273 }, { "epoch": 2.706271466734512, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7615323066711426, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8695639371871948, "num_tokens": 807856246.0, "step": 21274 }, { "epoch": 2.7063986770131025, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6548640727996826, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8866140246391296, "num_tokens": 807891417.0, "step": 21275 }, { "epoch": 2.706525887291693, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.763390302658081, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8874529600143433, "num_tokens": 807924831.0, "step": 21276 }, { "epoch": 2.7066530975702836, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6576297283172607, "learning_rate": 1e-05, "loss": 0.3757, "mean_token_accuracy": 0.8956547975540161, "num_tokens": 807955427.0, "step": 21277 }, { "epoch": 2.706780307848874, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7655255794525146, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8801315426826477, "num_tokens": 807990799.0, "step": 21278 }, { "epoch": 2.7069075181274647, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6527671813964844, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.867453932762146, "num_tokens": 808027028.0, "step": 21279 }, { "epoch": 2.707034728406055, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7954893112182617, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8795905113220215, "num_tokens": 808056750.0, "step": 21280 }, { "epoch": 2.7071619386846457, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.63271164894104, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8770290613174438, "num_tokens": 808096372.0, "step": 21281 }, { "epoch": 2.7072891489632362, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7877700328826904, "learning_rate": 1e-05, "loss": 0.4197, "mean_token_accuracy": 0.8794136643409729, "num_tokens": 808136783.0, "step": 21282 }, { "epoch": 2.7074163592418268, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6306591033935547, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8784554600715637, "num_tokens": 808172310.0, "step": 21283 }, { "epoch": 2.7075435695204173, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7946629524230957, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.885296106338501, "num_tokens": 808207745.0, "step": 21284 }, { "epoch": 2.707670779799008, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6509816646575928, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.877278745174408, "num_tokens": 808243135.0, "step": 21285 }, { "epoch": 2.7077979900775984, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7843666076660156, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8783577680587769, "num_tokens": 808280601.0, "step": 21286 }, { "epoch": 2.707925200356189, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.646808624267578, "learning_rate": 1e-05, "loss": 0.3902, "mean_token_accuracy": 0.8868707418441772, "num_tokens": 808315541.0, "step": 21287 }, { "epoch": 2.7080524106347794, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.787191867828369, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8725612759590149, "num_tokens": 808355670.0, "step": 21288 }, { "epoch": 2.70817962091337, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.638706684112549, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8859732151031494, "num_tokens": 808394189.0, "step": 21289 }, { "epoch": 2.7083068311919605, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.785837411880493, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.8904860019683838, "num_tokens": 808432861.0, "step": 21290 }, { "epoch": 2.708434041470551, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.662172317504883, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8762782216072083, "num_tokens": 808462423.0, "step": 21291 }, { "epoch": 2.708561251749141, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7884669303894043, "learning_rate": 1e-05, "loss": 0.3805, "mean_token_accuracy": 0.8947747945785522, "num_tokens": 808494112.0, "step": 21292 }, { "epoch": 2.708688462027732, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.63955020904541, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8839000463485718, "num_tokens": 808538399.0, "step": 21293 }, { "epoch": 2.708815672306322, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7878332138061523, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8806605339050293, "num_tokens": 808575857.0, "step": 21294 }, { "epoch": 2.708942882584913, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6465084552764893, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8754875659942627, "num_tokens": 808615529.0, "step": 21295 }, { "epoch": 2.709070092863503, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.780154228210449, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8926711082458496, "num_tokens": 808653899.0, "step": 21296 }, { "epoch": 2.7091973031420937, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6389412879943848, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8697912693023682, "num_tokens": 808697372.0, "step": 21297 }, { "epoch": 2.7093245134206843, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7970030307769775, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.87629234790802, "num_tokens": 808733854.0, "step": 21298 }, { "epoch": 2.709451723699275, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6266536712646484, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.879157543182373, "num_tokens": 808777026.0, "step": 21299 }, { "epoch": 2.7095789339778653, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.809399127960205, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8915058970451355, "num_tokens": 808814142.0, "step": 21300 }, { "epoch": 2.709706144256456, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6311771869659424, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8725998997688293, "num_tokens": 808853355.0, "step": 21301 }, { "epoch": 2.7098333545350464, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8120851516723633, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8802671432495117, "num_tokens": 808892782.0, "step": 21302 }, { "epoch": 2.709960564813637, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.633664608001709, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8725529313087463, "num_tokens": 808931935.0, "step": 21303 }, { "epoch": 2.7100877750922274, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.810899496078491, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8785414695739746, "num_tokens": 808973525.0, "step": 21304 }, { "epoch": 2.710214985370818, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.654299020767212, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8684861660003662, "num_tokens": 809008198.0, "step": 21305 }, { "epoch": 2.7103421956494085, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.812657594680786, "learning_rate": 1e-05, "loss": 0.3617, "mean_token_accuracy": 0.898352324962616, "num_tokens": 809038267.0, "step": 21306 }, { "epoch": 2.710469405927999, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6537184715270996, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8732088804244995, "num_tokens": 809073380.0, "step": 21307 }, { "epoch": 2.7105966162065895, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7912797927856445, "learning_rate": 1e-05, "loss": 0.3955, "mean_token_accuracy": 0.8886549472808838, "num_tokens": 809114635.0, "step": 21308 }, { "epoch": 2.71072382648518, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6380908489227295, "learning_rate": 1e-05, "loss": 0.3574, "mean_token_accuracy": 0.900643527507782, "num_tokens": 809149441.0, "step": 21309 }, { "epoch": 2.7108510367637706, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7970011234283447, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8860365152359009, "num_tokens": 809190902.0, "step": 21310 }, { "epoch": 2.710978247042361, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6273646354675293, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.8839026093482971, "num_tokens": 809239297.0, "step": 21311 }, { "epoch": 2.7111054573209517, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8126931190490723, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8844130039215088, "num_tokens": 809277304.0, "step": 21312 }, { "epoch": 2.711232667599542, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6548256874084473, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8675333261489868, "num_tokens": 809313975.0, "step": 21313 }, { "epoch": 2.7113598778781327, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7849252223968506, "learning_rate": 1e-05, "loss": 0.3714, "mean_token_accuracy": 0.8959041833877563, "num_tokens": 809354214.0, "step": 21314 }, { "epoch": 2.711487088156723, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.657367706298828, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8854570388793945, "num_tokens": 809388248.0, "step": 21315 }, { "epoch": 2.7116142984353138, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.77388596534729, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8795069456100464, "num_tokens": 809428256.0, "step": 21316 }, { "epoch": 2.711741508713904, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.66058611869812, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8848270773887634, "num_tokens": 809463706.0, "step": 21317 }, { "epoch": 2.711868718992495, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7717342376708984, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8787598609924316, "num_tokens": 809503172.0, "step": 21318 }, { "epoch": 2.711995929271085, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6551027297973633, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8765746355056763, "num_tokens": 809539894.0, "step": 21319 }, { "epoch": 2.712123139549676, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7657511234283447, "learning_rate": 1e-05, "loss": 0.4772, "mean_token_accuracy": 0.8638583421707153, "num_tokens": 809582781.0, "step": 21320 }, { "epoch": 2.712250349828266, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.651456356048584, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8830047845840454, "num_tokens": 809627077.0, "step": 21321 }, { "epoch": 2.7123775601068565, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.777059555053711, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8746887445449829, "num_tokens": 809663063.0, "step": 21322 }, { "epoch": 2.712504770385447, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6568233966827393, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8657281398773193, "num_tokens": 809702386.0, "step": 21323 }, { "epoch": 2.7126319806640375, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7643308639526367, "learning_rate": 1e-05, "loss": 0.431, "mean_token_accuracy": 0.8741481900215149, "num_tokens": 809745927.0, "step": 21324 }, { "epoch": 2.712759190942628, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6610498428344727, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.884998619556427, "num_tokens": 809777796.0, "step": 21325 }, { "epoch": 2.7128864012212186, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7640202045440674, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8873416185379028, "num_tokens": 809813450.0, "step": 21326 }, { "epoch": 2.713013611499809, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6674423217773438, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8669112324714661, "num_tokens": 809850740.0, "step": 21327 }, { "epoch": 2.7131408217783997, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.760570764541626, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8862663507461548, "num_tokens": 809886170.0, "step": 21328 }, { "epoch": 2.71326803205699, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6546742916107178, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8845903873443604, "num_tokens": 809922078.0, "step": 21329 }, { "epoch": 2.7133952423355807, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7554430961608887, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8774192333221436, "num_tokens": 809962495.0, "step": 21330 }, { "epoch": 2.7135224526141712, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.65419602394104, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8765162229537964, "num_tokens": 810002278.0, "step": 21331 }, { "epoch": 2.7136496628927618, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7628612518310547, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8826277256011963, "num_tokens": 810041236.0, "step": 21332 }, { "epoch": 2.7137768731713523, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.656926155090332, "learning_rate": 1e-05, "loss": 0.3769, "mean_token_accuracy": 0.8951790928840637, "num_tokens": 810078649.0, "step": 21333 }, { "epoch": 2.713904083449943, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.764469861984253, "learning_rate": 1e-05, "loss": 0.4596, "mean_token_accuracy": 0.8701998591423035, "num_tokens": 810118246.0, "step": 21334 }, { "epoch": 2.7140312937285334, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.649190902709961, "learning_rate": 1e-05, "loss": 0.3857, "mean_token_accuracy": 0.8925235271453857, "num_tokens": 810153407.0, "step": 21335 }, { "epoch": 2.714158504007124, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7759273052215576, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.879535973072052, "num_tokens": 810188110.0, "step": 21336 }, { "epoch": 2.7142857142857144, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.647024393081665, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.8730278015136719, "num_tokens": 810225290.0, "step": 21337 }, { "epoch": 2.714412924564305, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7733213901519775, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.885291576385498, "num_tokens": 810259801.0, "step": 21338 }, { "epoch": 2.7145401348428955, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6446549892425537, "learning_rate": 1e-05, "loss": 0.3922, "mean_token_accuracy": 0.8880658149719238, "num_tokens": 810296556.0, "step": 21339 }, { "epoch": 2.7146673451214856, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8213558197021484, "learning_rate": 1e-05, "loss": 0.4121, "mean_token_accuracy": 0.8817266821861267, "num_tokens": 810332282.0, "step": 21340 }, { "epoch": 2.7147945554000765, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.631427049636841, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8816636204719543, "num_tokens": 810368017.0, "step": 21341 }, { "epoch": 2.7149217656786666, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7957699298858643, "learning_rate": 1e-05, "loss": 0.4773, "mean_token_accuracy": 0.8628447651863098, "num_tokens": 810412239.0, "step": 21342 }, { "epoch": 2.7150489759572576, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.629535675048828, "learning_rate": 1e-05, "loss": 0.4352, "mean_token_accuracy": 0.873429536819458, "num_tokens": 810452319.0, "step": 21343 }, { "epoch": 2.7151761862358477, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8037309646606445, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8754316568374634, "num_tokens": 810490936.0, "step": 21344 }, { "epoch": 2.7153033965144386, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6279778480529785, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8805492520332336, "num_tokens": 810533565.0, "step": 21345 }, { "epoch": 2.7154306067930287, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8080596923828125, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8769869804382324, "num_tokens": 810574623.0, "step": 21346 }, { "epoch": 2.7155578170716193, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6324658393859863, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8860092163085938, "num_tokens": 810609423.0, "step": 21347 }, { "epoch": 2.71568502735021, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.820500135421753, "learning_rate": 1e-05, "loss": 0.4203, "mean_token_accuracy": 0.8851467967033386, "num_tokens": 810643814.0, "step": 21348 }, { "epoch": 2.7158122376288003, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.63026762008667, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8790137767791748, "num_tokens": 810682393.0, "step": 21349 }, { "epoch": 2.715939447907391, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.81748628616333, "learning_rate": 1e-05, "loss": 0.3853, "mean_token_accuracy": 0.8914531469345093, "num_tokens": 810718604.0, "step": 21350 }, { "epoch": 2.7160666581859814, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.650463819503784, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8850091695785522, "num_tokens": 810751259.0, "step": 21351 }, { "epoch": 2.716193868464572, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8376529216766357, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8801611661911011, "num_tokens": 810781741.0, "step": 21352 }, { "epoch": 2.7163210787431624, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6454098224639893, "learning_rate": 1e-05, "loss": 0.4972, "mean_token_accuracy": 0.853866457939148, "num_tokens": 810820258.0, "step": 21353 }, { "epoch": 2.716448289021753, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8236725330352783, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8871393799781799, "num_tokens": 810861341.0, "step": 21354 }, { "epoch": 2.7165754993003435, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6365835666656494, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8744597434997559, "num_tokens": 810897963.0, "step": 21355 }, { "epoch": 2.716702709578934, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.821131706237793, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8824514746665955, "num_tokens": 810938042.0, "step": 21356 }, { "epoch": 2.7168299198575245, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6455514430999756, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8810603618621826, "num_tokens": 810969905.0, "step": 21357 }, { "epoch": 2.716957130136115, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8146843910217285, "learning_rate": 1e-05, "loss": 0.3883, "mean_token_accuracy": 0.8872570395469666, "num_tokens": 811006818.0, "step": 21358 }, { "epoch": 2.7170843404147056, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.660595178604126, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8801935315132141, "num_tokens": 811045867.0, "step": 21359 }, { "epoch": 2.717211550693296, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8009729385375977, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8849117755889893, "num_tokens": 811084207.0, "step": 21360 }, { "epoch": 2.7173387609718866, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6702237129211426, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8689028024673462, "num_tokens": 811123093.0, "step": 21361 }, { "epoch": 2.717465971250477, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7870612144470215, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8829401731491089, "num_tokens": 811157951.0, "step": 21362 }, { "epoch": 2.7175931815290677, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.660564661026001, "learning_rate": 1e-05, "loss": 0.361, "mean_token_accuracy": 0.899279773235321, "num_tokens": 811192743.0, "step": 21363 }, { "epoch": 2.7177203918076582, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7787160873413086, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8814641237258911, "num_tokens": 811230042.0, "step": 21364 }, { "epoch": 2.7178476020862483, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.665480852127075, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8786667585372925, "num_tokens": 811270175.0, "step": 21365 }, { "epoch": 2.7179748123648393, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.785146713256836, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8840412497520447, "num_tokens": 811302926.0, "step": 21366 }, { "epoch": 2.7181020226434294, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6602935791015625, "learning_rate": 1e-05, "loss": 0.3578, "mean_token_accuracy": 0.9005729556083679, "num_tokens": 811334503.0, "step": 21367 }, { "epoch": 2.7182292329220203, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7692673206329346, "learning_rate": 1e-05, "loss": 0.4153, "mean_token_accuracy": 0.882319450378418, "num_tokens": 811371480.0, "step": 21368 }, { "epoch": 2.7183564432006104, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.661486864089966, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.879408597946167, "num_tokens": 811410044.0, "step": 21369 }, { "epoch": 2.718483653479201, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.791124105453491, "learning_rate": 1e-05, "loss": 0.4984, "mean_token_accuracy": 0.8532465100288391, "num_tokens": 811447815.0, "step": 21370 }, { "epoch": 2.7186108637577915, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.65623140335083, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8715575933456421, "num_tokens": 811484530.0, "step": 21371 }, { "epoch": 2.718738074036382, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7685818672180176, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8844533562660217, "num_tokens": 811523010.0, "step": 21372 }, { "epoch": 2.7188652843149725, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.663455009460449, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8820035457611084, "num_tokens": 811559031.0, "step": 21373 }, { "epoch": 2.718992494593563, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.767148733139038, "learning_rate": 1e-05, "loss": 0.4764, "mean_token_accuracy": 0.8616271615028381, "num_tokens": 811596501.0, "step": 21374 }, { "epoch": 2.7191197048721536, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6676738262176514, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8604130744934082, "num_tokens": 811634619.0, "step": 21375 }, { "epoch": 2.719246915150744, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7522902488708496, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8671971559524536, "num_tokens": 811678142.0, "step": 21376 }, { "epoch": 2.7193741254293347, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.688108205795288, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.874423623085022, "num_tokens": 811711072.0, "step": 21377 }, { "epoch": 2.719501335707925, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.740144968032837, "learning_rate": 1e-05, "loss": 0.3775, "mean_token_accuracy": 0.8950197696685791, "num_tokens": 811746372.0, "step": 21378 }, { "epoch": 2.7196285459865157, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.677150011062622, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8664706945419312, "num_tokens": 811788763.0, "step": 21379 }, { "epoch": 2.7197557562651062, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7420566082000732, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8867943286895752, "num_tokens": 811820866.0, "step": 21380 }, { "epoch": 2.7198829665436968, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6563804149627686, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8820594549179077, "num_tokens": 811866684.0, "step": 21381 }, { "epoch": 2.7200101768222873, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.745863676071167, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.882686972618103, "num_tokens": 811902728.0, "step": 21382 }, { "epoch": 2.720137387100878, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.653362512588501, "learning_rate": 1e-05, "loss": 0.4061, "mean_token_accuracy": 0.8829153180122375, "num_tokens": 811946051.0, "step": 21383 }, { "epoch": 2.7202645973794684, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.751254081726074, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.878211259841919, "num_tokens": 811984148.0, "step": 21384 }, { "epoch": 2.720391807658059, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.672595739364624, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8659441471099854, "num_tokens": 812020173.0, "step": 21385 }, { "epoch": 2.7205190179366494, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7518415451049805, "learning_rate": 1e-05, "loss": 0.3873, "mean_token_accuracy": 0.8892112970352173, "num_tokens": 812054509.0, "step": 21386 }, { "epoch": 2.72064622821524, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.660675287246704, "learning_rate": 1e-05, "loss": 0.4026, "mean_token_accuracy": 0.8850798010826111, "num_tokens": 812088717.0, "step": 21387 }, { "epoch": 2.7207734384938305, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7543528079986572, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8749134540557861, "num_tokens": 812126230.0, "step": 21388 }, { "epoch": 2.720900648772421, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6679370403289795, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8898663520812988, "num_tokens": 812157385.0, "step": 21389 }, { "epoch": 2.721027859051011, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7983367443084717, "learning_rate": 1e-05, "loss": 0.4879, "mean_token_accuracy": 0.861018180847168, "num_tokens": 812184653.0, "step": 21390 }, { "epoch": 2.721155069329602, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6330299377441406, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.8913565874099731, "num_tokens": 812226784.0, "step": 21391 }, { "epoch": 2.721282279608192, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7997148036956787, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.876766562461853, "num_tokens": 812261088.0, "step": 21392 }, { "epoch": 2.721409489886783, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6184353828430176, "learning_rate": 1e-05, "loss": 0.3832, "mean_token_accuracy": 0.8917027115821838, "num_tokens": 812296859.0, "step": 21393 }, { "epoch": 2.721536700165373, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.809368848800659, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8852822184562683, "num_tokens": 812334027.0, "step": 21394 }, { "epoch": 2.7216639104439637, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.1961669921875e-05, "grad_norm": 2.627744197845459, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8776262998580933, "num_tokens": 812370425.0, "step": 21395 }, { "epoch": 2.7217911207225542, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8237810134887695, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8811010122299194, "num_tokens": 812402067.0, "step": 21396 }, { "epoch": 2.7219183310011448, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6209301948547363, "learning_rate": 1e-05, "loss": 0.3706, "mean_token_accuracy": 0.8948493003845215, "num_tokens": 812436482.0, "step": 21397 }, { "epoch": 2.7220455412797353, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.825319290161133, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.866424024105072, "num_tokens": 812476059.0, "step": 21398 }, { "epoch": 2.722172751558326, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6092522144317627, "learning_rate": 1e-05, "loss": 0.3786, "mean_token_accuracy": 0.8923131823539734, "num_tokens": 812513846.0, "step": 21399 }, { "epoch": 2.7222999618369164, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8463947772979736, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8861621022224426, "num_tokens": 812550000.0, "step": 21400 }, { "epoch": 2.722427172115507, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.622762680053711, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8814554214477539, "num_tokens": 812589539.0, "step": 21401 }, { "epoch": 2.7225543823940974, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8515446186065674, "learning_rate": 1e-05, "loss": 0.3901, "mean_token_accuracy": 0.8905676007270813, "num_tokens": 812622699.0, "step": 21402 }, { "epoch": 2.722681592672688, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6222879886627197, "learning_rate": 1e-05, "loss": 0.3889, "mean_token_accuracy": 0.8873730897903442, "num_tokens": 812662945.0, "step": 21403 }, { "epoch": 2.7228088029512785, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.854623317718506, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8831363320350647, "num_tokens": 812700650.0, "step": 21404 }, { "epoch": 2.722936013229869, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6136162281036377, "learning_rate": 1e-05, "loss": 0.3815, "mean_token_accuracy": 0.8953368663787842, "num_tokens": 812742651.0, "step": 21405 }, { "epoch": 2.7230632235084595, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.864725351333618, "learning_rate": 1e-05, "loss": 0.4474, "mean_token_accuracy": 0.8717360496520996, "num_tokens": 812783792.0, "step": 21406 }, { "epoch": 2.72319043378705, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6201493740081787, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8729592561721802, "num_tokens": 812829059.0, "step": 21407 }, { "epoch": 2.7233176440656406, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.89009165763855, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8842439651489258, "num_tokens": 812859088.0, "step": 21408 }, { "epoch": 2.723444854344231, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.638071060180664, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8716793656349182, "num_tokens": 812899057.0, "step": 21409 }, { "epoch": 2.7235720646228216, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.9114227294921875e-05, "grad_norm": 2.88047194480896, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8713442087173462, "num_tokens": 812940711.0, "step": 21410 }, { "epoch": 2.723699274901412, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.633737325668335, "learning_rate": 1e-05, "loss": 0.4725, "mean_token_accuracy": 0.8608437776565552, "num_tokens": 812986967.0, "step": 21411 }, { "epoch": 2.7238264851800027, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 2.8762266635894775, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8800008296966553, "num_tokens": 813024396.0, "step": 21412 }, { "epoch": 2.723953695458593, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.654205560684204, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8734370470046997, "num_tokens": 813058204.0, "step": 21413 }, { "epoch": 2.7240809057371838, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.839897155761719e-05, "grad_norm": 2.8537046909332275, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8716106414794922, "num_tokens": 813098477.0, "step": 21414 }, { "epoch": 2.724208116015774, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6516246795654297, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8786040544509888, "num_tokens": 813138091.0, "step": 21415 }, { "epoch": 2.724335326294365, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.853217363357544, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8728487491607666, "num_tokens": 813170280.0, "step": 21416 }, { "epoch": 2.724462536572955, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.651543140411377, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8694605827331543, "num_tokens": 813209198.0, "step": 21417 }, { "epoch": 2.724589746851546, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.824882984161377, "learning_rate": 1e-05, "loss": 0.3581, "mean_token_accuracy": 0.9007949829101562, "num_tokens": 813248836.0, "step": 21418 }, { "epoch": 2.724716957130136, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6798267364501953, "learning_rate": 1e-05, "loss": 0.4372, "mean_token_accuracy": 0.8760879635810852, "num_tokens": 813279421.0, "step": 21419 }, { "epoch": 2.7248441674087265, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8069283962249756, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8696469664573669, "num_tokens": 813321615.0, "step": 21420 }, { "epoch": 2.724971377687317, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6687753200531006, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8816099166870117, "num_tokens": 813357338.0, "step": 21421 }, { "epoch": 2.7250985879659075, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8037166595458984, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.877941906452179, "num_tokens": 813397774.0, "step": 21422 }, { "epoch": 2.725225798244498, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6717441082000732, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8875764608383179, "num_tokens": 813432373.0, "step": 21423 }, { "epoch": 2.7253530085230886, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7690465450286865, "learning_rate": 1e-05, "loss": 0.3728, "mean_token_accuracy": 0.8983609676361084, "num_tokens": 813472354.0, "step": 21424 }, { "epoch": 2.725480218801679, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6814918518066406, "learning_rate": 1e-05, "loss": 0.3879, "mean_token_accuracy": 0.8881505727767944, "num_tokens": 813505498.0, "step": 21425 }, { "epoch": 2.7256074290802697, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7575104236602783, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8841944932937622, "num_tokens": 813546886.0, "step": 21426 }, { "epoch": 2.72573463935886, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.677175521850586, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8841102719306946, "num_tokens": 813583753.0, "step": 21427 }, { "epoch": 2.7258618496374507, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.773545265197754, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8775391578674316, "num_tokens": 813617561.0, "step": 21428 }, { "epoch": 2.7259890599160412, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6834099292755127, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8659904599189758, "num_tokens": 813652440.0, "step": 21429 }, { "epoch": 2.7261162701946318, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7373056411743164, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8857719898223877, "num_tokens": 813691118.0, "step": 21430 }, { "epoch": 2.7262434804732223, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.674469232559204, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8716150522232056, "num_tokens": 813733238.0, "step": 21431 }, { "epoch": 2.726370690751813, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.796720266342163, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8855663537979126, "num_tokens": 813770002.0, "step": 21432 }, { "epoch": 2.7264979010304033, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.653642177581787, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.876526951789856, "num_tokens": 813810778.0, "step": 21433 }, { "epoch": 2.726625111308994, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7815942764282227, "learning_rate": 1e-05, "loss": 0.4148, "mean_token_accuracy": 0.8784335851669312, "num_tokens": 813844004.0, "step": 21434 }, { "epoch": 2.7267523215875844, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6384332180023193, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8844645023345947, "num_tokens": 813884399.0, "step": 21435 }, { "epoch": 2.726879531866175, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7809338569641113, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8899756669998169, "num_tokens": 813920454.0, "step": 21436 }, { "epoch": 2.7270067421447655, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.645123243331909, "learning_rate": 1e-05, "loss": 0.3899, "mean_token_accuracy": 0.8874551057815552, "num_tokens": 813954629.0, "step": 21437 }, { "epoch": 2.7271339524233555, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7798542976379395, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8838115334510803, "num_tokens": 813992661.0, "step": 21438 }, { "epoch": 2.7272611627019465, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.662330150604248, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8821576833724976, "num_tokens": 814024401.0, "step": 21439 }, { "epoch": 2.7273883729805366, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7658886909484863, "learning_rate": 1e-05, "loss": 0.4241, "mean_token_accuracy": 0.8809834718704224, "num_tokens": 814064950.0, "step": 21440 }, { "epoch": 2.7275155832591276, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6434431076049805, "learning_rate": 1e-05, "loss": 0.412, "mean_token_accuracy": 0.8843304514884949, "num_tokens": 814109503.0, "step": 21441 }, { "epoch": 2.7276427935377177, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.778179883956909, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8754556775093079, "num_tokens": 814143981.0, "step": 21442 }, { "epoch": 2.7277700038163086, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6428611278533936, "learning_rate": 1e-05, "loss": 0.3862, "mean_token_accuracy": 0.8899296522140503, "num_tokens": 814180781.0, "step": 21443 }, { "epoch": 2.7278972140948987, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7702345848083496, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8783270716667175, "num_tokens": 814224841.0, "step": 21444 }, { "epoch": 2.7280244243734892, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6590170860290527, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8817166090011597, "num_tokens": 814258618.0, "step": 21445 }, { "epoch": 2.7281516346520798, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7681281566619873, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8836723566055298, "num_tokens": 814297286.0, "step": 21446 }, { "epoch": 2.7282788449306703, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6596884727478027, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8797343969345093, "num_tokens": 814331540.0, "step": 21447 }, { "epoch": 2.728406055209261, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7623071670532227, "learning_rate": 1e-05, "loss": 0.4626, "mean_token_accuracy": 0.8661118745803833, "num_tokens": 814371870.0, "step": 21448 }, { "epoch": 2.7285332654878514, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.659526824951172, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8854781985282898, "num_tokens": 814406561.0, "step": 21449 }, { "epoch": 2.728660475766442, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7604222297668457, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8844671249389648, "num_tokens": 814444176.0, "step": 21450 }, { "epoch": 2.7287876860450324, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6538968086242676, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8772127628326416, "num_tokens": 814485224.0, "step": 21451 }, { "epoch": 2.728914896323623, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7740325927734375, "learning_rate": 1e-05, "loss": 0.4013, "mean_token_accuracy": 0.8873535394668579, "num_tokens": 814518006.0, "step": 21452 }, { "epoch": 2.7290421066022135, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.646209955215454, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8768442273139954, "num_tokens": 814562883.0, "step": 21453 }, { "epoch": 2.729169316880804, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.762239933013916, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8857688903808594, "num_tokens": 814604619.0, "step": 21454 }, { "epoch": 2.7292965271593945, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.65476655960083, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.874554455280304, "num_tokens": 814641112.0, "step": 21455 }, { "epoch": 2.729423737437985, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7713823318481445, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8791123628616333, "num_tokens": 814674300.0, "step": 21456 }, { "epoch": 2.7295509477165756, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6449131965637207, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8856494426727295, "num_tokens": 814711013.0, "step": 21457 }, { "epoch": 2.729678157995166, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.767714262008667, "learning_rate": 1e-05, "loss": 0.3994, "mean_token_accuracy": 0.886589527130127, "num_tokens": 814750789.0, "step": 21458 }, { "epoch": 2.7298053682737566, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6459527015686035, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8794058561325073, "num_tokens": 814791498.0, "step": 21459 }, { "epoch": 2.729932578552347, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.774003505706787, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8820033073425293, "num_tokens": 814829811.0, "step": 21460 }, { "epoch": 2.7300597888309377, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6446166038513184, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8793901801109314, "num_tokens": 814864878.0, "step": 21461 }, { "epoch": 2.7301869991095282, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7721810340881348, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8855525851249695, "num_tokens": 814903843.0, "step": 21462 }, { "epoch": 2.7303142093881183, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6586456298828125, "learning_rate": 1e-05, "loss": 0.3988, "mean_token_accuracy": 0.8852429986000061, "num_tokens": 814943076.0, "step": 21463 }, { "epoch": 2.7304414196667093, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.767951011657715, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8875697255134583, "num_tokens": 814981257.0, "step": 21464 }, { "epoch": 2.7305686299452994, "ewc_loss": 0.078125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6430556774139404, "learning_rate": 1e-05, "loss": 0.3921, "mean_token_accuracy": 0.8870893120765686, "num_tokens": 815019763.0, "step": 21465 }, { "epoch": 2.7306958402238903, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.763302803039551, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8817213773727417, "num_tokens": 815061444.0, "step": 21466 }, { "epoch": 2.7308230505024804, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6409504413604736, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8761498332023621, "num_tokens": 815105141.0, "step": 21467 }, { "epoch": 2.730950260781071, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.781254291534424, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.877858579158783, "num_tokens": 815142839.0, "step": 21468 }, { "epoch": 2.7310774710596615, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.628933906555176, "learning_rate": 1e-05, "loss": 0.391, "mean_token_accuracy": 0.8884689807891846, "num_tokens": 815183077.0, "step": 21469 }, { "epoch": 2.731204681338252, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.798417329788208, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8810466527938843, "num_tokens": 815222601.0, "step": 21470 }, { "epoch": 2.7313318916168425, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6314620971679688, "learning_rate": 1e-05, "loss": 0.4396, "mean_token_accuracy": 0.876192569732666, "num_tokens": 815264587.0, "step": 21471 }, { "epoch": 2.731459101895433, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8013811111450195, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8782541751861572, "num_tokens": 815302575.0, "step": 21472 }, { "epoch": 2.7315863121740236, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.629423141479492, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8813366889953613, "num_tokens": 815342230.0, "step": 21473 }, { "epoch": 2.731713522452614, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.819589376449585, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.8872469067573547, "num_tokens": 815375906.0, "step": 21474 }, { "epoch": 2.7318407327312046, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6255595684051514, "learning_rate": 1e-05, "loss": 0.4156, "mean_token_accuracy": 0.8771835565567017, "num_tokens": 815415766.0, "step": 21475 }, { "epoch": 2.731967943009795, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8218796253204346, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8872005939483643, "num_tokens": 815449979.0, "step": 21476 }, { "epoch": 2.7320951532883857, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6323723793029785, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8745797276496887, "num_tokens": 815489736.0, "step": 21477 }, { "epoch": 2.7322223635669762, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.824706792831421, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8887425661087036, "num_tokens": 815525713.0, "step": 21478 }, { "epoch": 2.7323495738455668, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6284737586975098, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8749990463256836, "num_tokens": 815569231.0, "step": 21479 }, { "epoch": 2.7324767841241573, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8390369415283203, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8743166923522949, "num_tokens": 815607958.0, "step": 21480 }, { "epoch": 2.732603994402748, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.640648365020752, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8714102506637573, "num_tokens": 815645919.0, "step": 21481 }, { "epoch": 2.7327312046813383, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.823037624359131, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.874422550201416, "num_tokens": 815688568.0, "step": 21482 }, { "epoch": 2.732858414959929, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.662858009338379, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8739912509918213, "num_tokens": 815723130.0, "step": 21483 }, { "epoch": 2.7329856252385194, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8004140853881836, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.877442479133606, "num_tokens": 815764806.0, "step": 21484 }, { "epoch": 2.73311283551711, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.658282518386841, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8726866245269775, "num_tokens": 815800226.0, "step": 21485 }, { "epoch": 2.7332400457957005, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7987685203552246, "learning_rate": 1e-05, "loss": 0.3889, "mean_token_accuracy": 0.8906323313713074, "num_tokens": 815833625.0, "step": 21486 }, { "epoch": 2.733367256074291, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6469051837921143, "learning_rate": 1e-05, "loss": 0.3917, "mean_token_accuracy": 0.8874182105064392, "num_tokens": 815870617.0, "step": 21487 }, { "epoch": 2.733494466352881, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.800961494445801, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8757593631744385, "num_tokens": 815910854.0, "step": 21488 }, { "epoch": 2.733621676631472, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.655864715576172, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.88519287109375, "num_tokens": 815946467.0, "step": 21489 }, { "epoch": 2.733748886910062, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7949650287628174, "learning_rate": 1e-05, "loss": 0.4794, "mean_token_accuracy": 0.8631138801574707, "num_tokens": 815986924.0, "step": 21490 }, { "epoch": 2.733876097188653, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6561968326568604, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8862464427947998, "num_tokens": 816023960.0, "step": 21491 }, { "epoch": 2.734003307467243, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.799914598464966, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8759994506835938, "num_tokens": 816063911.0, "step": 21492 }, { "epoch": 2.7341305177458337, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6681368350982666, "learning_rate": 1e-05, "loss": 0.487, "mean_token_accuracy": 0.8584522008895874, "num_tokens": 816099979.0, "step": 21493 }, { "epoch": 2.7342577280244242, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7993953227996826, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.8687158823013306, "num_tokens": 816135226.0, "step": 21494 }, { "epoch": 2.7343849383030148, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6522293090820312, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8724413514137268, "num_tokens": 816174031.0, "step": 21495 }, { "epoch": 2.7345121485816053, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7822043895721436, "learning_rate": 1e-05, "loss": 0.4063, "mean_token_accuracy": 0.8848048448562622, "num_tokens": 816211194.0, "step": 21496 }, { "epoch": 2.734639358860196, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6500110626220703, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8807123899459839, "num_tokens": 816253849.0, "step": 21497 }, { "epoch": 2.7347665691387864, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7891743183135986, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8800209760665894, "num_tokens": 816290081.0, "step": 21498 }, { "epoch": 2.734893779417377, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.640624523162842, "learning_rate": 1e-05, "loss": 0.3553, "mean_token_accuracy": 0.898709237575531, "num_tokens": 816324818.0, "step": 21499 }, { "epoch": 2.7350209896959674, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7854111194610596, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8782356977462769, "num_tokens": 816369946.0, "step": 21500 }, { "epoch": 2.735148199974558, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.661602258682251, "learning_rate": 1e-05, "loss": 0.4667, "mean_token_accuracy": 0.8658939003944397, "num_tokens": 816405541.0, "step": 21501 }, { "epoch": 2.7352754102531485, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.791745185852051, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8756803274154663, "num_tokens": 816441232.0, "step": 21502 }, { "epoch": 2.735402620531739, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6657967567443848, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8759109973907471, "num_tokens": 816474538.0, "step": 21503 }, { "epoch": 2.7355298308103295, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.772382974624634, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8761318922042847, "num_tokens": 816512706.0, "step": 21504 }, { "epoch": 2.73565704108892, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.663498878479004, "learning_rate": 1e-05, "loss": 0.3972, "mean_token_accuracy": 0.8891869783401489, "num_tokens": 816548638.0, "step": 21505 }, { "epoch": 2.7357842513675106, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.770692825317383, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8731847405433655, "num_tokens": 816583721.0, "step": 21506 }, { "epoch": 2.735911461646101, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6498358249664307, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8819364309310913, "num_tokens": 816621873.0, "step": 21507 }, { "epoch": 2.7360386719246916, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.770695924758911, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8682428598403931, "num_tokens": 816659372.0, "step": 21508 }, { "epoch": 2.736165882203282, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.642997980117798, "learning_rate": 1e-05, "loss": 0.3786, "mean_token_accuracy": 0.8938871622085571, "num_tokens": 816696426.0, "step": 21509 }, { "epoch": 2.7362930924818727, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7789225578308105, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8836421966552734, "num_tokens": 816735344.0, "step": 21510 }, { "epoch": 2.7364203027604628, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6427903175354004, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.8908559679985046, "num_tokens": 816773026.0, "step": 21511 }, { "epoch": 2.7365475130390537, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7825212478637695, "learning_rate": 1e-05, "loss": 0.4086, "mean_token_accuracy": 0.8844629526138306, "num_tokens": 816809777.0, "step": 21512 }, { "epoch": 2.736674723317644, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.647247314453125, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.868419349193573, "num_tokens": 816849018.0, "step": 21513 }, { "epoch": 2.736801933596235, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.786984443664551, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8750026822090149, "num_tokens": 816888867.0, "step": 21514 }, { "epoch": 2.736929143874825, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.636176109313965, "learning_rate": 1e-05, "loss": 0.3918, "mean_token_accuracy": 0.8867419362068176, "num_tokens": 816929221.0, "step": 21515 }, { "epoch": 2.737056354153416, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7948315143585205, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8852359056472778, "num_tokens": 816965628.0, "step": 21516 }, { "epoch": 2.737183564432006, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6500258445739746, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8766210079193115, "num_tokens": 816997908.0, "step": 21517 }, { "epoch": 2.7373107747105965, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.794348955154419, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.882606029510498, "num_tokens": 817034339.0, "step": 21518 }, { "epoch": 2.737437984989187, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6503536701202393, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8726538419723511, "num_tokens": 817073006.0, "step": 21519 }, { "epoch": 2.7375651952677775, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7934207916259766, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8749892115592957, "num_tokens": 817114517.0, "step": 21520 }, { "epoch": 2.737692405546368, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.639970541000366, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8840383291244507, "num_tokens": 817158249.0, "step": 21521 }, { "epoch": 2.7378196158249586, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.79343318939209, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.884116530418396, "num_tokens": 817197178.0, "step": 21522 }, { "epoch": 2.737946826103549, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.650637149810791, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8862864971160889, "num_tokens": 817231406.0, "step": 21523 }, { "epoch": 2.7380740363821396, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.804333448410034, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.881392240524292, "num_tokens": 817269984.0, "step": 21524 }, { "epoch": 2.73820124666073, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.648195266723633, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8839791417121887, "num_tokens": 817308997.0, "step": 21525 }, { "epoch": 2.7383284569393207, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.793480157852173, "learning_rate": 1e-05, "loss": 0.4563, "mean_token_accuracy": 0.8705725073814392, "num_tokens": 817349329.0, "step": 21526 }, { "epoch": 2.7384556672179112, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6574416160583496, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8792648315429688, "num_tokens": 817384683.0, "step": 21527 }, { "epoch": 2.7385828774965018, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7848751544952393, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8799148797988892, "num_tokens": 817423202.0, "step": 21528 }, { "epoch": 2.7387100877750923, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.642967939376831, "learning_rate": 1e-05, "loss": 0.3915, "mean_token_accuracy": 0.8900866508483887, "num_tokens": 817462048.0, "step": 21529 }, { "epoch": 2.738837298053683, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7825682163238525, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8750854134559631, "num_tokens": 817508820.0, "step": 21530 }, { "epoch": 2.7389645083322733, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6399195194244385, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8758903741836548, "num_tokens": 817554140.0, "step": 21531 }, { "epoch": 2.739091718610864, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7918996810913086, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8679006099700928, "num_tokens": 817601629.0, "step": 21532 }, { "epoch": 2.7392189288894544, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6571204662323, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8745238780975342, "num_tokens": 817641900.0, "step": 21533 }, { "epoch": 2.739346139168045, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.779533863067627, "learning_rate": 1e-05, "loss": 0.391, "mean_token_accuracy": 0.8910420536994934, "num_tokens": 817681843.0, "step": 21534 }, { "epoch": 2.7394733494466355, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6435282230377197, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8624433875083923, "num_tokens": 817728548.0, "step": 21535 }, { "epoch": 2.7396005597252255, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.805910348892212, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8855040073394775, "num_tokens": 817759087.0, "step": 21536 }, { "epoch": 2.7397277700038165, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6417396068573, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8731414079666138, "num_tokens": 817799056.0, "step": 21537 }, { "epoch": 2.7398549802824066, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.807982921600342, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8808651566505432, "num_tokens": 817831859.0, "step": 21538 }, { "epoch": 2.7399821905609976, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.63757061958313, "learning_rate": 1e-05, "loss": 0.4162, "mean_token_accuracy": 0.8799103498458862, "num_tokens": 817869699.0, "step": 21539 }, { "epoch": 2.7401094008395877, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.81579852104187, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8714509010314941, "num_tokens": 817909218.0, "step": 21540 }, { "epoch": 2.7402366111181786, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6390457153320312, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.885802686214447, "num_tokens": 817948398.0, "step": 21541 }, { "epoch": 2.7403638213967687, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8159658908843994, "learning_rate": 1e-05, "loss": 0.4509, "mean_token_accuracy": 0.87115478515625, "num_tokens": 817987351.0, "step": 21542 }, { "epoch": 2.7404910316753592, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.622972249984741, "learning_rate": 1e-05, "loss": 0.4014, "mean_token_accuracy": 0.8863506317138672, "num_tokens": 818031306.0, "step": 21543 }, { "epoch": 2.7406182419539498, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8318936824798584, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8828972578048706, "num_tokens": 818067658.0, "step": 21544 }, { "epoch": 2.7407454522325403, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6289234161376953, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8682695031166077, "num_tokens": 818109766.0, "step": 21545 }, { "epoch": 2.740872662511131, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8465633392333984, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8688763976097107, "num_tokens": 818144793.0, "step": 21546 }, { "epoch": 2.7409998727897213, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003604888916015625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6326770782470703, "learning_rate": 1e-05, "loss": 0.4519, "mean_token_accuracy": 0.8736659288406372, "num_tokens": 818184140.0, "step": 21547 }, { "epoch": 2.741127083068312, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8326897621154785, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8754866123199463, "num_tokens": 818222216.0, "step": 21548 }, { "epoch": 2.7412542933469024, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6451568603515625, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8879896998405457, "num_tokens": 818256356.0, "step": 21549 }, { "epoch": 2.741381503625493, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8473997116088867, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8713463544845581, "num_tokens": 818300012.0, "step": 21550 }, { "epoch": 2.7415087139040835, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6401689052581787, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8771914839744568, "num_tokens": 818336090.0, "step": 21551 }, { "epoch": 2.741635924182674, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.833488702774048, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8803372383117676, "num_tokens": 818376475.0, "step": 21552 }, { "epoch": 2.7417631344612645, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.635488748550415, "learning_rate": 1e-05, "loss": 0.4007, "mean_token_accuracy": 0.8871131539344788, "num_tokens": 818417483.0, "step": 21553 }, { "epoch": 2.741890344739855, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.81853985786438, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8789269328117371, "num_tokens": 818460323.0, "step": 21554 }, { "epoch": 2.7420175550184456, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6502771377563477, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8786051273345947, "num_tokens": 818496989.0, "step": 21555 }, { "epoch": 2.742144765297036, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8189380168914795, "learning_rate": 1e-05, "loss": 0.3841, "mean_token_accuracy": 0.8917797207832336, "num_tokens": 818532755.0, "step": 21556 }, { "epoch": 2.7422719755756266, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6452505588531494, "learning_rate": 1e-05, "loss": 0.4727, "mean_token_accuracy": 0.8632107973098755, "num_tokens": 818575237.0, "step": 21557 }, { "epoch": 2.742399185854217, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.816012144088745, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8859920501708984, "num_tokens": 818607865.0, "step": 21558 }, { "epoch": 2.7425263961328077, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6410064697265625, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.8901251554489136, "num_tokens": 818647807.0, "step": 21559 }, { "epoch": 2.742653606411398, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.814089775085449, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8792022466659546, "num_tokens": 818686951.0, "step": 21560 }, { "epoch": 2.7427808166899883, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6471176147460938, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8754669427871704, "num_tokens": 818724813.0, "step": 21561 }, { "epoch": 2.7429080269685793, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8160040378570557, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.88289475440979, "num_tokens": 818757865.0, "step": 21562 }, { "epoch": 2.7430352372471694, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.652773141860962, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8698513507843018, "num_tokens": 818796994.0, "step": 21563 }, { "epoch": 2.7431624475257603, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.797501564025879, "learning_rate": 1e-05, "loss": 0.4015, "mean_token_accuracy": 0.8861019611358643, "num_tokens": 818836408.0, "step": 21564 }, { "epoch": 2.7432896578043504, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.65444016456604, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8818308115005493, "num_tokens": 818876134.0, "step": 21565 }, { "epoch": 2.743416868082941, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8049161434173584, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8794330954551697, "num_tokens": 818911357.0, "step": 21566 }, { "epoch": 2.7435440783615315, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6715006828308105, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.880700409412384, "num_tokens": 818942743.0, "step": 21567 }, { "epoch": 2.743671288640122, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7842774391174316, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.8780345916748047, "num_tokens": 818979217.0, "step": 21568 }, { "epoch": 2.7437984989187125, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.658363103866577, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8699865341186523, "num_tokens": 819019991.0, "step": 21569 }, { "epoch": 2.743925709197303, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.779611825942993, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8740265369415283, "num_tokens": 819059264.0, "step": 21570 }, { "epoch": 2.7440529194758936, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.648555278778076, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.880240797996521, "num_tokens": 819099649.0, "step": 21571 }, { "epoch": 2.744180129754484, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.772249221801758, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8758238554000854, "num_tokens": 819142434.0, "step": 21572 }, { "epoch": 2.7443073400330746, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.662721872329712, "learning_rate": 1e-05, "loss": 0.4824, "mean_token_accuracy": 0.860116720199585, "num_tokens": 819183908.0, "step": 21573 }, { "epoch": 2.744434550311665, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.772557020187378, "learning_rate": 1e-05, "loss": 0.3786, "mean_token_accuracy": 0.8941085338592529, "num_tokens": 819218975.0, "step": 21574 }, { "epoch": 2.7445617605902557, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6416168212890625, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8819468021392822, "num_tokens": 819264427.0, "step": 21575 }, { "epoch": 2.7446889708688462, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7785520553588867, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.871444821357727, "num_tokens": 819304947.0, "step": 21576 }, { "epoch": 2.7448161811474368, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.640177011489868, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.8873984813690186, "num_tokens": 819345540.0, "step": 21577 }, { "epoch": 2.7449433914260273, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.795269012451172, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8804394006729126, "num_tokens": 819383007.0, "step": 21578 }, { "epoch": 2.745070601704618, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6562507152557373, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.873568058013916, "num_tokens": 819415640.0, "step": 21579 }, { "epoch": 2.7451978119832083, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.776531457901001, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8683960437774658, "num_tokens": 819459497.0, "step": 21580 }, { "epoch": 2.745325022261799, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6706225872039795, "learning_rate": 1e-05, "loss": 0.4653, "mean_token_accuracy": 0.868253231048584, "num_tokens": 819497010.0, "step": 21581 }, { "epoch": 2.7454522325403894, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7626070976257324, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8797778487205505, "num_tokens": 819535659.0, "step": 21582 }, { "epoch": 2.74557944281898, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.659186840057373, "learning_rate": 1e-05, "loss": 0.43, "mean_token_accuracy": 0.8790034651756287, "num_tokens": 819572774.0, "step": 21583 }, { "epoch": 2.7457066530975704, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.753101110458374, "learning_rate": 1e-05, "loss": 0.3781, "mean_token_accuracy": 0.8926406502723694, "num_tokens": 819607461.0, "step": 21584 }, { "epoch": 2.745833863376161, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6600875854492188, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8856008648872375, "num_tokens": 819645213.0, "step": 21585 }, { "epoch": 2.745961073654751, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.763965606689453, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8827764391899109, "num_tokens": 819679537.0, "step": 21586 }, { "epoch": 2.746088283933342, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6522836685180664, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8849904537200928, "num_tokens": 819717414.0, "step": 21587 }, { "epoch": 2.746215494211932, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 3.3140077590942383, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8875351548194885, "num_tokens": 819757875.0, "step": 21588 }, { "epoch": 2.746342704490523, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.6300406455993652, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.876315712928772, "num_tokens": 819796002.0, "step": 21589 }, { "epoch": 2.746469914769113, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 2.9863617420196533, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.8656130433082581, "num_tokens": 819834736.0, "step": 21590 }, { "epoch": 2.7465971250477037, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.597165584564209, "learning_rate": 1e-05, "loss": 0.3829, "mean_token_accuracy": 0.8913201689720154, "num_tokens": 819877566.0, "step": 21591 }, { "epoch": 2.7467243353262942, "ewc_loss": 0.0908203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 5.435943603515625e-05, "grad_norm": 3.0325448513031006, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8766717910766602, "num_tokens": 819914843.0, "step": 21592 }, { "epoch": 2.7468515456048848, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.6677446365356445, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8791302442550659, "num_tokens": 819957181.0, "step": 21593 }, { "epoch": 2.7469787558834753, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 5.364418029785156e-05, "grad_norm": 3.007197856903076, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8825845718383789, "num_tokens": 819991291.0, "step": 21594 }, { "epoch": 2.747105966162066, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.6818668842315674, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8752434849739075, "num_tokens": 820033096.0, "step": 21595 }, { "epoch": 2.7472331764406563, "ewc_loss": 0.08984375, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 5.364418029785156e-05, "grad_norm": 2.9838130474090576, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.8872414231300354, "num_tokens": 820070200.0, "step": 21596 }, { "epoch": 2.747360386719247, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.7180397510528564, "learning_rate": 1e-05, "loss": 0.3896, "mean_token_accuracy": 0.8915055990219116, "num_tokens": 820100989.0, "step": 21597 }, { "epoch": 2.7474875969978374, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.000362396240234375, "ewc_loss_parallel": 5.1021575927734375e-05, "grad_norm": 2.9259157180786133, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8767381906509399, "num_tokens": 820145052.0, "step": 21598 }, { "epoch": 2.747614807276428, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.708742618560791, "learning_rate": 1e-05, "loss": 0.3854, "mean_token_accuracy": 0.8901139497756958, "num_tokens": 820181463.0, "step": 21599 }, { "epoch": 2.7477420175550185, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.9591064453125e-05, "grad_norm": 2.9075570106506348, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8771100044250488, "num_tokens": 820219226.0, "step": 21600 }, { "epoch": 2.747869227833609, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.701927661895752, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8753281235694885, "num_tokens": 820255953.0, "step": 21601 }, { "epoch": 2.7479964381121995, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.8777823448181152, "learning_rate": 1e-05, "loss": 0.443, "mean_token_accuracy": 0.8760479688644409, "num_tokens": 820296587.0, "step": 21602 }, { "epoch": 2.74812364839079, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.698831081390381, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8831095695495605, "num_tokens": 820333815.0, "step": 21603 }, { "epoch": 2.7482508586693806, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8371355533599854, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8808669447898865, "num_tokens": 820376401.0, "step": 21604 }, { "epoch": 2.748378068947971, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6895015239715576, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8834561109542847, "num_tokens": 820417355.0, "step": 21605 }, { "epoch": 2.7485052792265616, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.806804656982422, "learning_rate": 1e-05, "loss": 0.3438, "mean_token_accuracy": 0.90547776222229, "num_tokens": 820456788.0, "step": 21606 }, { "epoch": 2.748632489505152, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.668572425842285, "learning_rate": 1e-05, "loss": 0.3786, "mean_token_accuracy": 0.8945612907409668, "num_tokens": 820499767.0, "step": 21607 }, { "epoch": 2.7487596997837427, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7999765872955322, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8879594802856445, "num_tokens": 820543052.0, "step": 21608 }, { "epoch": 2.7488869100623328, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.663984537124634, "learning_rate": 1e-05, "loss": 0.3871, "mean_token_accuracy": 0.8917346000671387, "num_tokens": 820586162.0, "step": 21609 }, { "epoch": 2.7490141203409237, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8067827224731445, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8716423511505127, "num_tokens": 820626703.0, "step": 21610 }, { "epoch": 2.749141330619514, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6794025897979736, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8684248924255371, "num_tokens": 820663936.0, "step": 21611 }, { "epoch": 2.749268540898105, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7760355472564697, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8836200833320618, "num_tokens": 820703761.0, "step": 21612 }, { "epoch": 2.749395751176695, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6692471504211426, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8720721006393433, "num_tokens": 820742505.0, "step": 21613 }, { "epoch": 2.749522961455286, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7668776512145996, "learning_rate": 1e-05, "loss": 0.3923, "mean_token_accuracy": 0.8908709287643433, "num_tokens": 820784185.0, "step": 21614 }, { "epoch": 2.749650171733876, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6654300689697266, "learning_rate": 1e-05, "loss": 0.466, "mean_token_accuracy": 0.8625454902648926, "num_tokens": 820826163.0, "step": 21615 }, { "epoch": 2.7497773820124665, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.766021728515625, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8746803402900696, "num_tokens": 820861598.0, "step": 21616 }, { "epoch": 2.749904592291057, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6702933311462402, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8747061491012573, "num_tokens": 820901201.0, "step": 21617 }, { "epoch": 2.7500318025696475, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7514593601226807, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8895769119262695, "num_tokens": 820941773.0, "step": 21618 }, { "epoch": 2.750159012848238, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6754415035247803, "learning_rate": 1e-05, "loss": 0.4912, "mean_token_accuracy": 0.8576551675796509, "num_tokens": 820984610.0, "step": 21619 }, { "epoch": 2.7502862231268286, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.744863271713257, "learning_rate": 1e-05, "loss": 0.3867, "mean_token_accuracy": 0.889823853969574, "num_tokens": 821020814.0, "step": 21620 }, { "epoch": 2.750413433405419, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6802706718444824, "learning_rate": 1e-05, "loss": 0.3883, "mean_token_accuracy": 0.8887151479721069, "num_tokens": 821053073.0, "step": 21621 }, { "epoch": 2.7505406436840096, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.746155023574829, "learning_rate": 1e-05, "loss": 0.4277, "mean_token_accuracy": 0.8760525584220886, "num_tokens": 821088872.0, "step": 21622 }, { "epoch": 2.7506678539626, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.687612295150757, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8707776069641113, "num_tokens": 821124314.0, "step": 21623 }, { "epoch": 2.7507950642411907, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7611026763916016, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8805292844772339, "num_tokens": 821152025.0, "step": 21624 }, { "epoch": 2.750922274519781, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6579501628875732, "learning_rate": 1e-05, "loss": 0.4532, "mean_token_accuracy": 0.8677734732627869, "num_tokens": 821198138.0, "step": 21625 }, { "epoch": 2.7510494847983717, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7369751930236816, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8795150518417358, "num_tokens": 821241256.0, "step": 21626 }, { "epoch": 2.7511766950769623, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6778507232666016, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8861513137817383, "num_tokens": 821274932.0, "step": 21627 }, { "epoch": 2.751303905355553, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.745321273803711, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8794920444488525, "num_tokens": 821310769.0, "step": 21628 }, { "epoch": 2.7514311156341433, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6741936206817627, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.87889564037323, "num_tokens": 821346474.0, "step": 21629 }, { "epoch": 2.751558325912734, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7366716861724854, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8799905180931091, "num_tokens": 821385293.0, "step": 21630 }, { "epoch": 2.7516855361913244, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.67570424079895, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.8900552988052368, "num_tokens": 821421161.0, "step": 21631 }, { "epoch": 2.751812746469915, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7278058528900146, "learning_rate": 1e-05, "loss": 0.4037, "mean_token_accuracy": 0.889480471611023, "num_tokens": 821459937.0, "step": 21632 }, { "epoch": 2.7519399567485054, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.665496587753296, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8823024034500122, "num_tokens": 821503892.0, "step": 21633 }, { "epoch": 2.7520671670270955, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.747788429260254, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8863374590873718, "num_tokens": 821537781.0, "step": 21634 }, { "epoch": 2.7521943773056865, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6720144748687744, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8671081066131592, "num_tokens": 821576114.0, "step": 21635 }, { "epoch": 2.7523215875842766, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7456319332122803, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.885120689868927, "num_tokens": 821613587.0, "step": 21636 }, { "epoch": 2.7524487978628676, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6578056812286377, "learning_rate": 1e-05, "loss": 0.3455, "mean_token_accuracy": 0.903407871723175, "num_tokens": 821650956.0, "step": 21637 }, { "epoch": 2.7525760081414576, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7564949989318848, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.871628999710083, "num_tokens": 821692943.0, "step": 21638 }, { "epoch": 2.7527032184200486, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.663471221923828, "learning_rate": 1e-05, "loss": 0.47, "mean_token_accuracy": 0.8646280765533447, "num_tokens": 821735344.0, "step": 21639 }, { "epoch": 2.7528304286986387, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.756556510925293, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8813748955726624, "num_tokens": 821772910.0, "step": 21640 }, { "epoch": 2.7529576389772292, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6599864959716797, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8870093822479248, "num_tokens": 821810211.0, "step": 21641 }, { "epoch": 2.7530848492558198, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7649827003479004, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8801835775375366, "num_tokens": 821840782.0, "step": 21642 }, { "epoch": 2.7532120595344103, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.649634599685669, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.87778240442276, "num_tokens": 821879160.0, "step": 21643 }, { "epoch": 2.753339269813001, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7688236236572266, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.8892223238945007, "num_tokens": 821915344.0, "step": 21644 }, { "epoch": 2.7534664800915913, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6531903743743896, "learning_rate": 1e-05, "loss": 0.4743, "mean_token_accuracy": 0.8584907054901123, "num_tokens": 821955628.0, "step": 21645 }, { "epoch": 2.753593690370182, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7666168212890625, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8755192756652832, "num_tokens": 821996917.0, "step": 21646 }, { "epoch": 2.7537209006487724, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.646300792694092, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8827294111251831, "num_tokens": 822034222.0, "step": 21647 }, { "epoch": 2.753848110927363, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7641592025756836, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.876162052154541, "num_tokens": 822072350.0, "step": 21648 }, { "epoch": 2.7539753212059535, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6395552158355713, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8759253025054932, "num_tokens": 822111495.0, "step": 21649 }, { "epoch": 2.754102531484544, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7932400703430176, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.877131462097168, "num_tokens": 822148132.0, "step": 21650 }, { "epoch": 2.7542297417631345, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6282787322998047, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8807516098022461, "num_tokens": 822187440.0, "step": 21651 }, { "epoch": 2.754356952041725, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7884576320648193, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8800851702690125, "num_tokens": 822225936.0, "step": 21652 }, { "epoch": 2.7544841623203156, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.63312029838562, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8773908615112305, "num_tokens": 822266187.0, "step": 21653 }, { "epoch": 2.754611372598906, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7951533794403076, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8872823119163513, "num_tokens": 822302963.0, "step": 21654 }, { "epoch": 2.7547385828774966, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6488325595855713, "learning_rate": 1e-05, "loss": 0.3912, "mean_token_accuracy": 0.8902587890625, "num_tokens": 822336969.0, "step": 21655 }, { "epoch": 2.754865793156087, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.789929151535034, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8794097304344177, "num_tokens": 822376847.0, "step": 21656 }, { "epoch": 2.7549930034346777, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6423161029815674, "learning_rate": 1e-05, "loss": 0.4419, "mean_token_accuracy": 0.8739433288574219, "num_tokens": 822417858.0, "step": 21657 }, { "epoch": 2.755120213713268, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7975378036499023, "learning_rate": 1e-05, "loss": 0.4618, "mean_token_accuracy": 0.8694939017295837, "num_tokens": 822459946.0, "step": 21658 }, { "epoch": 2.7552474239918583, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6504263877868652, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8741573095321655, "num_tokens": 822498093.0, "step": 21659 }, { "epoch": 2.7553746342704493, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.796599864959717, "learning_rate": 1e-05, "loss": 0.4041, "mean_token_accuracy": 0.8865071535110474, "num_tokens": 822535918.0, "step": 21660 }, { "epoch": 2.7555018445490393, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6383888721466064, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8782055974006653, "num_tokens": 822574589.0, "step": 21661 }, { "epoch": 2.7556290548276303, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7958621978759766, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8861038684844971, "num_tokens": 822612437.0, "step": 21662 }, { "epoch": 2.7557562651062204, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6367228031158447, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8816365003585815, "num_tokens": 822654834.0, "step": 21663 }, { "epoch": 2.755883475384811, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7989420890808105, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.88913893699646, "num_tokens": 822689555.0, "step": 21664 }, { "epoch": 2.7560106856634015, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.638286828994751, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8656154274940491, "num_tokens": 822731609.0, "step": 21665 }, { "epoch": 2.756137895941992, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8034768104553223, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8720874786376953, "num_tokens": 822769881.0, "step": 21666 }, { "epoch": 2.7562651062205825, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6402065753936768, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8681650757789612, "num_tokens": 822810369.0, "step": 21667 }, { "epoch": 2.756392316499173, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.797487735748291, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8767112493515015, "num_tokens": 822852476.0, "step": 21668 }, { "epoch": 2.7565195267777636, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6527345180511475, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8637858629226685, "num_tokens": 822890745.0, "step": 21669 }, { "epoch": 2.756646737056354, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7835195064544678, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8786291480064392, "num_tokens": 822931243.0, "step": 21670 }, { "epoch": 2.7567739473349446, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6404354572296143, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.8784440755844116, "num_tokens": 822967560.0, "step": 21671 }, { "epoch": 2.756901157613535, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7882261276245117, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.882672905921936, "num_tokens": 823007129.0, "step": 21672 }, { "epoch": 2.7570283678921257, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.654236316680908, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8756736516952515, "num_tokens": 823043722.0, "step": 21673 }, { "epoch": 2.757155578170716, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8022985458374023, "learning_rate": 1e-05, "loss": 0.4828, "mean_token_accuracy": 0.8627467751502991, "num_tokens": 823079937.0, "step": 21674 }, { "epoch": 2.7572827884493067, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.635861396789551, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8762978315353394, "num_tokens": 823120455.0, "step": 21675 }, { "epoch": 2.7574099987278973, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8162150382995605, "learning_rate": 1e-05, "loss": 0.4406, "mean_token_accuracy": 0.8731284141540527, "num_tokens": 823151415.0, "step": 21676 }, { "epoch": 2.757537209006488, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.632554531097412, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8795647621154785, "num_tokens": 823188281.0, "step": 21677 }, { "epoch": 2.7576644192850783, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8112082481384277, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.8843451738357544, "num_tokens": 823226040.0, "step": 21678 }, { "epoch": 2.757791629563669, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.636711359024048, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8762936592102051, "num_tokens": 823261246.0, "step": 21679 }, { "epoch": 2.7579188398422594, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8133745193481445, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8803635835647583, "num_tokens": 823300034.0, "step": 21680 }, { "epoch": 2.75804605012085, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.640780448913574, "learning_rate": 1e-05, "loss": 0.3685, "mean_token_accuracy": 0.8941842317581177, "num_tokens": 823332152.0, "step": 21681 }, { "epoch": 2.7581732603994404, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.804349422454834, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8804569244384766, "num_tokens": 823373319.0, "step": 21682 }, { "epoch": 2.758300470678031, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6446614265441895, "learning_rate": 1e-05, "loss": 0.379, "mean_token_accuracy": 0.8941282033920288, "num_tokens": 823409875.0, "step": 21683 }, { "epoch": 2.758427680956621, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7920680046081543, "learning_rate": 1e-05, "loss": 0.384, "mean_token_accuracy": 0.8936948180198669, "num_tokens": 823450941.0, "step": 21684 }, { "epoch": 2.758554891235212, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6563332080841064, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8857361078262329, "num_tokens": 823493483.0, "step": 21685 }, { "epoch": 2.758682101513802, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.788710832595825, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8836574554443359, "num_tokens": 823531682.0, "step": 21686 }, { "epoch": 2.758809311792393, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6685142517089844, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.888670802116394, "num_tokens": 823564089.0, "step": 21687 }, { "epoch": 2.758936522070983, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7737765312194824, "learning_rate": 1e-05, "loss": 0.3665, "mean_token_accuracy": 0.8968198299407959, "num_tokens": 823598700.0, "step": 21688 }, { "epoch": 2.7590637323495737, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6577866077423096, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8748083114624023, "num_tokens": 823640248.0, "step": 21689 }, { "epoch": 2.7591909426281642, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.779340982437134, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8758216500282288, "num_tokens": 823676303.0, "step": 21690 }, { "epoch": 2.7593181529067548, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.656836748123169, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8740800619125366, "num_tokens": 823717736.0, "step": 21691 }, { "epoch": 2.7594453631853453, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.773980140686035, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8761791586875916, "num_tokens": 823755602.0, "step": 21692 }, { "epoch": 2.759572573463936, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6684694290161133, "learning_rate": 1e-05, "loss": 0.4674, "mean_token_accuracy": 0.8650869131088257, "num_tokens": 823792257.0, "step": 21693 }, { "epoch": 2.7596997837425263, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7492799758911133, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8803699612617493, "num_tokens": 823835655.0, "step": 21694 }, { "epoch": 2.759826994021117, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6690306663513184, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.874014139175415, "num_tokens": 823874840.0, "step": 21695 }, { "epoch": 2.7599542042997074, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7573633193969727, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8716039657592773, "num_tokens": 823910844.0, "step": 21696 }, { "epoch": 2.760081414578298, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6553163528442383, "learning_rate": 1e-05, "loss": 0.3827, "mean_token_accuracy": 0.8922477960586548, "num_tokens": 823948472.0, "step": 21697 }, { "epoch": 2.7602086248568884, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7467401027679443, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8897802829742432, "num_tokens": 823997532.0, "step": 21698 }, { "epoch": 2.760335835135479, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6624972820281982, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.870569109916687, "num_tokens": 824037426.0, "step": 21699 }, { "epoch": 2.7604630454140695, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7566468715667725, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8740304112434387, "num_tokens": 824076731.0, "step": 21700 }, { "epoch": 2.76059025569266, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6686182022094727, "learning_rate": 1e-05, "loss": 0.4718, "mean_token_accuracy": 0.8646420240402222, "num_tokens": 824119445.0, "step": 21701 }, { "epoch": 2.7607174659712506, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.761094570159912, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8691767454147339, "num_tokens": 824157372.0, "step": 21702 }, { "epoch": 2.760844676249841, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.640418529510498, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.877436101436615, "num_tokens": 824204312.0, "step": 21703 }, { "epoch": 2.7609718865284316, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7698426246643066, "learning_rate": 1e-05, "loss": 0.4029, "mean_token_accuracy": 0.8853603005409241, "num_tokens": 824245951.0, "step": 21704 }, { "epoch": 2.761099096807022, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6529648303985596, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.8787448406219482, "num_tokens": 824288742.0, "step": 21705 }, { "epoch": 2.7612263070856127, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.775658369064331, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8738270998001099, "num_tokens": 824326435.0, "step": 21706 }, { "epoch": 2.7613535173642028, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.658278703689575, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8796537518501282, "num_tokens": 824359442.0, "step": 21707 }, { "epoch": 2.7614807276427937, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7970731258392334, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.8679789304733276, "num_tokens": 824392387.0, "step": 21708 }, { "epoch": 2.761607937921384, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6512234210968018, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8742550611495972, "num_tokens": 824427860.0, "step": 21709 }, { "epoch": 2.761735148199975, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7732203006744385, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8778764009475708, "num_tokens": 824466011.0, "step": 21710 }, { "epoch": 2.761862358478565, "ewc_loss": 0.07861328125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6346466541290283, "learning_rate": 1e-05, "loss": 0.3808, "mean_token_accuracy": 0.8906011581420898, "num_tokens": 824507235.0, "step": 21711 }, { "epoch": 2.761989568757156, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.803748607635498, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8840867280960083, "num_tokens": 824538807.0, "step": 21712 }, { "epoch": 2.762116779035746, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.645294666290283, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8797909021377563, "num_tokens": 824575047.0, "step": 21713 }, { "epoch": 2.7622439893143365, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.802273988723755, "learning_rate": 1e-05, "loss": 0.3669, "mean_token_accuracy": 0.8973449468612671, "num_tokens": 824604113.0, "step": 21714 }, { "epoch": 2.762371199592927, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6470630168914795, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8672064542770386, "num_tokens": 824640912.0, "step": 21715 }, { "epoch": 2.7624984098715175, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.792470932006836, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8857551217079163, "num_tokens": 824682739.0, "step": 21716 }, { "epoch": 2.762625620150108, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6553492546081543, "learning_rate": 1e-05, "loss": 0.4766, "mean_token_accuracy": 0.8623162508010864, "num_tokens": 824719973.0, "step": 21717 }, { "epoch": 2.7627528304286986, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7857065200805664, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8834455013275146, "num_tokens": 824761214.0, "step": 21718 }, { "epoch": 2.762880040707289, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6451776027679443, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8691730499267578, "num_tokens": 824801114.0, "step": 21719 }, { "epoch": 2.7630072509858796, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003643035888671875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8074636459350586, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8744137287139893, "num_tokens": 824832734.0, "step": 21720 }, { "epoch": 2.76313446126447, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6807548999786377, "learning_rate": 1e-05, "loss": 0.432, "mean_token_accuracy": 0.8731942772865295, "num_tokens": 824867089.0, "step": 21721 }, { "epoch": 2.7632616715430607, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7644362449645996, "learning_rate": 1e-05, "loss": 0.387, "mean_token_accuracy": 0.890968918800354, "num_tokens": 824909173.0, "step": 21722 }, { "epoch": 2.763388881821651, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6617650985717773, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8873323202133179, "num_tokens": 824943815.0, "step": 21723 }, { "epoch": 2.7635160921002417, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7676477432250977, "learning_rate": 1e-05, "loss": 0.3791, "mean_token_accuracy": 0.8950818777084351, "num_tokens": 824978708.0, "step": 21724 }, { "epoch": 2.7636433023788323, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.671590566635132, "learning_rate": 1e-05, "loss": 0.4848, "mean_token_accuracy": 0.8640382885932922, "num_tokens": 825015226.0, "step": 21725 }, { "epoch": 2.763770512657423, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.756117582321167, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8698506951332092, "num_tokens": 825054417.0, "step": 21726 }, { "epoch": 2.7638977229360133, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6621878147125244, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8929042220115662, "num_tokens": 825093757.0, "step": 21727 }, { "epoch": 2.764024933214604, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7705001831054688, "learning_rate": 1e-05, "loss": 0.4837, "mean_token_accuracy": 0.8629717230796814, "num_tokens": 825129955.0, "step": 21728 }, { "epoch": 2.7641521434931944, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.649394989013672, "learning_rate": 1e-05, "loss": 0.39, "mean_token_accuracy": 0.8899895548820496, "num_tokens": 825168426.0, "step": 21729 }, { "epoch": 2.764279353771785, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7631587982177734, "learning_rate": 1e-05, "loss": 0.4672, "mean_token_accuracy": 0.8661112189292908, "num_tokens": 825213475.0, "step": 21730 }, { "epoch": 2.7644065640503754, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6550703048706055, "learning_rate": 1e-05, "loss": 0.3833, "mean_token_accuracy": 0.8889374136924744, "num_tokens": 825249146.0, "step": 21731 }, { "epoch": 2.7645337743289655, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.765495777130127, "learning_rate": 1e-05, "loss": 0.3733, "mean_token_accuracy": 0.8976566791534424, "num_tokens": 825286504.0, "step": 21732 }, { "epoch": 2.7646609846075565, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6495022773742676, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8758490085601807, "num_tokens": 825327820.0, "step": 21733 }, { "epoch": 2.7647881948861466, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.772867441177368, "learning_rate": 1e-05, "loss": 0.3928, "mean_token_accuracy": 0.8910154700279236, "num_tokens": 825365967.0, "step": 21734 }, { "epoch": 2.7649154051647375, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6468539237976074, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8787578344345093, "num_tokens": 825406959.0, "step": 21735 }, { "epoch": 2.7650426154433276, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7863738536834717, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.868274450302124, "num_tokens": 825447355.0, "step": 21736 }, { "epoch": 2.7651698257219186, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6510002613067627, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8689777851104736, "num_tokens": 825485618.0, "step": 21737 }, { "epoch": 2.7652970360005087, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7801523208618164, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8853173851966858, "num_tokens": 825522287.0, "step": 21738 }, { "epoch": 2.765424246279099, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.637474536895752, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.879859209060669, "num_tokens": 825566214.0, "step": 21739 }, { "epoch": 2.7655514565576897, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.785259962081909, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8806672096252441, "num_tokens": 825608169.0, "step": 21740 }, { "epoch": 2.7656786668362803, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6521973609924316, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8856626152992249, "num_tokens": 825643076.0, "step": 21741 }, { "epoch": 2.765805877114871, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.788886308670044, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8781875371932983, "num_tokens": 825679868.0, "step": 21742 }, { "epoch": 2.7659330873934613, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6486213207244873, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.878936767578125, "num_tokens": 825719391.0, "step": 21743 }, { "epoch": 2.766060297672052, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7938649654388428, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8785321712493896, "num_tokens": 825761972.0, "step": 21744 }, { "epoch": 2.7661875079506424, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6473841667175293, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8773723840713501, "num_tokens": 825802672.0, "step": 21745 }, { "epoch": 2.766314718229233, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7988462448120117, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.875961184501648, "num_tokens": 825840219.0, "step": 21746 }, { "epoch": 2.7664419285078234, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6428611278533936, "learning_rate": 1e-05, "loss": 0.4219, "mean_token_accuracy": 0.8788023591041565, "num_tokens": 825876793.0, "step": 21747 }, { "epoch": 2.766569138786414, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8034608364105225, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.873225748538971, "num_tokens": 825913231.0, "step": 21748 }, { "epoch": 2.7666963490650045, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.638810396194458, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8674024343490601, "num_tokens": 825952600.0, "step": 21749 }, { "epoch": 2.766823559343595, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7912440299987793, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8786529302597046, "num_tokens": 825997239.0, "step": 21750 }, { "epoch": 2.7669507696221856, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6354215145111084, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8848431706428528, "num_tokens": 826037492.0, "step": 21751 }, { "epoch": 2.767077979900776, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7943027019500732, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8815294504165649, "num_tokens": 826077607.0, "step": 21752 }, { "epoch": 2.7672051901793666, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.649522542953491, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8735305070877075, "num_tokens": 826115503.0, "step": 21753 }, { "epoch": 2.767332400457957, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7945899963378906, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.874470055103302, "num_tokens": 826154526.0, "step": 21754 }, { "epoch": 2.7674596107365477, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6482038497924805, "learning_rate": 1e-05, "loss": 0.4599, "mean_token_accuracy": 0.8708427548408508, "num_tokens": 826194945.0, "step": 21755 }, { "epoch": 2.767586821015138, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.792898654937744, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8824967741966248, "num_tokens": 826231000.0, "step": 21756 }, { "epoch": 2.7677140312937283, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6514337062835693, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8903363943099976, "num_tokens": 826266453.0, "step": 21757 }, { "epoch": 2.7678412415723193, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.782026767730713, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8842503428459167, "num_tokens": 826310442.0, "step": 21758 }, { "epoch": 2.7679684518509093, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.67046856880188, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8795238733291626, "num_tokens": 826344595.0, "step": 21759 }, { "epoch": 2.7680956621295003, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7795276641845703, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8808385133743286, "num_tokens": 826384556.0, "step": 21760 }, { "epoch": 2.7682228724080904, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.673922538757324, "learning_rate": 1e-05, "loss": 0.4752, "mean_token_accuracy": 0.8567322492599487, "num_tokens": 826418344.0, "step": 21761 }, { "epoch": 2.768350082686681, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.76135516166687, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.879032552242279, "num_tokens": 826460027.0, "step": 21762 }, { "epoch": 2.7684772929652715, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.663104295730591, "learning_rate": 1e-05, "loss": 0.4275, "mean_token_accuracy": 0.877344012260437, "num_tokens": 826497422.0, "step": 21763 }, { "epoch": 2.768604503243862, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7662136554718018, "learning_rate": 1e-05, "loss": 0.463, "mean_token_accuracy": 0.8712542057037354, "num_tokens": 826536624.0, "step": 21764 }, { "epoch": 2.7687317135224525, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6633009910583496, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8746613264083862, "num_tokens": 826572303.0, "step": 21765 }, { "epoch": 2.768858923801043, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.762138843536377, "learning_rate": 1e-05, "loss": 0.4783, "mean_token_accuracy": 0.8609670400619507, "num_tokens": 826613326.0, "step": 21766 }, { "epoch": 2.7689861340796336, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.660924196243286, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8761701583862305, "num_tokens": 826649059.0, "step": 21767 }, { "epoch": 2.769113344358224, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7676894664764404, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8827563524246216, "num_tokens": 826682983.0, "step": 21768 }, { "epoch": 2.7692405546368146, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.67608380317688, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8766870498657227, "num_tokens": 826715179.0, "step": 21769 }, { "epoch": 2.769367764915405, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.773538589477539, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8684605956077576, "num_tokens": 826748442.0, "step": 21770 }, { "epoch": 2.7694949751939957, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6466526985168457, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8884227871894836, "num_tokens": 826791014.0, "step": 21771 }, { "epoch": 2.769622185472586, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7732224464416504, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8729905486106873, "num_tokens": 826827987.0, "step": 21772 }, { "epoch": 2.7697493957511767, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.646381139755249, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8832290172576904, "num_tokens": 826869315.0, "step": 21773 }, { "epoch": 2.7698766060297673, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7726118564605713, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8769890069961548, "num_tokens": 826912679.0, "step": 21774 }, { "epoch": 2.770003816308358, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6536049842834473, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8829070329666138, "num_tokens": 826949276.0, "step": 21775 }, { "epoch": 2.7701310265869483, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7717738151550293, "learning_rate": 1e-05, "loss": 0.4061, "mean_token_accuracy": 0.8859018683433533, "num_tokens": 826989441.0, "step": 21776 }, { "epoch": 2.770258236865539, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.660947799682617, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8723471164703369, "num_tokens": 827025030.0, "step": 21777 }, { "epoch": 2.7703854471441294, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.777202606201172, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8769926428794861, "num_tokens": 827060252.0, "step": 21778 }, { "epoch": 2.77051265742272, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.659507989883423, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8799784779548645, "num_tokens": 827100063.0, "step": 21779 }, { "epoch": 2.77063986770131, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7773666381835938, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.8918389678001404, "num_tokens": 827131772.0, "step": 21780 }, { "epoch": 2.770767077979901, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.65958571434021, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8672014474868774, "num_tokens": 827167659.0, "step": 21781 }, { "epoch": 2.770894288258491, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7618212699890137, "learning_rate": 1e-05, "loss": 0.4462, "mean_token_accuracy": 0.8716675043106079, "num_tokens": 827211215.0, "step": 21782 }, { "epoch": 2.771021498537082, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.671717405319214, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8702535629272461, "num_tokens": 827248075.0, "step": 21783 }, { "epoch": 2.771148708815672, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7582333087921143, "learning_rate": 1e-05, "loss": 0.4032, "mean_token_accuracy": 0.8844119310379028, "num_tokens": 827283270.0, "step": 21784 }, { "epoch": 2.771275919094263, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6775028705596924, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8833824992179871, "num_tokens": 827317520.0, "step": 21785 }, { "epoch": 2.771403129372853, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.761125326156616, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8723095655441284, "num_tokens": 827355899.0, "step": 21786 }, { "epoch": 2.7715303396514437, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.689824342727661, "learning_rate": 1e-05, "loss": 0.4882, "mean_token_accuracy": 0.860095739364624, "num_tokens": 827390038.0, "step": 21787 }, { "epoch": 2.771657549930034, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.744520425796509, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.8744421005249023, "num_tokens": 827428563.0, "step": 21788 }, { "epoch": 2.7717847602086247, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6663084030151367, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8863756656646729, "num_tokens": 827469658.0, "step": 21789 }, { "epoch": 2.7719119704872153, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7496845722198486, "learning_rate": 1e-05, "loss": 0.4044, "mean_token_accuracy": 0.8827115297317505, "num_tokens": 827504277.0, "step": 21790 }, { "epoch": 2.772039180765806, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6753437519073486, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8737226128578186, "num_tokens": 827544840.0, "step": 21791 }, { "epoch": 2.7721663910443963, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7437117099761963, "learning_rate": 1e-05, "loss": 0.4694, "mean_token_accuracy": 0.8655512928962708, "num_tokens": 827590607.0, "step": 21792 }, { "epoch": 2.772293601322987, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.674456834793091, "learning_rate": 1e-05, "loss": 0.3642, "mean_token_accuracy": 0.8949934244155884, "num_tokens": 827622039.0, "step": 21793 }, { "epoch": 2.7724208116015774, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.743429660797119, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8860248327255249, "num_tokens": 827659234.0, "step": 21794 }, { "epoch": 2.772548021880168, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6675453186035156, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8826286792755127, "num_tokens": 827702044.0, "step": 21795 }, { "epoch": 2.7726752321587584, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7549569606781006, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8836537003517151, "num_tokens": 827738189.0, "step": 21796 }, { "epoch": 2.772802442437349, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6699256896972656, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8703055381774902, "num_tokens": 827780483.0, "step": 21797 }, { "epoch": 2.7729296527159395, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.762741804122925, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8825874328613281, "num_tokens": 827817156.0, "step": 21798 }, { "epoch": 2.77305686299453, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.666332960128784, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8797462582588196, "num_tokens": 827851591.0, "step": 21799 }, { "epoch": 2.7731840732731206, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.743417978286743, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8861637115478516, "num_tokens": 827891289.0, "step": 21800 }, { "epoch": 2.773311283551711, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6565942764282227, "learning_rate": 1e-05, "loss": 0.3578, "mean_token_accuracy": 0.9002734422683716, "num_tokens": 827926466.0, "step": 21801 }, { "epoch": 2.7734384938303016, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.765901565551758, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.8530564904212952, "num_tokens": 827964885.0, "step": 21802 }, { "epoch": 2.773565704108892, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6546061038970947, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8701042532920837, "num_tokens": 828006063.0, "step": 21803 }, { "epoch": 2.7736929143874827, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7844414710998535, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8869640231132507, "num_tokens": 828037533.0, "step": 21804 }, { "epoch": 2.7738201246660728, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6455323696136475, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8871790170669556, "num_tokens": 828072298.0, "step": 21805 }, { "epoch": 2.7739473349446637, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7826082706451416, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8869023323059082, "num_tokens": 828107900.0, "step": 21806 }, { "epoch": 2.774074545223254, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.640310764312744, "learning_rate": 1e-05, "loss": 0.4069, "mean_token_accuracy": 0.8835737109184265, "num_tokens": 828143323.0, "step": 21807 }, { "epoch": 2.774201755501845, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7756154537200928, "learning_rate": 1e-05, "loss": 0.3867, "mean_token_accuracy": 0.8934153318405151, "num_tokens": 828183439.0, "step": 21808 }, { "epoch": 2.774328965780435, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.646874189376831, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.883962869644165, "num_tokens": 828218360.0, "step": 21809 }, { "epoch": 2.774456176059026, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.800173282623291, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8730584383010864, "num_tokens": 828254042.0, "step": 21810 }, { "epoch": 2.774583386337616, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6433465480804443, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8783866763114929, "num_tokens": 828290992.0, "step": 21811 }, { "epoch": 2.7747105966162064, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.882538080215454, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.884029746055603, "num_tokens": 828323351.0, "step": 21812 }, { "epoch": 2.774837806894797, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6325230598449707, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.8844256401062012, "num_tokens": 828354555.0, "step": 21813 }, { "epoch": 2.7749650171733875, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8312392234802246, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8844115734100342, "num_tokens": 828395641.0, "step": 21814 }, { "epoch": 2.775092227451978, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6085381507873535, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8819630146026611, "num_tokens": 828437871.0, "step": 21815 }, { "epoch": 2.7752194377305686, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.852435350418091, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8658992052078247, "num_tokens": 828480545.0, "step": 21816 }, { "epoch": 2.775346648009159, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.626809597015381, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8810386061668396, "num_tokens": 828513519.0, "step": 21817 }, { "epoch": 2.7754738582877496, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.862412691116333, "learning_rate": 1e-05, "loss": 0.4363, "mean_token_accuracy": 0.8744833469390869, "num_tokens": 828549329.0, "step": 21818 }, { "epoch": 2.77560106856634, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.634206771850586, "learning_rate": 1e-05, "loss": 0.4666, "mean_token_accuracy": 0.8661870360374451, "num_tokens": 828592357.0, "step": 21819 }, { "epoch": 2.7757282788449307, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.8672871589660645, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8772871494293213, "num_tokens": 828627622.0, "step": 21820 }, { "epoch": 2.775855489123521, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6299479007720947, "learning_rate": 1e-05, "loss": 0.3924, "mean_token_accuracy": 0.8897754549980164, "num_tokens": 828668680.0, "step": 21821 }, { "epoch": 2.7759826994021117, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.9114227294921875e-05, "grad_norm": 2.866609573364258, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8778542280197144, "num_tokens": 828705233.0, "step": 21822 }, { "epoch": 2.7761099096807023, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.641634941101074, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8698247671127319, "num_tokens": 828745183.0, "step": 21823 }, { "epoch": 2.776237119959293, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.8570311069488525, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8733006715774536, "num_tokens": 828787769.0, "step": 21824 }, { "epoch": 2.7763643302378833, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.642608165740967, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8758981227874756, "num_tokens": 828826511.0, "step": 21825 }, { "epoch": 2.776491540516474, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.844827890396118, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8803002834320068, "num_tokens": 828868632.0, "step": 21826 }, { "epoch": 2.7766187507950644, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6666629314422607, "learning_rate": 1e-05, "loss": 0.4481, "mean_token_accuracy": 0.8690078854560852, "num_tokens": 828904342.0, "step": 21827 }, { "epoch": 2.776745961073655, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.835627317428589, "learning_rate": 1e-05, "loss": 0.4696, "mean_token_accuracy": 0.8754249215126038, "num_tokens": 828947074.0, "step": 21828 }, { "epoch": 2.7768731713522454, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.650143623352051, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8818773031234741, "num_tokens": 828989583.0, "step": 21829 }, { "epoch": 2.7770003816308355, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.836110830307007, "learning_rate": 1e-05, "loss": 0.5037, "mean_token_accuracy": 0.8516002893447876, "num_tokens": 829028848.0, "step": 21830 }, { "epoch": 2.7771275919094265, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6547553539276123, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8791434168815613, "num_tokens": 829063904.0, "step": 21831 }, { "epoch": 2.7772548021880166, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8315443992614746, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8660379648208618, "num_tokens": 829102416.0, "step": 21832 }, { "epoch": 2.7773820124666075, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.666722297668457, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8735034465789795, "num_tokens": 829138870.0, "step": 21833 }, { "epoch": 2.7775092227451976, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8280560970306396, "learning_rate": 1e-05, "loss": 0.4636, "mean_token_accuracy": 0.8710975050926208, "num_tokens": 829176704.0, "step": 21834 }, { "epoch": 2.7776364330237886, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6621968746185303, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8854261636734009, "num_tokens": 829210170.0, "step": 21835 }, { "epoch": 2.7777636433023787, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8012945652008057, "learning_rate": 1e-05, "loss": 0.3918, "mean_token_accuracy": 0.8927465677261353, "num_tokens": 829249692.0, "step": 21836 }, { "epoch": 2.777890853580969, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.663142442703247, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8874720335006714, "num_tokens": 829285342.0, "step": 21837 }, { "epoch": 2.7780180638595597, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.811875104904175, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8806617259979248, "num_tokens": 829320281.0, "step": 21838 }, { "epoch": 2.7781452741381503, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.659480094909668, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8735005855560303, "num_tokens": 829359990.0, "step": 21839 }, { "epoch": 2.778272484416741, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8038289546966553, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.8881815671920776, "num_tokens": 829395330.0, "step": 21840 }, { "epoch": 2.7783996946953313, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.664832353591919, "learning_rate": 1e-05, "loss": 0.4448, "mean_token_accuracy": 0.8766253590583801, "num_tokens": 829432127.0, "step": 21841 }, { "epoch": 2.778526904973922, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7919094562530518, "learning_rate": 1e-05, "loss": 0.4376, "mean_token_accuracy": 0.8756926655769348, "num_tokens": 829469735.0, "step": 21842 }, { "epoch": 2.7786541152525124, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6646461486816406, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8774019479751587, "num_tokens": 829502893.0, "step": 21843 }, { "epoch": 2.778781325531103, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7817392349243164, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8862794637680054, "num_tokens": 829538307.0, "step": 21844 }, { "epoch": 2.7789085358096934, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6468334197998047, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.8883488178253174, "num_tokens": 829575992.0, "step": 21845 }, { "epoch": 2.779035746088284, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.79583740234375, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8717951774597168, "num_tokens": 829610167.0, "step": 21846 }, { "epoch": 2.7791629563668745, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.660665273666382, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8712381720542908, "num_tokens": 829646828.0, "step": 21847 }, { "epoch": 2.779290166645465, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7874598503112793, "learning_rate": 1e-05, "loss": 0.455, "mean_token_accuracy": 0.8711374998092651, "num_tokens": 829683593.0, "step": 21848 }, { "epoch": 2.7794173769240555, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6556031703948975, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8764082193374634, "num_tokens": 829721080.0, "step": 21849 }, { "epoch": 2.779544587202646, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.788264036178589, "learning_rate": 1e-05, "loss": 0.3779, "mean_token_accuracy": 0.8937265872955322, "num_tokens": 829755871.0, "step": 21850 }, { "epoch": 2.7796717974812366, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.653510808944702, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8804230690002441, "num_tokens": 829793377.0, "step": 21851 }, { "epoch": 2.779799007759827, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.781329393386841, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8778078556060791, "num_tokens": 829833165.0, "step": 21852 }, { "epoch": 2.7799262180384177, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6586921215057373, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8834895491600037, "num_tokens": 829869558.0, "step": 21853 }, { "epoch": 2.780053428317008, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.784977912902832, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8784694075584412, "num_tokens": 829906385.0, "step": 21854 }, { "epoch": 2.7801806385955983, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.7058234214782715, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8694682121276855, "num_tokens": 829944216.0, "step": 21855 }, { "epoch": 2.7803078488741892, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7822375297546387, "learning_rate": 1e-05, "loss": 0.4945, "mean_token_accuracy": 0.8577205538749695, "num_tokens": 829976817.0, "step": 21856 }, { "epoch": 2.7804350591527793, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.656940221786499, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.8713221549987793, "num_tokens": 830020434.0, "step": 21857 }, { "epoch": 2.7805622694313703, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.77694034576416, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8799406886100769, "num_tokens": 830054469.0, "step": 21858 }, { "epoch": 2.7806894797099604, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003662109375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6506354808807373, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8800122737884521, "num_tokens": 830091536.0, "step": 21859 }, { "epoch": 2.780816689988551, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.777949810028076, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8706668615341187, "num_tokens": 830129425.0, "step": 21860 }, { "epoch": 2.7809439002671414, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.661813974380493, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8723776936531067, "num_tokens": 830171137.0, "step": 21861 }, { "epoch": 2.781071110545732, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.763468027114868, "learning_rate": 1e-05, "loss": 0.3937, "mean_token_accuracy": 0.8916192054748535, "num_tokens": 830211434.0, "step": 21862 }, { "epoch": 2.7811983208243225, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.649271249771118, "learning_rate": 1e-05, "loss": 0.3798, "mean_token_accuracy": 0.892768383026123, "num_tokens": 830250843.0, "step": 21863 }, { "epoch": 2.781325531102913, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7695181369781494, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8692729473114014, "num_tokens": 830292131.0, "step": 21864 }, { "epoch": 2.7814527413815036, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6509664058685303, "learning_rate": 1e-05, "loss": 0.3876, "mean_token_accuracy": 0.8881274461746216, "num_tokens": 830329678.0, "step": 21865 }, { "epoch": 2.781579951660094, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.778061628341675, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8880558013916016, "num_tokens": 830367822.0, "step": 21866 }, { "epoch": 2.7817071619386846, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6505494117736816, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.875826358795166, "num_tokens": 830408170.0, "step": 21867 }, { "epoch": 2.781834372217275, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.766655445098877, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8834571838378906, "num_tokens": 830447402.0, "step": 21868 }, { "epoch": 2.7819615824958657, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6690266132354736, "learning_rate": 1e-05, "loss": 0.492, "mean_token_accuracy": 0.8586224317550659, "num_tokens": 830485411.0, "step": 21869 }, { "epoch": 2.782088792774456, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.763465404510498, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8796097040176392, "num_tokens": 830522562.0, "step": 21870 }, { "epoch": 2.7822160030530467, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.651066541671753, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.867545485496521, "num_tokens": 830565979.0, "step": 21871 }, { "epoch": 2.7823432133316373, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.782874345779419, "learning_rate": 1e-05, "loss": 0.4297, "mean_token_accuracy": 0.8746098279953003, "num_tokens": 830602771.0, "step": 21872 }, { "epoch": 2.782470423610228, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.652676582336426, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8700430393218994, "num_tokens": 830639891.0, "step": 21873 }, { "epoch": 2.7825976338888183, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.773338556289673, "learning_rate": 1e-05, "loss": 0.3748, "mean_token_accuracy": 0.8906335830688477, "num_tokens": 830676044.0, "step": 21874 }, { "epoch": 2.782724844167409, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.641829013824463, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8770979046821594, "num_tokens": 830718252.0, "step": 21875 }, { "epoch": 2.7828520544459994, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7717490196228027, "learning_rate": 1e-05, "loss": 0.3913, "mean_token_accuracy": 0.8906995058059692, "num_tokens": 830757563.0, "step": 21876 }, { "epoch": 2.78297926472459, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6530468463897705, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.8726834058761597, "num_tokens": 830797808.0, "step": 21877 }, { "epoch": 2.78310647500318, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.77239727973938, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8760883808135986, "num_tokens": 830838691.0, "step": 21878 }, { "epoch": 2.783233685281771, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.644693613052368, "learning_rate": 1e-05, "loss": 0.3896, "mean_token_accuracy": 0.8918991088867188, "num_tokens": 830877016.0, "step": 21879 }, { "epoch": 2.783360895560361, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.795736312866211, "learning_rate": 1e-05, "loss": 0.4283, "mean_token_accuracy": 0.8826336860656738, "num_tokens": 830908841.0, "step": 21880 }, { "epoch": 2.783488105838952, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.639608144760132, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8743282556533813, "num_tokens": 830947580.0, "step": 21881 }, { "epoch": 2.783615316117542, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.80293869972229, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8787358999252319, "num_tokens": 830984385.0, "step": 21882 }, { "epoch": 2.783742526396133, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.638554096221924, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8896538019180298, "num_tokens": 831022280.0, "step": 21883 }, { "epoch": 2.783869736674723, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8046464920043945, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8811981081962585, "num_tokens": 831055764.0, "step": 21884 }, { "epoch": 2.7839969469533137, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.650054454803467, "learning_rate": 1e-05, "loss": 0.4866, "mean_token_accuracy": 0.867904007434845, "num_tokens": 831093397.0, "step": 21885 }, { "epoch": 2.784124157231904, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8056423664093018, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.878747820854187, "num_tokens": 831130720.0, "step": 21886 }, { "epoch": 2.7842513675104947, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.64963436126709, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8791780471801758, "num_tokens": 831164092.0, "step": 21887 }, { "epoch": 2.7843785777890853, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7977795600891113, "learning_rate": 1e-05, "loss": 0.4364, "mean_token_accuracy": 0.874482274055481, "num_tokens": 831202570.0, "step": 21888 }, { "epoch": 2.784505788067676, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.648076295852661, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8652303218841553, "num_tokens": 831242596.0, "step": 21889 }, { "epoch": 2.7846329983462663, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7866005897521973, "learning_rate": 1e-05, "loss": 0.3869, "mean_token_accuracy": 0.8912439346313477, "num_tokens": 831283305.0, "step": 21890 }, { "epoch": 2.784760208624857, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6593291759490967, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.8922820687294006, "num_tokens": 831322355.0, "step": 21891 }, { "epoch": 2.7848874189034474, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.793363332748413, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8773893713951111, "num_tokens": 831361410.0, "step": 21892 }, { "epoch": 2.785014629182038, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6489202976226807, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8724547028541565, "num_tokens": 831402493.0, "step": 21893 }, { "epoch": 2.7851418394606284, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8051505088806152, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.8538656830787659, "num_tokens": 831443231.0, "step": 21894 }, { "epoch": 2.785269049739219, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6508896350860596, "learning_rate": 1e-05, "loss": 0.3596, "mean_token_accuracy": 0.9032720327377319, "num_tokens": 831474716.0, "step": 21895 }, { "epoch": 2.7853962600178095, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7966439723968506, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8625048398971558, "num_tokens": 831519410.0, "step": 21896 }, { "epoch": 2.7855234702964, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6400325298309326, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.883844256401062, "num_tokens": 831562040.0, "step": 21897 }, { "epoch": 2.7856506805749905, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.793710947036743, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8701015710830688, "num_tokens": 831604956.0, "step": 21898 }, { "epoch": 2.785777890853581, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6624319553375244, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8835776448249817, "num_tokens": 831637270.0, "step": 21899 }, { "epoch": 2.7859051011321716, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.796245574951172, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8849202394485474, "num_tokens": 831671342.0, "step": 21900 }, { "epoch": 2.786032311410762, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6573150157928467, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.877230167388916, "num_tokens": 831709222.0, "step": 21901 }, { "epoch": 2.7861595216893527, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.784403085708618, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8869504332542419, "num_tokens": 831749690.0, "step": 21902 }, { "epoch": 2.7862867319679427, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6657068729400635, "learning_rate": 1e-05, "loss": 0.4777, "mean_token_accuracy": 0.8671874403953552, "num_tokens": 831787963.0, "step": 21903 }, { "epoch": 2.7864139422465337, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7846975326538086, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.885098934173584, "num_tokens": 831820925.0, "step": 21904 }, { "epoch": 2.786541152525124, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.668095350265503, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8741414546966553, "num_tokens": 831856190.0, "step": 21905 }, { "epoch": 2.7866683628037148, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7722699642181396, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8670741319656372, "num_tokens": 831900339.0, "step": 21906 }, { "epoch": 2.786795573082305, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6860644817352295, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8716204166412354, "num_tokens": 831932166.0, "step": 21907 }, { "epoch": 2.786922783360896, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7686052322387695, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8822176456451416, "num_tokens": 831964796.0, "step": 21908 }, { "epoch": 2.787049993639486, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6593968868255615, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8753210306167603, "num_tokens": 832004603.0, "step": 21909 }, { "epoch": 2.7871772039180764, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7568774223327637, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8903711438179016, "num_tokens": 832042847.0, "step": 21910 }, { "epoch": 2.787304414196667, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6647346019744873, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8803509473800659, "num_tokens": 832079746.0, "step": 21911 }, { "epoch": 2.7874316244752575, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7718162536621094, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8724803328514099, "num_tokens": 832116300.0, "step": 21912 }, { "epoch": 2.787558834753848, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.662698984146118, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8902701139450073, "num_tokens": 832151985.0, "step": 21913 }, { "epoch": 2.7876860450324386, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7910234928131104, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8786242604255676, "num_tokens": 832181327.0, "step": 21914 }, { "epoch": 2.787813255311029, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6539957523345947, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8840292096138, "num_tokens": 832217710.0, "step": 21915 }, { "epoch": 2.7879404655896196, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.804124116897583, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8744630813598633, "num_tokens": 832255628.0, "step": 21916 }, { "epoch": 2.78806767586821, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6403796672821045, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8787423372268677, "num_tokens": 832294239.0, "step": 21917 }, { "epoch": 2.7881948861468007, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.801115036010742, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8829865455627441, "num_tokens": 832331270.0, "step": 21918 }, { "epoch": 2.788322096425391, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6446166038513184, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8697776198387146, "num_tokens": 832372276.0, "step": 21919 }, { "epoch": 2.7884493067039817, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.7961578369140625, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8865717649459839, "num_tokens": 832410352.0, "step": 21920 }, { "epoch": 2.7885765169825723, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6454145908355713, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.871315598487854, "num_tokens": 832452633.0, "step": 21921 }, { "epoch": 2.788703727261163, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.789344072341919, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8841755390167236, "num_tokens": 832492100.0, "step": 21922 }, { "epoch": 2.7888309375397533, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6588165760040283, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.8757320642471313, "num_tokens": 832529972.0, "step": 21923 }, { "epoch": 2.788958147818344, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.777209758758545, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8880593776702881, "num_tokens": 832571042.0, "step": 21924 }, { "epoch": 2.7890853580969344, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6599795818328857, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8707382678985596, "num_tokens": 832610575.0, "step": 21925 }, { "epoch": 2.789212568375525, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.770853042602539, "learning_rate": 1e-05, "loss": 0.3836, "mean_token_accuracy": 0.8909018039703369, "num_tokens": 832649477.0, "step": 21926 }, { "epoch": 2.7893397786541154, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.649143695831299, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8851189613342285, "num_tokens": 832688958.0, "step": 21927 }, { "epoch": 2.7894669889327055, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7780535221099854, "learning_rate": 1e-05, "loss": 0.3915, "mean_token_accuracy": 0.8906077742576599, "num_tokens": 832724368.0, "step": 21928 }, { "epoch": 2.7895941992112965, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.653068780899048, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8753513097763062, "num_tokens": 832765950.0, "step": 21929 }, { "epoch": 2.7897214094898866, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7730824947357178, "learning_rate": 1e-05, "loss": 0.3703, "mean_token_accuracy": 0.894092321395874, "num_tokens": 832800666.0, "step": 21930 }, { "epoch": 2.7898486197684775, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6668756008148193, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8691136837005615, "num_tokens": 832835830.0, "step": 21931 }, { "epoch": 2.7899758300470676, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.781146764755249, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8902186155319214, "num_tokens": 832866370.0, "step": 21932 }, { "epoch": 2.7901030403256586, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.638483762741089, "learning_rate": 1e-05, "loss": 0.4344, "mean_token_accuracy": 0.8784404993057251, "num_tokens": 832914174.0, "step": 21933 }, { "epoch": 2.7902302506042487, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7773847579956055, "learning_rate": 1e-05, "loss": 0.3869, "mean_token_accuracy": 0.8900294899940491, "num_tokens": 832951397.0, "step": 21934 }, { "epoch": 2.790357460882839, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6626927852630615, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8844924569129944, "num_tokens": 832986282.0, "step": 21935 }, { "epoch": 2.7904846711614297, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7785041332244873, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.8874613642692566, "num_tokens": 833023288.0, "step": 21936 }, { "epoch": 2.7906118814400203, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.676504135131836, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8752286434173584, "num_tokens": 833053744.0, "step": 21937 }, { "epoch": 2.790739091718611, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7503414154052734, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8890702724456787, "num_tokens": 833096499.0, "step": 21938 }, { "epoch": 2.7908663019972013, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.661395311355591, "learning_rate": 1e-05, "loss": 0.3948, "mean_token_accuracy": 0.8878804445266724, "num_tokens": 833132405.0, "step": 21939 }, { "epoch": 2.790993512275792, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.74519681930542, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8794887065887451, "num_tokens": 833175581.0, "step": 21940 }, { "epoch": 2.7911207225543824, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.673551321029663, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8697493076324463, "num_tokens": 833213466.0, "step": 21941 }, { "epoch": 2.791247932832973, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7474896907806396, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8742727041244507, "num_tokens": 833250344.0, "step": 21942 }, { "epoch": 2.7913751431115634, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6655023097991943, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8797007203102112, "num_tokens": 833291032.0, "step": 21943 }, { "epoch": 2.791502353390154, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7429757118225098, "learning_rate": 1e-05, "loss": 0.4071, "mean_token_accuracy": 0.8834861516952515, "num_tokens": 833331213.0, "step": 21944 }, { "epoch": 2.7916295636687445, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.657750368118286, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8869755864143372, "num_tokens": 833373906.0, "step": 21945 }, { "epoch": 2.791756773947335, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7589309215545654, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8720385432243347, "num_tokens": 833411724.0, "step": 21946 }, { "epoch": 2.7918839842259255, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.671412467956543, "learning_rate": 1e-05, "loss": 0.4011, "mean_token_accuracy": 0.8876060843467712, "num_tokens": 833446337.0, "step": 21947 }, { "epoch": 2.792011194504516, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.756983995437622, "learning_rate": 1e-05, "loss": 0.4577, "mean_token_accuracy": 0.866466760635376, "num_tokens": 833484444.0, "step": 21948 }, { "epoch": 2.7921384047831066, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.662301540374756, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8814973831176758, "num_tokens": 833521375.0, "step": 21949 }, { "epoch": 2.792265615061697, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7591447830200195, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8877236843109131, "num_tokens": 833557784.0, "step": 21950 }, { "epoch": 2.7923928253402877, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6557023525238037, "learning_rate": 1e-05, "loss": 0.3756, "mean_token_accuracy": 0.8975775837898254, "num_tokens": 833598872.0, "step": 21951 }, { "epoch": 2.792520035618878, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7644851207733154, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8868775367736816, "num_tokens": 833635799.0, "step": 21952 }, { "epoch": 2.7926472458974683, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6532881259918213, "learning_rate": 1e-05, "loss": 0.3989, "mean_token_accuracy": 0.8866345286369324, "num_tokens": 833671814.0, "step": 21953 }, { "epoch": 2.7927744561760592, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7590491771698, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8683381676673889, "num_tokens": 833718081.0, "step": 21954 }, { "epoch": 2.7929016664546493, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6449427604675293, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8770143389701843, "num_tokens": 833759197.0, "step": 21955 }, { "epoch": 2.7930288767332403, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.800652027130127, "learning_rate": 1e-05, "loss": 0.3935, "mean_token_accuracy": 0.885877251625061, "num_tokens": 833785582.0, "step": 21956 }, { "epoch": 2.7931560870118304, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.626394033432007, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8863468170166016, "num_tokens": 833826611.0, "step": 21957 }, { "epoch": 2.793283297290421, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7952542304992676, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8759873509407043, "num_tokens": 833870657.0, "step": 21958 }, { "epoch": 2.7934105075690114, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.634678602218628, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8786882162094116, "num_tokens": 833907621.0, "step": 21959 }, { "epoch": 2.793537717847602, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8075711727142334, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8847911357879639, "num_tokens": 833944364.0, "step": 21960 }, { "epoch": 2.7936649281261925, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6259896755218506, "learning_rate": 1e-05, "loss": 0.3894, "mean_token_accuracy": 0.8903341889381409, "num_tokens": 833983840.0, "step": 21961 }, { "epoch": 2.793792138404783, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.820082902908325, "learning_rate": 1e-05, "loss": 0.3883, "mean_token_accuracy": 0.8910287022590637, "num_tokens": 834016367.0, "step": 21962 }, { "epoch": 2.7939193486833735, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6424593925476074, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8875526189804077, "num_tokens": 834050190.0, "step": 21963 }, { "epoch": 2.794046558961964, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.839855909347534, "learning_rate": 1e-05, "loss": 0.4575, "mean_token_accuracy": 0.8718986511230469, "num_tokens": 834087013.0, "step": 21964 }, { "epoch": 2.7941737692405546, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.63272762298584, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8763391971588135, "num_tokens": 834131771.0, "step": 21965 }, { "epoch": 2.794300979519145, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8361899852752686, "learning_rate": 1e-05, "loss": 0.4104, "mean_token_accuracy": 0.8848690390586853, "num_tokens": 834169055.0, "step": 21966 }, { "epoch": 2.7944281897977357, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.64125657081604, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8848416805267334, "num_tokens": 834206081.0, "step": 21967 }, { "epoch": 2.794555400076326, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.838696002960205, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.888096809387207, "num_tokens": 834240846.0, "step": 21968 }, { "epoch": 2.7946826103549167, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6481730937957764, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8789556622505188, "num_tokens": 834275037.0, "step": 21969 }, { "epoch": 2.7948098206335072, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8226428031921387, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8803983926773071, "num_tokens": 834316408.0, "step": 21970 }, { "epoch": 2.7949370309120978, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.634091854095459, "learning_rate": 1e-05, "loss": 0.3918, "mean_token_accuracy": 0.8888810873031616, "num_tokens": 834358473.0, "step": 21971 }, { "epoch": 2.7950642411906883, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8272366523742676, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.8786429166793823, "num_tokens": 834399982.0, "step": 21972 }, { "epoch": 2.795191451469279, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.65474796295166, "learning_rate": 1e-05, "loss": 0.3904, "mean_token_accuracy": 0.8888623118400574, "num_tokens": 834429731.0, "step": 21973 }, { "epoch": 2.7953186617478694, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.880774736404419, "learning_rate": 1e-05, "loss": 0.4356, "mean_token_accuracy": 0.8772330284118652, "num_tokens": 834476958.0, "step": 21974 }, { "epoch": 2.79544587202646, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.633781909942627, "learning_rate": 1e-05, "loss": 0.4091, "mean_token_accuracy": 0.8843162059783936, "num_tokens": 834516820.0, "step": 21975 }, { "epoch": 2.79557308230505, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8497085571289062, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8792662620544434, "num_tokens": 834552455.0, "step": 21976 }, { "epoch": 2.795700292583641, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6269803047180176, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8759371042251587, "num_tokens": 834593059.0, "step": 21977 }, { "epoch": 2.795827502862231, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.852065324783325, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.869979739189148, "num_tokens": 834640134.0, "step": 21978 }, { "epoch": 2.795954713140822, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.644040107727051, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8779032826423645, "num_tokens": 834679172.0, "step": 21979 }, { "epoch": 2.796081923419412, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8659486770629883, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8683716058731079, "num_tokens": 834713463.0, "step": 21980 }, { "epoch": 2.796209133698003, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.6443448066711426, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8587836027145386, "num_tokens": 834757782.0, "step": 21981 }, { "epoch": 2.796336343976593, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8648195266723633, "learning_rate": 1e-05, "loss": 0.4661, "mean_token_accuracy": 0.8719388246536255, "num_tokens": 834791598.0, "step": 21982 }, { "epoch": 2.7964635542551837, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.644066572189331, "learning_rate": 1e-05, "loss": 0.4549, "mean_token_accuracy": 0.8715516328811646, "num_tokens": 834835205.0, "step": 21983 }, { "epoch": 2.796590764533774, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.858330726623535, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8900356888771057, "num_tokens": 834867450.0, "step": 21984 }, { "epoch": 2.7967179748123647, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.643183469772339, "learning_rate": 1e-05, "loss": 0.4355, "mean_token_accuracy": 0.8769110441207886, "num_tokens": 834911067.0, "step": 21985 }, { "epoch": 2.7968451850909553, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.845999002456665, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8801051378250122, "num_tokens": 834949117.0, "step": 21986 }, { "epoch": 2.796972395369546, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.660242795944214, "learning_rate": 1e-05, "loss": 0.4377, "mean_token_accuracy": 0.876593291759491, "num_tokens": 834988909.0, "step": 21987 }, { "epoch": 2.7970996056481363, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8223485946655273, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8849806785583496, "num_tokens": 835031168.0, "step": 21988 }, { "epoch": 2.797226815926727, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.661198377609253, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8859686255455017, "num_tokens": 835072110.0, "step": 21989 }, { "epoch": 2.7973540262053174, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8246755599975586, "learning_rate": 1e-05, "loss": 0.3945, "mean_token_accuracy": 0.8913565278053284, "num_tokens": 835104605.0, "step": 21990 }, { "epoch": 2.797481236483908, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.668449878692627, "learning_rate": 1e-05, "loss": 0.4748, "mean_token_accuracy": 0.8651372790336609, "num_tokens": 835141970.0, "step": 21991 }, { "epoch": 2.7976084467624984, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8267478942871094, "learning_rate": 1e-05, "loss": 0.4774, "mean_token_accuracy": 0.8632721900939941, "num_tokens": 835177469.0, "step": 21992 }, { "epoch": 2.797735657041089, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.643008232116699, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8771134614944458, "num_tokens": 835220286.0, "step": 21993 }, { "epoch": 2.7978628673196795, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8211357593536377, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8832610845565796, "num_tokens": 835254290.0, "step": 21994 }, { "epoch": 2.79799007759827, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.665207624435425, "learning_rate": 1e-05, "loss": 0.422, "mean_token_accuracy": 0.8773348927497864, "num_tokens": 835286375.0, "step": 21995 }, { "epoch": 2.7981172878768605, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.9466426372528076, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.881428599357605, "num_tokens": 835315472.0, "step": 21996 }, { "epoch": 2.798244498155451, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6284234523773193, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8868849277496338, "num_tokens": 835355865.0, "step": 21997 }, { "epoch": 2.7983717084340416, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.872311592102051, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8785437345504761, "num_tokens": 835391251.0, "step": 21998 }, { "epoch": 2.798498918712632, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.639068365097046, "learning_rate": 1e-05, "loss": 0.3829, "mean_token_accuracy": 0.8910008668899536, "num_tokens": 835424869.0, "step": 21999 }, { "epoch": 2.7986261289912227, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.9591064453125e-05, "grad_norm": 2.874377727508545, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8812453746795654, "num_tokens": 835464916.0, "step": 22000 }, { "epoch": 2.7987533392698127, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6546547412872314, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.870869517326355, "num_tokens": 835503114.0, "step": 22001 }, { "epoch": 2.7988805495484037, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 2.867908239364624, "learning_rate": 1e-05, "loss": 0.4704, "mean_token_accuracy": 0.8688998222351074, "num_tokens": 835541446.0, "step": 22002 }, { "epoch": 2.799007759826994, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.654050350189209, "learning_rate": 1e-05, "loss": 0.3835, "mean_token_accuracy": 0.893744707107544, "num_tokens": 835579261.0, "step": 22003 }, { "epoch": 2.7991349701055848, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.887580871582031e-05, "grad_norm": 2.8641295433044434, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8758610486984253, "num_tokens": 835615847.0, "step": 22004 }, { "epoch": 2.799262180384175, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.6640894412994385, "learning_rate": 1e-05, "loss": 0.4869, "mean_token_accuracy": 0.8598504066467285, "num_tokens": 835658461.0, "step": 22005 }, { "epoch": 2.799389390662766, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8381083011627197, "learning_rate": 1e-05, "loss": 0.3934, "mean_token_accuracy": 0.8873589038848877, "num_tokens": 835693489.0, "step": 22006 }, { "epoch": 2.799516600941356, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.652033567428589, "learning_rate": 1e-05, "loss": 0.38, "mean_token_accuracy": 0.8935287594795227, "num_tokens": 835733925.0, "step": 22007 }, { "epoch": 2.7996438112199464, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8294451236724854, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8730152249336243, "num_tokens": 835775071.0, "step": 22008 }, { "epoch": 2.799771021498537, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.6673717498779297, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8770936131477356, "num_tokens": 835813940.0, "step": 22009 }, { "epoch": 2.7998982317771275, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8359720706939697, "learning_rate": 1e-05, "loss": 0.4734, "mean_token_accuracy": 0.8643841743469238, "num_tokens": 835849667.0, "step": 22010 }, { "epoch": 2.800025442055718, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.744297742843628, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8833632469177246, "num_tokens": 835880928.0, "step": 22011 }, { "epoch": 2.8001526523343085, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.761369466781616, "learning_rate": 1e-05, "loss": 0.3696, "mean_token_accuracy": 0.8973146080970764, "num_tokens": 835917502.0, "step": 22012 }, { "epoch": 2.800279862612899, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.6947031021118164, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8702107667922974, "num_tokens": 835958547.0, "step": 22013 }, { "epoch": 2.8004070728914896, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7483012676239014, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8775596618652344, "num_tokens": 835997427.0, "step": 22014 }, { "epoch": 2.80053428317008, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6828906536102295, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8852660655975342, "num_tokens": 836039096.0, "step": 22015 }, { "epoch": 2.8006614934486707, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.740243434906006, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.8791090250015259, "num_tokens": 836079613.0, "step": 22016 }, { "epoch": 2.800788703727261, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.689408540725708, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8846856951713562, "num_tokens": 836117500.0, "step": 22017 }, { "epoch": 2.8009159140058517, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.768794536590576, "learning_rate": 1e-05, "loss": 0.4167, "mean_token_accuracy": 0.8795255422592163, "num_tokens": 836149680.0, "step": 22018 }, { "epoch": 2.8010431242844422, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6820523738861084, "learning_rate": 1e-05, "loss": 0.3653, "mean_token_accuracy": 0.8964301347732544, "num_tokens": 836179869.0, "step": 22019 }, { "epoch": 2.8011703345630328, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7422571182250977, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.874984622001648, "num_tokens": 836222305.0, "step": 22020 }, { "epoch": 2.8012975448416233, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6874895095825195, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8744254112243652, "num_tokens": 836257687.0, "step": 22021 }, { "epoch": 2.801424755120214, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7341127395629883, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8781256675720215, "num_tokens": 836299105.0, "step": 22022 }, { "epoch": 2.8015519653988044, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.684431791305542, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.8819659948348999, "num_tokens": 836337661.0, "step": 22023 }, { "epoch": 2.801679175677395, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.727022171020508, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.8708441257476807, "num_tokens": 836378792.0, "step": 22024 }, { "epoch": 2.8018063859559854, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.708554983139038, "learning_rate": 1e-05, "loss": 0.4368, "mean_token_accuracy": 0.8748005032539368, "num_tokens": 836412754.0, "step": 22025 }, { "epoch": 2.8019335962345755, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.716536283493042, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8841820955276489, "num_tokens": 836452173.0, "step": 22026 }, { "epoch": 2.8020608065131665, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6907079219818115, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8876943588256836, "num_tokens": 836491398.0, "step": 22027 }, { "epoch": 2.8021880167917566, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7383036613464355, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8698408603668213, "num_tokens": 836525130.0, "step": 22028 }, { "epoch": 2.8023152270703475, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.702075481414795, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8771344423294067, "num_tokens": 836561288.0, "step": 22029 }, { "epoch": 2.8024424373489376, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7273268699645996, "learning_rate": 1e-05, "loss": 0.4768, "mean_token_accuracy": 0.864445686340332, "num_tokens": 836603595.0, "step": 22030 }, { "epoch": 2.8025696476275286, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.708460569381714, "learning_rate": 1e-05, "loss": 0.4537, "mean_token_accuracy": 0.873046875, "num_tokens": 836640460.0, "step": 22031 }, { "epoch": 2.8026968579061187, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.725461959838867, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8701593279838562, "num_tokens": 836678024.0, "step": 22032 }, { "epoch": 2.802824068184709, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6828482151031494, "learning_rate": 1e-05, "loss": 0.38, "mean_token_accuracy": 0.8945025205612183, "num_tokens": 836718383.0, "step": 22033 }, { "epoch": 2.8029512784632997, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.732593297958374, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8658064603805542, "num_tokens": 836755564.0, "step": 22034 }, { "epoch": 2.8030784887418903, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6936192512512207, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8773115873336792, "num_tokens": 836792891.0, "step": 22035 }, { "epoch": 2.803205699020481, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7292122840881348, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8753933906555176, "num_tokens": 836830024.0, "step": 22036 }, { "epoch": 2.8033329092990713, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6901779174804688, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8704589009284973, "num_tokens": 836871600.0, "step": 22037 }, { "epoch": 2.803460119577662, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.714001178741455, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.88653564453125, "num_tokens": 836911495.0, "step": 22038 }, { "epoch": 2.8035873298562524, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.680732488632202, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8800990581512451, "num_tokens": 836952840.0, "step": 22039 }, { "epoch": 2.803714540134843, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7324681282043457, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8745040893554688, "num_tokens": 836988355.0, "step": 22040 }, { "epoch": 2.8038417504134334, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6929242610931396, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8797271251678467, "num_tokens": 837021559.0, "step": 22041 }, { "epoch": 2.803968960692024, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.714648485183716, "learning_rate": 1e-05, "loss": 0.4318, "mean_token_accuracy": 0.875552773475647, "num_tokens": 837060401.0, "step": 22042 }, { "epoch": 2.8040961709706145, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6968653202056885, "learning_rate": 1e-05, "loss": 0.3909, "mean_token_accuracy": 0.889586329460144, "num_tokens": 837094769.0, "step": 22043 }, { "epoch": 2.804223381249205, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.726062297821045, "learning_rate": 1e-05, "loss": 0.3777, "mean_token_accuracy": 0.8925485610961914, "num_tokens": 837125820.0, "step": 22044 }, { "epoch": 2.8043505915277955, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.675968647003174, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8875986337661743, "num_tokens": 837167392.0, "step": 22045 }, { "epoch": 2.804477801806386, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.833915948867798, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8788240551948547, "num_tokens": 837207439.0, "step": 22046 }, { "epoch": 2.8046050120849766, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6516005992889404, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8838385343551636, "num_tokens": 837240038.0, "step": 22047 }, { "epoch": 2.804732222363567, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7814438343048096, "learning_rate": 1e-05, "loss": 0.4528, "mean_token_accuracy": 0.8722979426383972, "num_tokens": 837278842.0, "step": 22048 }, { "epoch": 2.8048594326421576, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.617323160171509, "learning_rate": 1e-05, "loss": 0.3728, "mean_token_accuracy": 0.8956986665725708, "num_tokens": 837317519.0, "step": 22049 }, { "epoch": 2.804986642920748, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8144350051879883, "learning_rate": 1e-05, "loss": 0.3936, "mean_token_accuracy": 0.8891953229904175, "num_tokens": 837349053.0, "step": 22050 }, { "epoch": 2.8051138531993383, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6309309005737305, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8762164115905762, "num_tokens": 837386018.0, "step": 22051 }, { "epoch": 2.8052410634779292, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8243753910064697, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8739436864852905, "num_tokens": 837423640.0, "step": 22052 }, { "epoch": 2.8053682737565193, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.627822160720825, "learning_rate": 1e-05, "loss": 0.416, "mean_token_accuracy": 0.8803260922431946, "num_tokens": 837462198.0, "step": 22053 }, { "epoch": 2.8054954840351103, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.849515914916992, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8768085241317749, "num_tokens": 837492722.0, "step": 22054 }, { "epoch": 2.8056226943137004, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6232032775878906, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8850810527801514, "num_tokens": 837531598.0, "step": 22055 }, { "epoch": 2.805749904592291, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.852471351623535, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8736425638198853, "num_tokens": 837572580.0, "step": 22056 }, { "epoch": 2.8058771148708814, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6285083293914795, "learning_rate": 1e-05, "loss": 0.4012, "mean_token_accuracy": 0.8870954513549805, "num_tokens": 837610398.0, "step": 22057 }, { "epoch": 2.806004325149472, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8568308353424072, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8761038780212402, "num_tokens": 837653518.0, "step": 22058 }, { "epoch": 2.8061315354280625, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6249921321868896, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8767973780632019, "num_tokens": 837697397.0, "step": 22059 }, { "epoch": 2.806258745706653, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8609983921051025, "learning_rate": 1e-05, "loss": 0.3785, "mean_token_accuracy": 0.8955118656158447, "num_tokens": 837736565.0, "step": 22060 }, { "epoch": 2.8063859559852435, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.638880729675293, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.8802666664123535, "num_tokens": 837772514.0, "step": 22061 }, { "epoch": 2.806513166263834, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8612020015716553, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.8882185220718384, "num_tokens": 837805147.0, "step": 22062 }, { "epoch": 2.8066403765424246, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.647028684616089, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8660176992416382, "num_tokens": 837846795.0, "step": 22063 }, { "epoch": 2.806767586821015, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8548004627227783, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8822619318962097, "num_tokens": 837882703.0, "step": 22064 }, { "epoch": 2.8068947970996057, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6622045040130615, "learning_rate": 1e-05, "loss": 0.4708, "mean_token_accuracy": 0.8648583292961121, "num_tokens": 837924698.0, "step": 22065 }, { "epoch": 2.807022007378196, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.841813087463379, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8818923234939575, "num_tokens": 837958553.0, "step": 22066 }, { "epoch": 2.8071492176567867, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6436264514923096, "learning_rate": 1e-05, "loss": 0.353, "mean_token_accuracy": 0.9025925397872925, "num_tokens": 837998674.0, "step": 22067 }, { "epoch": 2.8072764279353772, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.839616298675537, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8811902403831482, "num_tokens": 838035111.0, "step": 22068 }, { "epoch": 2.8074036382139678, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.651278495788574, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8760552406311035, "num_tokens": 838075444.0, "step": 22069 }, { "epoch": 2.8075308484925583, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.83469557762146, "learning_rate": 1e-05, "loss": 0.5008, "mean_token_accuracy": 0.859122633934021, "num_tokens": 838117972.0, "step": 22070 }, { "epoch": 2.807658058771149, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6537182331085205, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.8682842254638672, "num_tokens": 838158797.0, "step": 22071 }, { "epoch": 2.8077852690497394, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.829710006713867, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.8740835785865784, "num_tokens": 838198958.0, "step": 22072 }, { "epoch": 2.80791247932833, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.657552480697632, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8762289881706238, "num_tokens": 838233936.0, "step": 22073 }, { "epoch": 2.80803968960692, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8201117515563965, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.875139594078064, "num_tokens": 838274333.0, "step": 22074 }, { "epoch": 2.808166899885511, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6710331439971924, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8828833699226379, "num_tokens": 838304932.0, "step": 22075 }, { "epoch": 2.808294110164101, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8081488609313965, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8684079051017761, "num_tokens": 838345377.0, "step": 22076 }, { "epoch": 2.808421320442692, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6671175956726074, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8780920505523682, "num_tokens": 838383105.0, "step": 22077 }, { "epoch": 2.808548530721282, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.802309036254883, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8658995628356934, "num_tokens": 838421857.0, "step": 22078 }, { "epoch": 2.808675740999873, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.661437511444092, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.867637574672699, "num_tokens": 838459439.0, "step": 22079 }, { "epoch": 2.808802951278463, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.800915002822876, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8763746023178101, "num_tokens": 838501176.0, "step": 22080 }, { "epoch": 2.8089301615570537, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6639959812164307, "learning_rate": 1e-05, "loss": 0.461, "mean_token_accuracy": 0.8663966059684753, "num_tokens": 838538725.0, "step": 22081 }, { "epoch": 2.809057371835644, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7998390197753906, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8728455901145935, "num_tokens": 838577365.0, "step": 22082 }, { "epoch": 2.8091845821142347, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6549181938171387, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8855398893356323, "num_tokens": 838617872.0, "step": 22083 }, { "epoch": 2.8093117923928252, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7932870388031006, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8734588623046875, "num_tokens": 838656599.0, "step": 22084 }, { "epoch": 2.8094390026714158, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.653616189956665, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8772512674331665, "num_tokens": 838693270.0, "step": 22085 }, { "epoch": 2.8095662129500063, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8022539615631104, "learning_rate": 1e-05, "loss": 0.4515, "mean_token_accuracy": 0.8730084300041199, "num_tokens": 838728493.0, "step": 22086 }, { "epoch": 2.809693423228597, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6501855850219727, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8799771070480347, "num_tokens": 838767136.0, "step": 22087 }, { "epoch": 2.8098206335071874, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.80778431892395, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8772549033164978, "num_tokens": 838801894.0, "step": 22088 }, { "epoch": 2.809947843785778, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6378118991851807, "learning_rate": 1e-05, "loss": 0.364, "mean_token_accuracy": 0.8991950154304504, "num_tokens": 838838732.0, "step": 22089 }, { "epoch": 2.8100750540643684, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8058388233184814, "learning_rate": 1e-05, "loss": 0.3581, "mean_token_accuracy": 0.9018003940582275, "num_tokens": 838876331.0, "step": 22090 }, { "epoch": 2.810202264342959, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6501271724700928, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8820983171463013, "num_tokens": 838909613.0, "step": 22091 }, { "epoch": 2.8103294746215495, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8124146461486816, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8708707094192505, "num_tokens": 838944694.0, "step": 22092 }, { "epoch": 2.81045668490014, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.642517328262329, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8845467567443848, "num_tokens": 838984595.0, "step": 22093 }, { "epoch": 2.8105838951787305, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.814445734024048, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8757987022399902, "num_tokens": 839018528.0, "step": 22094 }, { "epoch": 2.810711105457321, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.649876594543457, "learning_rate": 1e-05, "loss": 0.4311, "mean_token_accuracy": 0.8769811391830444, "num_tokens": 839056447.0, "step": 22095 }, { "epoch": 2.8108383157359116, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8171422481536865, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.8858798742294312, "num_tokens": 839093612.0, "step": 22096 }, { "epoch": 2.810965526014502, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6475067138671875, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8777744770050049, "num_tokens": 839130192.0, "step": 22097 }, { "epoch": 2.8110927362930926, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8258962631225586, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8742946982383728, "num_tokens": 839167821.0, "step": 22098 }, { "epoch": 2.8112199465716827, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6705498695373535, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8699950575828552, "num_tokens": 839200124.0, "step": 22099 }, { "epoch": 2.8113471568502737, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.794983148574829, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8752949833869934, "num_tokens": 839242787.0, "step": 22100 }, { "epoch": 2.811474367128864, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.653627872467041, "learning_rate": 1e-05, "loss": 0.3811, "mean_token_accuracy": 0.894078254699707, "num_tokens": 839277517.0, "step": 22101 }, { "epoch": 2.8116015774074548, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7751855850219727, "learning_rate": 1e-05, "loss": 0.388, "mean_token_accuracy": 0.8925120234489441, "num_tokens": 839318609.0, "step": 22102 }, { "epoch": 2.811728787686045, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6810901165008545, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8739863634109497, "num_tokens": 839353964.0, "step": 22103 }, { "epoch": 2.811855997964636, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.760647773742676, "learning_rate": 1e-05, "loss": 0.3815, "mean_token_accuracy": 0.8945797681808472, "num_tokens": 839393926.0, "step": 22104 }, { "epoch": 2.811983208243226, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6692328453063965, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8784298896789551, "num_tokens": 839437857.0, "step": 22105 }, { "epoch": 2.8121104185218164, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.769848108291626, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8691814541816711, "num_tokens": 839478444.0, "step": 22106 }, { "epoch": 2.812237628800407, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6777288913726807, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8824924230575562, "num_tokens": 839512752.0, "step": 22107 }, { "epoch": 2.8123648390789975, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7731943130493164, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8663604259490967, "num_tokens": 839550024.0, "step": 22108 }, { "epoch": 2.812492049357588, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6772751808166504, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.887136697769165, "num_tokens": 839582612.0, "step": 22109 }, { "epoch": 2.8126192596361785, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7560267448425293, "learning_rate": 1e-05, "loss": 0.41, "mean_token_accuracy": 0.8823500871658325, "num_tokens": 839621791.0, "step": 22110 }, { "epoch": 2.812746469914769, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.677379846572876, "learning_rate": 1e-05, "loss": 0.4687, "mean_token_accuracy": 0.8657127618789673, "num_tokens": 839660829.0, "step": 22111 }, { "epoch": 2.8128736801933596, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7464888095855713, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.880760669708252, "num_tokens": 839701675.0, "step": 22112 }, { "epoch": 2.81300089047195, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6740832328796387, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8712868690490723, "num_tokens": 839742958.0, "step": 22113 }, { "epoch": 2.8131281007505406, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7346043586730957, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8792634010314941, "num_tokens": 839787552.0, "step": 22114 }, { "epoch": 2.813255311029131, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6776204109191895, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8769069314002991, "num_tokens": 839824630.0, "step": 22115 }, { "epoch": 2.8133825213077217, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.739556074142456, "learning_rate": 1e-05, "loss": 0.3927, "mean_token_accuracy": 0.8891174793243408, "num_tokens": 839857671.0, "step": 22116 }, { "epoch": 2.8135097315863122, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.681152105331421, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8757256269454956, "num_tokens": 839896395.0, "step": 22117 }, { "epoch": 2.8136369418649028, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.722891330718994, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8789166212081909, "num_tokens": 839942188.0, "step": 22118 }, { "epoch": 2.8137641521434933, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.696220874786377, "learning_rate": 1e-05, "loss": 0.3639, "mean_token_accuracy": 0.8956549167633057, "num_tokens": 839970510.0, "step": 22119 }, { "epoch": 2.813891362422084, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7229409217834473, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8806449174880981, "num_tokens": 840007458.0, "step": 22120 }, { "epoch": 2.8140185727006743, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.680912733078003, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8777660131454468, "num_tokens": 840049729.0, "step": 22121 }, { "epoch": 2.814145782979265, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.712799549102783, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8813657760620117, "num_tokens": 840089883.0, "step": 22122 }, { "epoch": 2.8142729932578554, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.746079683303833, "learning_rate": 1e-05, "loss": 0.3754, "mean_token_accuracy": 0.8963700532913208, "num_tokens": 840121237.0, "step": 22123 }, { "epoch": 2.8144002035364455, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6961114406585693, "learning_rate": 1e-05, "loss": 0.4503, "mean_token_accuracy": 0.8702495098114014, "num_tokens": 840157213.0, "step": 22124 }, { "epoch": 2.8145274138150365, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.720517158508301, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8838645815849304, "num_tokens": 840191404.0, "step": 22125 }, { "epoch": 2.8146546240936265, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.701107978820801, "learning_rate": 1e-05, "loss": 0.4552, "mean_token_accuracy": 0.8721289038658142, "num_tokens": 840227803.0, "step": 22126 }, { "epoch": 2.8147818343722175, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7282938957214355, "learning_rate": 1e-05, "loss": 0.4943, "mean_token_accuracy": 0.858110249042511, "num_tokens": 840266686.0, "step": 22127 }, { "epoch": 2.8149090446508076, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6710317134857178, "learning_rate": 1e-05, "loss": 0.4016, "mean_token_accuracy": 0.8859202861785889, "num_tokens": 840310437.0, "step": 22128 }, { "epoch": 2.815036254929398, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7314202785491943, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.87811279296875, "num_tokens": 840349599.0, "step": 22129 }, { "epoch": 2.8151634652079887, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.665827751159668, "learning_rate": 1e-05, "loss": 0.4125, "mean_token_accuracy": 0.8832799792289734, "num_tokens": 840389143.0, "step": 22130 }, { "epoch": 2.815290675486579, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7328946590423584, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8857842683792114, "num_tokens": 840429440.0, "step": 22131 }, { "epoch": 2.8154178857651697, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6824963092803955, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8789265155792236, "num_tokens": 840466056.0, "step": 22132 }, { "epoch": 2.8155450960437602, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7379531860351562, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8779828548431396, "num_tokens": 840503039.0, "step": 22133 }, { "epoch": 2.8156723063223508, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6705162525177, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8830119371414185, "num_tokens": 840546809.0, "step": 22134 }, { "epoch": 2.8157995166009413, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7451372146606445, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.875537097454071, "num_tokens": 840582617.0, "step": 22135 }, { "epoch": 2.815926726879532, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6656055450439453, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8816527128219604, "num_tokens": 840620279.0, "step": 22136 }, { "epoch": 2.8160539371581224, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7611143589019775, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8815589547157288, "num_tokens": 840655925.0, "step": 22137 }, { "epoch": 2.816181147436713, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6676695346832275, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.880901575088501, "num_tokens": 840692328.0, "step": 22138 }, { "epoch": 2.8163083577153034, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7591049671173096, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.880771815776825, "num_tokens": 840730674.0, "step": 22139 }, { "epoch": 2.816435567993894, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6709775924682617, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8729657530784607, "num_tokens": 840764791.0, "step": 22140 }, { "epoch": 2.8165627782724845, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.756549596786499, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8808767795562744, "num_tokens": 840803527.0, "step": 22141 }, { "epoch": 2.816689988551075, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.7051806449890137, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.871740460395813, "num_tokens": 840843823.0, "step": 22142 }, { "epoch": 2.8168171988296655, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7415988445281982, "learning_rate": 1e-05, "loss": 0.4522, "mean_token_accuracy": 0.869225025177002, "num_tokens": 840879648.0, "step": 22143 }, { "epoch": 2.816944409108256, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6717636585235596, "learning_rate": 1e-05, "loss": 0.3755, "mean_token_accuracy": 0.892978847026825, "num_tokens": 840914979.0, "step": 22144 }, { "epoch": 2.8170716193868466, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.734685182571411, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8825269937515259, "num_tokens": 840951534.0, "step": 22145 }, { "epoch": 2.817198829665437, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.682884693145752, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8717048168182373, "num_tokens": 840987416.0, "step": 22146 }, { "epoch": 2.8173260399440276, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7433977127075195, "learning_rate": 1e-05, "loss": 0.475, "mean_token_accuracy": 0.8629417419433594, "num_tokens": 841026734.0, "step": 22147 }, { "epoch": 2.817453250222618, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.666945457458496, "learning_rate": 1e-05, "loss": 0.3793, "mean_token_accuracy": 0.8929717540740967, "num_tokens": 841065333.0, "step": 22148 }, { "epoch": 2.8175804605012083, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7434754371643066, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8870402574539185, "num_tokens": 841102257.0, "step": 22149 }, { "epoch": 2.8177076707797992, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.679713010787964, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.8808854222297668, "num_tokens": 841140253.0, "step": 22150 }, { "epoch": 2.8178348810583893, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.73584246635437, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.8840851783752441, "num_tokens": 841181675.0, "step": 22151 }, { "epoch": 2.8179620913369803, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6701695919036865, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8850250840187073, "num_tokens": 841219922.0, "step": 22152 }, { "epoch": 2.8180893016155704, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7475202083587646, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8870252370834351, "num_tokens": 841256296.0, "step": 22153 }, { "epoch": 2.818216511894161, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6638636589050293, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8858044147491455, "num_tokens": 841296737.0, "step": 22154 }, { "epoch": 2.8183437221727514, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.777721405029297, "learning_rate": 1e-05, "loss": 0.476, "mean_token_accuracy": 0.8662542104721069, "num_tokens": 841329863.0, "step": 22155 }, { "epoch": 2.818470932451342, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.663767099380493, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8751824498176575, "num_tokens": 841365057.0, "step": 22156 }, { "epoch": 2.8185981427299325, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7603347301483154, "learning_rate": 1e-05, "loss": 0.4133, "mean_token_accuracy": 0.8830238580703735, "num_tokens": 841404885.0, "step": 22157 }, { "epoch": 2.818725353008523, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6626312732696533, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8757212162017822, "num_tokens": 841445067.0, "step": 22158 }, { "epoch": 2.8188525632871135, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.757042169570923, "learning_rate": 1e-05, "loss": 0.3694, "mean_token_accuracy": 0.8949390649795532, "num_tokens": 841480569.0, "step": 22159 }, { "epoch": 2.818979773565704, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6546630859375, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8870095014572144, "num_tokens": 841520540.0, "step": 22160 }, { "epoch": 2.8191069838442946, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.760842800140381, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.887149453163147, "num_tokens": 841557765.0, "step": 22161 }, { "epoch": 2.819234194122885, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6698598861694336, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8759347796440125, "num_tokens": 841593345.0, "step": 22162 }, { "epoch": 2.8193614044014756, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7683756351470947, "learning_rate": 1e-05, "loss": 0.4597, "mean_token_accuracy": 0.8718359470367432, "num_tokens": 841629977.0, "step": 22163 }, { "epoch": 2.819488614680066, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6547391414642334, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8808439373970032, "num_tokens": 841671901.0, "step": 22164 }, { "epoch": 2.8196158249586567, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7699923515319824, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8799453377723694, "num_tokens": 841706072.0, "step": 22165 }, { "epoch": 2.8197430352372472, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6543421745300293, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8701732158660889, "num_tokens": 841745707.0, "step": 22166 }, { "epoch": 2.8198702455158378, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7822673320770264, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.8809981346130371, "num_tokens": 841781468.0, "step": 22167 }, { "epoch": 2.8199974557944283, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.646177291870117, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8883585929870605, "num_tokens": 841821890.0, "step": 22168 }, { "epoch": 2.820124666073019, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.797044277191162, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8697509765625, "num_tokens": 841857668.0, "step": 22169 }, { "epoch": 2.8202518763516093, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.653059244155884, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.878543496131897, "num_tokens": 841892595.0, "step": 22170 }, { "epoch": 2.8203790866302, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7977890968322754, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8925211429595947, "num_tokens": 841928077.0, "step": 22171 }, { "epoch": 2.82050629690879, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6360812187194824, "learning_rate": 1e-05, "loss": 0.4094, "mean_token_accuracy": 0.8815895915031433, "num_tokens": 841967847.0, "step": 22172 }, { "epoch": 2.820633507187381, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8081984519958496, "learning_rate": 1e-05, "loss": 0.4812, "mean_token_accuracy": 0.8677430748939514, "num_tokens": 842008931.0, "step": 22173 }, { "epoch": 2.820760717465971, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6519570350646973, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8778904676437378, "num_tokens": 842042870.0, "step": 22174 }, { "epoch": 2.820887927744562, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.809290885925293, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8810440301895142, "num_tokens": 842080193.0, "step": 22175 }, { "epoch": 2.821015138023152, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6313610076904297, "learning_rate": 1e-05, "loss": 0.3651, "mean_token_accuracy": 0.8970312476158142, "num_tokens": 842120953.0, "step": 22176 }, { "epoch": 2.821142348301743, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.799654722213745, "learning_rate": 1e-05, "loss": 0.3902, "mean_token_accuracy": 0.8897280693054199, "num_tokens": 842158466.0, "step": 22177 }, { "epoch": 2.821269558580333, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6493144035339355, "learning_rate": 1e-05, "loss": 0.4279, "mean_token_accuracy": 0.876204252243042, "num_tokens": 842194164.0, "step": 22178 }, { "epoch": 2.8213967688589237, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8049163818359375, "learning_rate": 1e-05, "loss": 0.4316, "mean_token_accuracy": 0.8777225017547607, "num_tokens": 842232137.0, "step": 22179 }, { "epoch": 2.821523979137514, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.663339853286743, "learning_rate": 1e-05, "loss": 0.454, "mean_token_accuracy": 0.8685980439186096, "num_tokens": 842263770.0, "step": 22180 }, { "epoch": 2.8216511894161047, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.887655258178711, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.870608389377594, "num_tokens": 842309420.0, "step": 22181 }, { "epoch": 2.8217783996946952, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.628077745437622, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.864112138748169, "num_tokens": 842350533.0, "step": 22182 }, { "epoch": 2.8219056099732858, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.840911388397217, "learning_rate": 1e-05, "loss": 0.4576, "mean_token_accuracy": 0.868148922920227, "num_tokens": 842389734.0, "step": 22183 }, { "epoch": 2.8220328202518763, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6525778770446777, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8700066804885864, "num_tokens": 842424425.0, "step": 22184 }, { "epoch": 2.822160030530467, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8283188343048096, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8792193531990051, "num_tokens": 842460269.0, "step": 22185 }, { "epoch": 2.8222872408090574, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.653064489364624, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.874407172203064, "num_tokens": 842497018.0, "step": 22186 }, { "epoch": 2.822414451087648, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8111279010772705, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8819916248321533, "num_tokens": 842537563.0, "step": 22187 }, { "epoch": 2.8225416613662384, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.647350549697876, "learning_rate": 1e-05, "loss": 0.3683, "mean_token_accuracy": 0.8985211849212646, "num_tokens": 842573842.0, "step": 22188 }, { "epoch": 2.822668871644829, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8170454502105713, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8683971166610718, "num_tokens": 842614216.0, "step": 22189 }, { "epoch": 2.8227960819234195, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.672711133956909, "learning_rate": 1e-05, "loss": 0.4926, "mean_token_accuracy": 0.8612430095672607, "num_tokens": 842651823.0, "step": 22190 }, { "epoch": 2.82292329220201, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.839273691177368, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8615624904632568, "num_tokens": 842695113.0, "step": 22191 }, { "epoch": 2.8230505024806005, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.651136875152588, "learning_rate": 1e-05, "loss": 0.3734, "mean_token_accuracy": 0.8944785594940186, "num_tokens": 842731341.0, "step": 22192 }, { "epoch": 2.823177712759191, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.824247360229492, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8702672719955444, "num_tokens": 842769611.0, "step": 22193 }, { "epoch": 2.8233049230377816, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.663874626159668, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8704943060874939, "num_tokens": 842807106.0, "step": 22194 }, { "epoch": 2.823432133316372, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8138177394866943, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8829444646835327, "num_tokens": 842847957.0, "step": 22195 }, { "epoch": 2.8235593435949626, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6578598022460938, "learning_rate": 1e-05, "loss": 0.4647, "mean_token_accuracy": 0.8667219281196594, "num_tokens": 842889222.0, "step": 22196 }, { "epoch": 2.8236865538735527, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8009653091430664, "learning_rate": 1e-05, "loss": 0.4021, "mean_token_accuracy": 0.8859075307846069, "num_tokens": 842925164.0, "step": 22197 }, { "epoch": 2.8238137641521437, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.666963815689087, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.8651553392410278, "num_tokens": 842963832.0, "step": 22198 }, { "epoch": 2.8239409744307338, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.7876079082489014, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.874365508556366, "num_tokens": 843004357.0, "step": 22199 }, { "epoch": 2.8240681847093247, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6673622131347656, "learning_rate": 1e-05, "loss": 0.4829, "mean_token_accuracy": 0.8586888909339905, "num_tokens": 843045241.0, "step": 22200 }, { "epoch": 2.824195394987915, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.792433500289917, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8725812435150146, "num_tokens": 843084473.0, "step": 22201 }, { "epoch": 2.824322605266506, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.662224531173706, "learning_rate": 1e-05, "loss": 0.433, "mean_token_accuracy": 0.8776814937591553, "num_tokens": 843122118.0, "step": 22202 }, { "epoch": 2.824449815545096, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.795384407043457, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8667545318603516, "num_tokens": 843159936.0, "step": 22203 }, { "epoch": 2.8245770258236864, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6713337898254395, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8748605847358704, "num_tokens": 843196212.0, "step": 22204 }, { "epoch": 2.824704236102277, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7733867168426514, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8820732831954956, "num_tokens": 843236870.0, "step": 22205 }, { "epoch": 2.8248314463808675, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6689136028289795, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.8834219574928284, "num_tokens": 843270701.0, "step": 22206 }, { "epoch": 2.824958656659458, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7687108516693115, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8841278553009033, "num_tokens": 843308547.0, "step": 22207 }, { "epoch": 2.8250858669380485, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6471643447875977, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8905489444732666, "num_tokens": 843354496.0, "step": 22208 }, { "epoch": 2.825213077216639, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7902050018310547, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8703556060791016, "num_tokens": 843389995.0, "step": 22209 }, { "epoch": 2.8253402874952296, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6613929271698, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8757113218307495, "num_tokens": 843428444.0, "step": 22210 }, { "epoch": 2.82546749777382, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.774953842163086, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8828208446502686, "num_tokens": 843466390.0, "step": 22211 }, { "epoch": 2.8255947080524106, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.673086643218994, "learning_rate": 1e-05, "loss": 0.4969, "mean_token_accuracy": 0.8625521063804626, "num_tokens": 843503424.0, "step": 22212 }, { "epoch": 2.825721918331001, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7723445892333984, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8680647611618042, "num_tokens": 843544669.0, "step": 22213 }, { "epoch": 2.8258491286095917, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.665856122970581, "learning_rate": 1e-05, "loss": 0.3764, "mean_token_accuracy": 0.8935333490371704, "num_tokens": 843579777.0, "step": 22214 }, { "epoch": 2.8259763388881822, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7725179195404053, "learning_rate": 1e-05, "loss": 0.4155, "mean_token_accuracy": 0.8792430758476257, "num_tokens": 843613481.0, "step": 22215 }, { "epoch": 2.8261035491667728, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6539430618286133, "learning_rate": 1e-05, "loss": 0.3797, "mean_token_accuracy": 0.8932772874832153, "num_tokens": 843656870.0, "step": 22216 }, { "epoch": 2.8262307594453633, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7738988399505615, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8803732395172119, "num_tokens": 843693859.0, "step": 22217 }, { "epoch": 2.826357969723954, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.661997079849243, "learning_rate": 1e-05, "loss": 0.3797, "mean_token_accuracy": 0.8946211338043213, "num_tokens": 843726999.0, "step": 22218 }, { "epoch": 2.8264851800025443, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7724621295928955, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8726771473884583, "num_tokens": 843767244.0, "step": 22219 }, { "epoch": 2.826612390281135, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6790592670440674, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8772667050361633, "num_tokens": 843799828.0, "step": 22220 }, { "epoch": 2.8267396005597254, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7617688179016113, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8790832757949829, "num_tokens": 843837485.0, "step": 22221 }, { "epoch": 2.8268668108383155, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6533985137939453, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.885482132434845, "num_tokens": 843878796.0, "step": 22222 }, { "epoch": 2.8269940211169065, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.775066614151001, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8738212585449219, "num_tokens": 843920200.0, "step": 22223 }, { "epoch": 2.8271212313954965, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.646989583969116, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8870894908905029, "num_tokens": 843961719.0, "step": 22224 }, { "epoch": 2.8272484416740875, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7769339084625244, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8709657192230225, "num_tokens": 844005085.0, "step": 22225 }, { "epoch": 2.8273756519526776, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6751585006713867, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8704172968864441, "num_tokens": 844036964.0, "step": 22226 }, { "epoch": 2.827502862231268, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7706997394561768, "learning_rate": 1e-05, "loss": 0.4627, "mean_token_accuracy": 0.8729395866394043, "num_tokens": 844078297.0, "step": 22227 }, { "epoch": 2.8276300725098586, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6572787761688232, "learning_rate": 1e-05, "loss": 0.384, "mean_token_accuracy": 0.8916682004928589, "num_tokens": 844113749.0, "step": 22228 }, { "epoch": 2.827757282788449, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.76749587059021, "learning_rate": 1e-05, "loss": 0.4525, "mean_token_accuracy": 0.8710662126541138, "num_tokens": 844155993.0, "step": 22229 }, { "epoch": 2.8278844930670397, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.648833751678467, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8787453174591064, "num_tokens": 844204302.0, "step": 22230 }, { "epoch": 2.8280117033456302, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.784252166748047, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8834993243217468, "num_tokens": 844241490.0, "step": 22231 }, { "epoch": 2.8281389136242208, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.651367425918579, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.882049560546875, "num_tokens": 844282891.0, "step": 22232 }, { "epoch": 2.8282661239028113, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.779031753540039, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8754165172576904, "num_tokens": 844324407.0, "step": 22233 }, { "epoch": 2.828393334181402, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.669269323348999, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8820527791976929, "num_tokens": 844356656.0, "step": 22234 }, { "epoch": 2.8285205444599923, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.790255546569824, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8729802966117859, "num_tokens": 844389789.0, "step": 22235 }, { "epoch": 2.828647754738583, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6585590839385986, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.8795394897460938, "num_tokens": 844423599.0, "step": 22236 }, { "epoch": 2.8287749650171734, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7828822135925293, "learning_rate": 1e-05, "loss": 0.4187, "mean_token_accuracy": 0.8823387622833252, "num_tokens": 844460117.0, "step": 22237 }, { "epoch": 2.828902175295764, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.642658233642578, "learning_rate": 1e-05, "loss": 0.4461, "mean_token_accuracy": 0.8691574335098267, "num_tokens": 844501263.0, "step": 22238 }, { "epoch": 2.8290293855743545, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.781078577041626, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.8811376690864563, "num_tokens": 844541573.0, "step": 22239 }, { "epoch": 2.829156595852945, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.643036127090454, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8794323801994324, "num_tokens": 844583728.0, "step": 22240 }, { "epoch": 2.8292838061315355, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.805633544921875, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8812710046768188, "num_tokens": 844616611.0, "step": 22241 }, { "epoch": 2.829411016410126, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6620991230010986, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8629799485206604, "num_tokens": 844652067.0, "step": 22242 }, { "epoch": 2.8295382266887166, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.796853542327881, "learning_rate": 1e-05, "loss": 0.4429, "mean_token_accuracy": 0.8726420998573303, "num_tokens": 844688582.0, "step": 22243 }, { "epoch": 2.829665436967307, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.640019416809082, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8793911933898926, "num_tokens": 844727120.0, "step": 22244 }, { "epoch": 2.8297926472458976, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.786625385284424, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8799278736114502, "num_tokens": 844772328.0, "step": 22245 }, { "epoch": 2.829919857524488, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6269783973693848, "learning_rate": 1e-05, "loss": 0.3719, "mean_token_accuracy": 0.8929835557937622, "num_tokens": 844812528.0, "step": 22246 }, { "epoch": 2.8300470678030782, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.80415940284729, "learning_rate": 1e-05, "loss": 0.4486, "mean_token_accuracy": 0.8732572793960571, "num_tokens": 844857208.0, "step": 22247 }, { "epoch": 2.830174278081669, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6560330390930176, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.870952844619751, "num_tokens": 844890362.0, "step": 22248 }, { "epoch": 2.8303014883602593, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8038384914398193, "learning_rate": 1e-05, "loss": 0.4659, "mean_token_accuracy": 0.8662033081054688, "num_tokens": 844928664.0, "step": 22249 }, { "epoch": 2.8304286986388503, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6673946380615234, "learning_rate": 1e-05, "loss": 0.3708, "mean_token_accuracy": 0.8976641893386841, "num_tokens": 844957043.0, "step": 22250 }, { "epoch": 2.8305559089174404, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.80246639251709, "learning_rate": 1e-05, "loss": 0.4385, "mean_token_accuracy": 0.87569260597229, "num_tokens": 844991808.0, "step": 22251 }, { "epoch": 2.830683119196031, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.663390636444092, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8806270360946655, "num_tokens": 845025571.0, "step": 22252 }, { "epoch": 2.8308103294746214, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.78955340385437, "learning_rate": 1e-05, "loss": 0.4601, "mean_token_accuracy": 0.8696000576019287, "num_tokens": 845063154.0, "step": 22253 }, { "epoch": 2.830937539753212, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6566572189331055, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.8860321044921875, "num_tokens": 845100256.0, "step": 22254 }, { "epoch": 2.8310647500318025, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7802393436431885, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8788065910339355, "num_tokens": 845138111.0, "step": 22255 }, { "epoch": 2.831191960310393, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.658064842224121, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8830824494361877, "num_tokens": 845175759.0, "step": 22256 }, { "epoch": 2.8313191705889835, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7891311645507812, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8788384795188904, "num_tokens": 845212232.0, "step": 22257 }, { "epoch": 2.831446380867574, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6509621143341064, "learning_rate": 1e-05, "loss": 0.3706, "mean_token_accuracy": 0.8963704109191895, "num_tokens": 845247141.0, "step": 22258 }, { "epoch": 2.8315735911461646, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7840304374694824, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8847030401229858, "num_tokens": 845287908.0, "step": 22259 }, { "epoch": 2.831700801424755, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.648588180541992, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8834167122840881, "num_tokens": 845328134.0, "step": 22260 }, { "epoch": 2.8318280117033456, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7926411628723145, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8843236565589905, "num_tokens": 845366096.0, "step": 22261 }, { "epoch": 2.831955221981936, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.661508083343506, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8807978630065918, "num_tokens": 845397737.0, "step": 22262 }, { "epoch": 2.8320824322605267, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7761948108673096, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8823739886283875, "num_tokens": 845437015.0, "step": 22263 }, { "epoch": 2.8322096425391172, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6474008560180664, "learning_rate": 1e-05, "loss": 0.3937, "mean_token_accuracy": 0.88932204246521, "num_tokens": 845473778.0, "step": 22264 }, { "epoch": 2.8323368528177078, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7827484607696533, "learning_rate": 1e-05, "loss": 0.3858, "mean_token_accuracy": 0.8914953470230103, "num_tokens": 845509254.0, "step": 22265 }, { "epoch": 2.8324640630962983, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6609883308410645, "learning_rate": 1e-05, "loss": 0.4487, "mean_token_accuracy": 0.8725250959396362, "num_tokens": 845547682.0, "step": 22266 }, { "epoch": 2.832591273374889, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7676117420196533, "learning_rate": 1e-05, "loss": 0.3847, "mean_token_accuracy": 0.8891063928604126, "num_tokens": 845585315.0, "step": 22267 }, { "epoch": 2.8327184836534793, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6605801582336426, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8841866254806519, "num_tokens": 845625065.0, "step": 22268 }, { "epoch": 2.83284569393207, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.757427453994751, "learning_rate": 1e-05, "loss": 0.3937, "mean_token_accuracy": 0.8888947367668152, "num_tokens": 845664148.0, "step": 22269 }, { "epoch": 2.83297290421066, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6801912784576416, "learning_rate": 1e-05, "loss": 0.3937, "mean_token_accuracy": 0.8840522766113281, "num_tokens": 845695986.0, "step": 22270 }, { "epoch": 2.833100114489251, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7604143619537354, "learning_rate": 1e-05, "loss": 0.4326, "mean_token_accuracy": 0.8793199062347412, "num_tokens": 845732807.0, "step": 22271 }, { "epoch": 2.833227324767841, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6887857913970947, "learning_rate": 1e-05, "loss": 0.4025, "mean_token_accuracy": 0.8853144645690918, "num_tokens": 845763238.0, "step": 22272 }, { "epoch": 2.833354535046432, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.740280866622925, "learning_rate": 1e-05, "loss": 0.3861, "mean_token_accuracy": 0.8920773267745972, "num_tokens": 845802218.0, "step": 22273 }, { "epoch": 2.833481745325022, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.687494993209839, "learning_rate": 1e-05, "loss": 0.4849, "mean_token_accuracy": 0.8595271110534668, "num_tokens": 845845056.0, "step": 22274 }, { "epoch": 2.833608955603613, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.737399101257324, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.8762544989585876, "num_tokens": 845884909.0, "step": 22275 }, { "epoch": 2.833736165882203, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.686612606048584, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8785505890846252, "num_tokens": 845923894.0, "step": 22276 }, { "epoch": 2.8338633761607936, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.742678642272949, "learning_rate": 1e-05, "loss": 0.4819, "mean_token_accuracy": 0.8613042235374451, "num_tokens": 845962452.0, "step": 22277 }, { "epoch": 2.833990586439384, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6787056922912598, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8787164688110352, "num_tokens": 845998031.0, "step": 22278 }, { "epoch": 2.8341177967179747, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.731458902359009, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8758491277694702, "num_tokens": 846036920.0, "step": 22279 }, { "epoch": 2.8342450069965652, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6913552284240723, "learning_rate": 1e-05, "loss": 0.4632, "mean_token_accuracy": 0.868239164352417, "num_tokens": 846075978.0, "step": 22280 }, { "epoch": 2.8343722172751558, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.735168933868408, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8782395124435425, "num_tokens": 846111893.0, "step": 22281 }, { "epoch": 2.8344994275537463, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.672433376312256, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8758842349052429, "num_tokens": 846153417.0, "step": 22282 }, { "epoch": 2.834626637832337, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.736410617828369, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8730306029319763, "num_tokens": 846194375.0, "step": 22283 }, { "epoch": 2.8347538481109273, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6743271350860596, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8839069604873657, "num_tokens": 846231040.0, "step": 22284 }, { "epoch": 2.834881058389518, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7320146560668945, "learning_rate": 1e-05, "loss": 0.3919, "mean_token_accuracy": 0.889116644859314, "num_tokens": 846270645.0, "step": 22285 }, { "epoch": 2.8350082686681084, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6818788051605225, "learning_rate": 1e-05, "loss": 0.399, "mean_token_accuracy": 0.885061502456665, "num_tokens": 846304038.0, "step": 22286 }, { "epoch": 2.835135478946699, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7483530044555664, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8704672455787659, "num_tokens": 846338136.0, "step": 22287 }, { "epoch": 2.8352626892252895, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6801791191101074, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8769451379776001, "num_tokens": 846372654.0, "step": 22288 }, { "epoch": 2.83538989950388, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7342233657836914, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8764686584472656, "num_tokens": 846409180.0, "step": 22289 }, { "epoch": 2.8355171097824705, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.671527862548828, "learning_rate": 1e-05, "loss": 0.4623, "mean_token_accuracy": 0.8673248887062073, "num_tokens": 846453410.0, "step": 22290 }, { "epoch": 2.835644320061061, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.738551616668701, "learning_rate": 1e-05, "loss": 0.4019, "mean_token_accuracy": 0.8870548605918884, "num_tokens": 846488271.0, "step": 22291 }, { "epoch": 2.8357715303396516, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.659853935241699, "learning_rate": 1e-05, "loss": 0.3932, "mean_token_accuracy": 0.8862085342407227, "num_tokens": 846528730.0, "step": 22292 }, { "epoch": 2.835898740618242, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.741502285003662, "learning_rate": 1e-05, "loss": 0.3762, "mean_token_accuracy": 0.893025279045105, "num_tokens": 846567276.0, "step": 22293 }, { "epoch": 2.8360259508968326, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6581647396087646, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8858471512794495, "num_tokens": 846608482.0, "step": 22294 }, { "epoch": 2.8361531611754227, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7564194202423096, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8745535612106323, "num_tokens": 846648582.0, "step": 22295 }, { "epoch": 2.8362803714540137, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6525490283966064, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8881822824478149, "num_tokens": 846687392.0, "step": 22296 }, { "epoch": 2.8364075817326038, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7745420932769775, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8840838670730591, "num_tokens": 846721542.0, "step": 22297 }, { "epoch": 2.8365347920111947, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6494200229644775, "learning_rate": 1e-05, "loss": 0.4569, "mean_token_accuracy": 0.8691251873970032, "num_tokens": 846759262.0, "step": 22298 }, { "epoch": 2.836662002289785, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7751033306121826, "learning_rate": 1e-05, "loss": 0.4457, "mean_token_accuracy": 0.8742685317993164, "num_tokens": 846796979.0, "step": 22299 }, { "epoch": 2.836789212568376, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6537227630615234, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8782941102981567, "num_tokens": 846832417.0, "step": 22300 }, { "epoch": 2.836916422846966, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7788798809051514, "learning_rate": 1e-05, "loss": 0.4513, "mean_token_accuracy": 0.8739527463912964, "num_tokens": 846870289.0, "step": 22301 }, { "epoch": 2.8370436331255564, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6432442665100098, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8805865049362183, "num_tokens": 846909821.0, "step": 22302 }, { "epoch": 2.837170843404147, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7909820079803467, "learning_rate": 1e-05, "loss": 0.4252, "mean_token_accuracy": 0.8808765411376953, "num_tokens": 846945512.0, "step": 22303 }, { "epoch": 2.8372980536827375, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.643662452697754, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8855249285697937, "num_tokens": 846980834.0, "step": 22304 }, { "epoch": 2.837425263961328, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7855417728424072, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8815600275993347, "num_tokens": 847025320.0, "step": 22305 }, { "epoch": 2.8375524742399185, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6370065212249756, "learning_rate": 1e-05, "loss": 0.3891, "mean_token_accuracy": 0.8897418975830078, "num_tokens": 847063637.0, "step": 22306 }, { "epoch": 2.837679684518509, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.79308819770813, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8867058753967285, "num_tokens": 847102314.0, "step": 22307 }, { "epoch": 2.8378068947970996, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6397528648376465, "learning_rate": 1e-05, "loss": 0.41, "mean_token_accuracy": 0.8824763298034668, "num_tokens": 847141956.0, "step": 22308 }, { "epoch": 2.83793410507569, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.806506633758545, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8708798885345459, "num_tokens": 847178855.0, "step": 22309 }, { "epoch": 2.8380613153542806, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.648860216140747, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8807482719421387, "num_tokens": 847214228.0, "step": 22310 }, { "epoch": 2.838188525632871, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.798394203186035, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.8769884705543518, "num_tokens": 847255942.0, "step": 22311 }, { "epoch": 2.8383157359114617, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.649310827255249, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8654974699020386, "num_tokens": 847294742.0, "step": 22312 }, { "epoch": 2.838442946190052, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8011410236358643, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8769453167915344, "num_tokens": 847330105.0, "step": 22313 }, { "epoch": 2.8385701564686427, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.638209104537964, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8736571073532104, "num_tokens": 847373291.0, "step": 22314 }, { "epoch": 2.8386973667472333, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.802483558654785, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8912610411643982, "num_tokens": 847409059.0, "step": 22315 }, { "epoch": 2.838824577025824, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6430673599243164, "learning_rate": 1e-05, "loss": 0.418, "mean_token_accuracy": 0.8841224312782288, "num_tokens": 847445869.0, "step": 22316 }, { "epoch": 2.8389517873044143, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.80134916305542, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8824340105056763, "num_tokens": 847486816.0, "step": 22317 }, { "epoch": 2.839078997583005, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6359503269195557, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.886143684387207, "num_tokens": 847527482.0, "step": 22318 }, { "epoch": 2.8392062078615954, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.811767339706421, "learning_rate": 1e-05, "loss": 0.467, "mean_token_accuracy": 0.8666918873786926, "num_tokens": 847570850.0, "step": 22319 }, { "epoch": 2.8393334181401855, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6437253952026367, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.875268816947937, "num_tokens": 847608872.0, "step": 22320 }, { "epoch": 2.8394606284187764, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.811523199081421, "learning_rate": 1e-05, "loss": 0.4432, "mean_token_accuracy": 0.873589038848877, "num_tokens": 847647813.0, "step": 22321 }, { "epoch": 2.8395878386973665, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6389529705047607, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8817564249038696, "num_tokens": 847685726.0, "step": 22322 }, { "epoch": 2.8397150489759575, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8178935050964355, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8730773329734802, "num_tokens": 847723029.0, "step": 22323 }, { "epoch": 2.8398422592545476, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6346750259399414, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8858364820480347, "num_tokens": 847761922.0, "step": 22324 }, { "epoch": 2.839969469533138, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8173046112060547, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8751890063285828, "num_tokens": 847806229.0, "step": 22325 }, { "epoch": 2.8400966798117286, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6474034786224365, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8780763745307922, "num_tokens": 847844448.0, "step": 22326 }, { "epoch": 2.840223890090319, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8211238384246826, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8697115182876587, "num_tokens": 847889387.0, "step": 22327 }, { "epoch": 2.8403511003689097, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.647387981414795, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8764603137969971, "num_tokens": 847926993.0, "step": 22328 }, { "epoch": 2.8404783106475002, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8160977363586426, "learning_rate": 1e-05, "loss": 0.398, "mean_token_accuracy": 0.8884807825088501, "num_tokens": 847958884.0, "step": 22329 }, { "epoch": 2.8406055209260908, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.654768228530884, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8773541450500488, "num_tokens": 847997894.0, "step": 22330 }, { "epoch": 2.8407327312046813, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8043079376220703, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8791229724884033, "num_tokens": 848037392.0, "step": 22331 }, { "epoch": 2.840859941483272, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.65207576751709, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8810060024261475, "num_tokens": 848072844.0, "step": 22332 }, { "epoch": 2.8409871517618623, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8169407844543457, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8659957647323608, "num_tokens": 848107477.0, "step": 22333 }, { "epoch": 2.841114362040453, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.653437852859497, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8795100450515747, "num_tokens": 848143124.0, "step": 22334 }, { "epoch": 2.8412415723190434, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.798182487487793, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.881428599357605, "num_tokens": 848181454.0, "step": 22335 }, { "epoch": 2.841368782597634, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.645979881286621, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8865205645561218, "num_tokens": 848219300.0, "step": 22336 }, { "epoch": 2.8414959928762245, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.796898365020752, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8831095695495605, "num_tokens": 848256650.0, "step": 22337 }, { "epoch": 2.841623203154815, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.661827802658081, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8706896305084229, "num_tokens": 848294836.0, "step": 22338 }, { "epoch": 2.8417504134334055, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8072350025177, "learning_rate": 1e-05, "loss": 0.4243, "mean_token_accuracy": 0.8815661668777466, "num_tokens": 848326306.0, "step": 22339 }, { "epoch": 2.841877623711996, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6547460556030273, "learning_rate": 1e-05, "loss": 0.3972, "mean_token_accuracy": 0.889563798904419, "num_tokens": 848360886.0, "step": 22340 }, { "epoch": 2.8420048339905866, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8129279613494873, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.877873420715332, "num_tokens": 848399364.0, "step": 22341 }, { "epoch": 2.842132044269177, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6419992446899414, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8857588768005371, "num_tokens": 848436978.0, "step": 22342 }, { "epoch": 2.8422592545477676, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.808669328689575, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8664227724075317, "num_tokens": 848474661.0, "step": 22343 }, { "epoch": 2.842386464826358, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.648073434829712, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.8900157809257507, "num_tokens": 848514469.0, "step": 22344 }, { "epoch": 2.8425136751049482, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8161284923553467, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8841012716293335, "num_tokens": 848548606.0, "step": 22345 }, { "epoch": 2.842640885383539, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.659877061843872, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.871869683265686, "num_tokens": 848585309.0, "step": 22346 }, { "epoch": 2.8427680956621293, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.80130934715271, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8803645968437195, "num_tokens": 848625062.0, "step": 22347 }, { "epoch": 2.8428953059407203, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6627066135406494, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8772161602973938, "num_tokens": 848661914.0, "step": 22348 }, { "epoch": 2.8430225162193103, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8005263805389404, "learning_rate": 1e-05, "loss": 0.5195, "mean_token_accuracy": 0.8503190875053406, "num_tokens": 848704343.0, "step": 22349 }, { "epoch": 2.843149726497901, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6577398777008057, "learning_rate": 1e-05, "loss": 0.3819, "mean_token_accuracy": 0.8936875462532043, "num_tokens": 848737415.0, "step": 22350 }, { "epoch": 2.8432769367764914, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7985379695892334, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8772462010383606, "num_tokens": 848770795.0, "step": 22351 }, { "epoch": 2.843404147055082, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.638200044631958, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.8862694501876831, "num_tokens": 848815086.0, "step": 22352 }, { "epoch": 2.8435313573336725, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8017938137054443, "learning_rate": 1e-05, "loss": 0.3699, "mean_token_accuracy": 0.8957206010818481, "num_tokens": 848853545.0, "step": 22353 }, { "epoch": 2.843658567612263, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.655428409576416, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8854971528053284, "num_tokens": 848886047.0, "step": 22354 }, { "epoch": 2.8437857778908535, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.792915105819702, "learning_rate": 1e-05, "loss": 0.4115, "mean_token_accuracy": 0.8836228847503662, "num_tokens": 848929147.0, "step": 22355 }, { "epoch": 2.843912988169444, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6686995029449463, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.8775509595870972, "num_tokens": 848962912.0, "step": 22356 }, { "epoch": 2.8440401984480346, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7939038276672363, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8793381452560425, "num_tokens": 848997954.0, "step": 22357 }, { "epoch": 2.844167408726625, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6410529613494873, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8759909272193909, "num_tokens": 849044418.0, "step": 22358 }, { "epoch": 2.8442946190052156, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.791881799697876, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.878337025642395, "num_tokens": 849085591.0, "step": 22359 }, { "epoch": 2.844421829283806, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6569228172302246, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8760324716567993, "num_tokens": 849128046.0, "step": 22360 }, { "epoch": 2.8445490395623967, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8005523681640625, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8723909854888916, "num_tokens": 849165231.0, "step": 22361 }, { "epoch": 2.844676249840987, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6506879329681396, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8813201189041138, "num_tokens": 849203414.0, "step": 22362 }, { "epoch": 2.8448034601195777, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.793175458908081, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8841148018836975, "num_tokens": 849242463.0, "step": 22363 }, { "epoch": 2.8449306703981683, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6542558670043945, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8836305737495422, "num_tokens": 849280098.0, "step": 22364 }, { "epoch": 2.845057880676759, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7924156188964844, "learning_rate": 1e-05, "loss": 0.4055, "mean_token_accuracy": 0.8833746314048767, "num_tokens": 849317460.0, "step": 22365 }, { "epoch": 2.8451850909553493, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6614415645599365, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8833873867988586, "num_tokens": 849354852.0, "step": 22366 }, { "epoch": 2.84531230123394, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.789205551147461, "learning_rate": 1e-05, "loss": 0.4289, "mean_token_accuracy": 0.8788421154022217, "num_tokens": 849397132.0, "step": 22367 }, { "epoch": 2.84543951151253, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.649390459060669, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8817440271377563, "num_tokens": 849441751.0, "step": 22368 }, { "epoch": 2.845566721791121, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8016726970672607, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8736028671264648, "num_tokens": 849480177.0, "step": 22369 }, { "epoch": 2.845693932069711, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6591882705688477, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8731787204742432, "num_tokens": 849520211.0, "step": 22370 }, { "epoch": 2.845821142348302, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8074111938476562, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8796987533569336, "num_tokens": 849551591.0, "step": 22371 }, { "epoch": 2.845948352626892, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6525282859802246, "learning_rate": 1e-05, "loss": 0.4455, "mean_token_accuracy": 0.8723146319389343, "num_tokens": 849592700.0, "step": 22372 }, { "epoch": 2.846075562905483, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.802901029586792, "learning_rate": 1e-05, "loss": 0.465, "mean_token_accuracy": 0.8693851232528687, "num_tokens": 849631220.0, "step": 22373 }, { "epoch": 2.846202773184073, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6428048610687256, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8855361938476562, "num_tokens": 849670349.0, "step": 22374 }, { "epoch": 2.8463299834626636, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.815091133117676, "learning_rate": 1e-05, "loss": 0.507, "mean_token_accuracy": 0.8527522087097168, "num_tokens": 849711000.0, "step": 22375 }, { "epoch": 2.846457193741254, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6511809825897217, "learning_rate": 1e-05, "loss": 0.4544, "mean_token_accuracy": 0.8667826652526855, "num_tokens": 849751199.0, "step": 22376 }, { "epoch": 2.8465844040198447, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.801135540008545, "learning_rate": 1e-05, "loss": 0.3611, "mean_token_accuracy": 0.9023240208625793, "num_tokens": 849785416.0, "step": 22377 }, { "epoch": 2.8467116142984352, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6631455421447754, "learning_rate": 1e-05, "loss": 0.456, "mean_token_accuracy": 0.8701134324073792, "num_tokens": 849818553.0, "step": 22378 }, { "epoch": 2.8468388245770258, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.79263973236084, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8855644464492798, "num_tokens": 849854994.0, "step": 22379 }, { "epoch": 2.8469660348556163, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6530416011810303, "learning_rate": 1e-05, "loss": 0.4668, "mean_token_accuracy": 0.8611290454864502, "num_tokens": 849896603.0, "step": 22380 }, { "epoch": 2.847093245134207, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.804100275039673, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8680867552757263, "num_tokens": 849935999.0, "step": 22381 }, { "epoch": 2.8472204554127973, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6389145851135254, "learning_rate": 1e-05, "loss": 0.4013, "mean_token_accuracy": 0.8843965530395508, "num_tokens": 849976858.0, "step": 22382 }, { "epoch": 2.847347665691388, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.796790361404419, "learning_rate": 1e-05, "loss": 0.389, "mean_token_accuracy": 0.8926975727081299, "num_tokens": 850012802.0, "step": 22383 }, { "epoch": 2.8474748759699784, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.672463893890381, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8723028898239136, "num_tokens": 850047330.0, "step": 22384 }, { "epoch": 2.847602086248569, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.793001890182495, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.87333083152771, "num_tokens": 850087543.0, "step": 22385 }, { "epoch": 2.8477292965271594, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.656296968460083, "learning_rate": 1e-05, "loss": 0.4008, "mean_token_accuracy": 0.8840209245681763, "num_tokens": 850126864.0, "step": 22386 }, { "epoch": 2.84785650680575, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.79429030418396, "learning_rate": 1e-05, "loss": 0.3867, "mean_token_accuracy": 0.8908678889274597, "num_tokens": 850162476.0, "step": 22387 }, { "epoch": 2.8479837170843405, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.669879913330078, "learning_rate": 1e-05, "loss": 0.429, "mean_token_accuracy": 0.8767164945602417, "num_tokens": 850197606.0, "step": 22388 }, { "epoch": 2.848110927362931, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.8828954696655273, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8785340785980225, "num_tokens": 850232223.0, "step": 22389 }, { "epoch": 2.8482381376415216, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.634424924850464, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8859156370162964, "num_tokens": 850268936.0, "step": 22390 }, { "epoch": 2.848365347920112, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8465726375579834, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8734647035598755, "num_tokens": 850301652.0, "step": 22391 }, { "epoch": 2.8484925581987026, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6427693367004395, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8824396133422852, "num_tokens": 850335033.0, "step": 22392 }, { "epoch": 2.8486197684772927, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8446033000946045, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8781914710998535, "num_tokens": 850369677.0, "step": 22393 }, { "epoch": 2.8487469787558837, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.644533395767212, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.865358829498291, "num_tokens": 850411464.0, "step": 22394 }, { "epoch": 2.8488741890344738, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8453006744384766, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8812054991722107, "num_tokens": 850447578.0, "step": 22395 }, { "epoch": 2.8490013993130647, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6329710483551025, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8844536542892456, "num_tokens": 850489689.0, "step": 22396 }, { "epoch": 2.849128609591655, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.8421449661254883, "learning_rate": 1e-05, "loss": 0.4068, "mean_token_accuracy": 0.8852240443229675, "num_tokens": 850528598.0, "step": 22397 }, { "epoch": 2.849255819870246, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.663147211074829, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8808286786079407, "num_tokens": 850566953.0, "step": 22398 }, { "epoch": 2.849383030148836, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8333747386932373, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.8873419165611267, "num_tokens": 850604286.0, "step": 22399 }, { "epoch": 2.8495102404274264, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6689860820770264, "learning_rate": 1e-05, "loss": 0.4454, "mean_token_accuracy": 0.8755366802215576, "num_tokens": 850640724.0, "step": 22400 }, { "epoch": 2.849637450706017, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.815009593963623, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8783729672431946, "num_tokens": 850682377.0, "step": 22401 }, { "epoch": 2.8497646609846075, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.656653881072998, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8774938583374023, "num_tokens": 850721637.0, "step": 22402 }, { "epoch": 2.849891871263198, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.82429838180542, "learning_rate": 1e-05, "loss": 0.4416, "mean_token_accuracy": 0.8728971481323242, "num_tokens": 850754884.0, "step": 22403 }, { "epoch": 2.8500190815417885, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6603851318359375, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8743205666542053, "num_tokens": 850793706.0, "step": 22404 }, { "epoch": 2.850146291820379, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.827505588531494, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8688120245933533, "num_tokens": 850827708.0, "step": 22405 }, { "epoch": 2.8502735020989696, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6459336280822754, "learning_rate": 1e-05, "loss": 0.4399, "mean_token_accuracy": 0.8712129592895508, "num_tokens": 850870808.0, "step": 22406 }, { "epoch": 2.85040071237756, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.826355218887329, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8808528780937195, "num_tokens": 850906129.0, "step": 22407 }, { "epoch": 2.8505279226561506, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6526730060577393, "learning_rate": 1e-05, "loss": 0.4313, "mean_token_accuracy": 0.8804982900619507, "num_tokens": 850944021.0, "step": 22408 }, { "epoch": 2.850655132934741, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.818673610687256, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8844693899154663, "num_tokens": 850983040.0, "step": 22409 }, { "epoch": 2.8507823432133317, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.660829544067383, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.879925012588501, "num_tokens": 851017400.0, "step": 22410 }, { "epoch": 2.850909553491922, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.816283702850342, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8807413578033447, "num_tokens": 851055559.0, "step": 22411 }, { "epoch": 2.8510367637705127, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6605770587921143, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.888791561126709, "num_tokens": 851088810.0, "step": 22412 }, { "epoch": 2.8511639740491033, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.818028688430786, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8788326978683472, "num_tokens": 851121531.0, "step": 22413 }, { "epoch": 2.851291184327694, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6556341648101807, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8796712160110474, "num_tokens": 851160601.0, "step": 22414 }, { "epoch": 2.8514183946062843, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.822476625442505, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.867920994758606, "num_tokens": 851195887.0, "step": 22415 }, { "epoch": 2.851545604884875, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6597001552581787, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.8853604197502136, "num_tokens": 851231761.0, "step": 22416 }, { "epoch": 2.8516728151634654, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8528456687927246, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8628469705581665, "num_tokens": 851270765.0, "step": 22417 }, { "epoch": 2.8518000254420555, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.651975631713867, "learning_rate": 1e-05, "loss": 0.4845, "mean_token_accuracy": 0.864945650100708, "num_tokens": 851312200.0, "step": 22418 }, { "epoch": 2.8519272357206464, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8204808235168457, "learning_rate": 1e-05, "loss": 0.4428, "mean_token_accuracy": 0.8741133213043213, "num_tokens": 851350854.0, "step": 22419 }, { "epoch": 2.8520544459992365, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.655444383621216, "learning_rate": 1e-05, "loss": 0.4529, "mean_token_accuracy": 0.8668512105941772, "num_tokens": 851389242.0, "step": 22420 }, { "epoch": 2.8521816562778275, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8174636363983154, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8770994544029236, "num_tokens": 851425450.0, "step": 22421 }, { "epoch": 2.8523088665564176, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6667802333831787, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.877339243888855, "num_tokens": 851459470.0, "step": 22422 }, { "epoch": 2.852436076835008, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7999889850616455, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8688779473304749, "num_tokens": 851500849.0, "step": 22423 }, { "epoch": 2.8525632871135986, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.665003538131714, "learning_rate": 1e-05, "loss": 0.4265, "mean_token_accuracy": 0.8792883157730103, "num_tokens": 851540048.0, "step": 22424 }, { "epoch": 2.852690497392189, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.810894250869751, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8763834238052368, "num_tokens": 851573343.0, "step": 22425 }, { "epoch": 2.8528177076707797, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6629621982574463, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.863771378993988, "num_tokens": 851613658.0, "step": 22426 }, { "epoch": 2.85294491794937, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8009233474731445, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8878806233406067, "num_tokens": 851649961.0, "step": 22427 }, { "epoch": 2.8530721282279607, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6642277240753174, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.8779274225234985, "num_tokens": 851685465.0, "step": 22428 }, { "epoch": 2.8531993385065513, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.802072763442993, "learning_rate": 1e-05, "loss": 0.3765, "mean_token_accuracy": 0.8937901854515076, "num_tokens": 851718434.0, "step": 22429 }, { "epoch": 2.853326548785142, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6597940921783447, "learning_rate": 1e-05, "loss": 0.4082, "mean_token_accuracy": 0.8855515718460083, "num_tokens": 851756207.0, "step": 22430 }, { "epoch": 2.8534537590637323, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8236284255981445, "learning_rate": 1e-05, "loss": 0.4489, "mean_token_accuracy": 0.8741008043289185, "num_tokens": 851795398.0, "step": 22431 }, { "epoch": 2.853580969342323, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.669621229171753, "learning_rate": 1e-05, "loss": 0.4579, "mean_token_accuracy": 0.8686968088150024, "num_tokens": 851832569.0, "step": 22432 }, { "epoch": 2.8537081796209134, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.786149024963379, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.8826946020126343, "num_tokens": 851870749.0, "step": 22433 }, { "epoch": 2.853835389899504, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.646822452545166, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8824343681335449, "num_tokens": 851912591.0, "step": 22434 }, { "epoch": 2.8539626001780944, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8049464225769043, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.877864420413971, "num_tokens": 851948792.0, "step": 22435 }, { "epoch": 2.854089810456685, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6462693214416504, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8708779215812683, "num_tokens": 851988811.0, "step": 22436 }, { "epoch": 2.8542170207352755, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.802171468734741, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.8848440051078796, "num_tokens": 852024781.0, "step": 22437 }, { "epoch": 2.854344231013866, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.655555248260498, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8853456974029541, "num_tokens": 852056676.0, "step": 22438 }, { "epoch": 2.8544714412924566, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7999932765960693, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8680362701416016, "num_tokens": 852095500.0, "step": 22439 }, { "epoch": 2.854598651571047, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.639954090118408, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.8894224762916565, "num_tokens": 852133658.0, "step": 22440 }, { "epoch": 2.8547258618496376, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8023293018341064, "learning_rate": 1e-05, "loss": 0.4662, "mean_token_accuracy": 0.8631041049957275, "num_tokens": 852173963.0, "step": 22441 }, { "epoch": 2.854853072128228, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6522750854492188, "learning_rate": 1e-05, "loss": 0.428, "mean_token_accuracy": 0.8774609565734863, "num_tokens": 852209957.0, "step": 22442 }, { "epoch": 2.8549802824068182, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.7873799800872803, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8805065155029297, "num_tokens": 852252281.0, "step": 22443 }, { "epoch": 2.855107492685409, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6575968265533447, "learning_rate": 1e-05, "loss": 0.3708, "mean_token_accuracy": 0.8967216610908508, "num_tokens": 852285287.0, "step": 22444 }, { "epoch": 2.8552347029639993, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.793475866317749, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8691915273666382, "num_tokens": 852326596.0, "step": 22445 }, { "epoch": 2.8553619132425903, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6730616092681885, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8848737478256226, "num_tokens": 852358062.0, "step": 22446 }, { "epoch": 2.8554891235211803, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.792834520339966, "learning_rate": 1e-05, "loss": 0.4613, "mean_token_accuracy": 0.8699905276298523, "num_tokens": 852395828.0, "step": 22447 }, { "epoch": 2.855616333799771, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6537561416625977, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8739831447601318, "num_tokens": 852439466.0, "step": 22448 }, { "epoch": 2.8557435440783614, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7854480743408203, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8808479905128479, "num_tokens": 852477992.0, "step": 22449 }, { "epoch": 2.855870754356952, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6483588218688965, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8851174712181091, "num_tokens": 852514814.0, "step": 22450 }, { "epoch": 2.8559979646355425, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.799532890319824, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8810514211654663, "num_tokens": 852553866.0, "step": 22451 }, { "epoch": 2.856125174914133, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.655963659286499, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8673498034477234, "num_tokens": 852594291.0, "step": 22452 }, { "epoch": 2.8562523851927235, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8018596172332764, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8835774064064026, "num_tokens": 852630586.0, "step": 22453 }, { "epoch": 2.856379595471314, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.651841163635254, "learning_rate": 1e-05, "loss": 0.4323, "mean_token_accuracy": 0.8747434616088867, "num_tokens": 852670013.0, "step": 22454 }, { "epoch": 2.8565068057499046, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.809037923812866, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8783900141716003, "num_tokens": 852704097.0, "step": 22455 }, { "epoch": 2.856634016028495, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6659531593322754, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8733515739440918, "num_tokens": 852739771.0, "step": 22456 }, { "epoch": 2.8567612263070856, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8063406944274902, "learning_rate": 1e-05, "loss": 0.464, "mean_token_accuracy": 0.8698660731315613, "num_tokens": 852778604.0, "step": 22457 }, { "epoch": 2.856888436585676, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6682353019714355, "learning_rate": 1e-05, "loss": 0.4927, "mean_token_accuracy": 0.8576703667640686, "num_tokens": 852814966.0, "step": 22458 }, { "epoch": 2.8570156468642667, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.805506706237793, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8690389394760132, "num_tokens": 852849562.0, "step": 22459 }, { "epoch": 2.857142857142857, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6598901748657227, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8832417726516724, "num_tokens": 852884183.0, "step": 22460 }, { "epoch": 2.8572700674214477, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7992680072784424, "learning_rate": 1e-05, "loss": 0.472, "mean_token_accuracy": 0.8704330921173096, "num_tokens": 852925121.0, "step": 22461 }, { "epoch": 2.8573972777000383, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.649733304977417, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.884343147277832, "num_tokens": 852965962.0, "step": 22462 }, { "epoch": 2.857524487978629, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.804291009902954, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8774334192276001, "num_tokens": 853004886.0, "step": 22463 }, { "epoch": 2.8576516982572193, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.660311222076416, "learning_rate": 1e-05, "loss": 0.4335, "mean_token_accuracy": 0.8741284608840942, "num_tokens": 853038898.0, "step": 22464 }, { "epoch": 2.85777890853581, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8024978637695312, "learning_rate": 1e-05, "loss": 0.4756, "mean_token_accuracy": 0.8622554540634155, "num_tokens": 853079064.0, "step": 22465 }, { "epoch": 2.8579061188144, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.651151180267334, "learning_rate": 1e-05, "loss": 0.4831, "mean_token_accuracy": 0.8608921766281128, "num_tokens": 853125704.0, "step": 22466 }, { "epoch": 2.858033329092991, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8222436904907227, "learning_rate": 1e-05, "loss": 0.4418, "mean_token_accuracy": 0.8753819465637207, "num_tokens": 853168760.0, "step": 22467 }, { "epoch": 2.858160539371581, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.650061845779419, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8785964250564575, "num_tokens": 853207146.0, "step": 22468 }, { "epoch": 2.858287749650172, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.802657127380371, "learning_rate": 1e-05, "loss": 0.4089, "mean_token_accuracy": 0.8846473097801208, "num_tokens": 853248764.0, "step": 22469 }, { "epoch": 2.858414959928762, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.658780336380005, "learning_rate": 1e-05, "loss": 0.4006, "mean_token_accuracy": 0.8842728137969971, "num_tokens": 853283325.0, "step": 22470 }, { "epoch": 2.858542170207353, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8374838829040527, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8774759769439697, "num_tokens": 853316579.0, "step": 22471 }, { "epoch": 2.858669380485943, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.648864269256592, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8812618851661682, "num_tokens": 853353109.0, "step": 22472 }, { "epoch": 2.8587965907645336, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.820044994354248, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.880867063999176, "num_tokens": 853394974.0, "step": 22473 }, { "epoch": 2.858923801043124, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.643423080444336, "learning_rate": 1e-05, "loss": 0.3781, "mean_token_accuracy": 0.8946102857589722, "num_tokens": 853433819.0, "step": 22474 }, { "epoch": 2.8590510113217147, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.826721668243408, "learning_rate": 1e-05, "loss": 0.4269, "mean_token_accuracy": 0.8778084516525269, "num_tokens": 853471947.0, "step": 22475 }, { "epoch": 2.859178221600305, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6629621982574463, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8786975741386414, "num_tokens": 853505936.0, "step": 22476 }, { "epoch": 2.8593054318788957, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.810014486312866, "learning_rate": 1e-05, "loss": 0.3766, "mean_token_accuracy": 0.896817147731781, "num_tokens": 853540957.0, "step": 22477 }, { "epoch": 2.8594326421574863, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6617836952209473, "learning_rate": 1e-05, "loss": 0.3866, "mean_token_accuracy": 0.8920605182647705, "num_tokens": 853574117.0, "step": 22478 }, { "epoch": 2.859559852436077, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8138139247894287, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.872678279876709, "num_tokens": 853612424.0, "step": 22479 }, { "epoch": 2.8596870627146673, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.659074544906616, "learning_rate": 1e-05, "loss": 0.4307, "mean_token_accuracy": 0.8765771389007568, "num_tokens": 853652659.0, "step": 22480 }, { "epoch": 2.859814272993258, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.798947811126709, "learning_rate": 1e-05, "loss": 0.4064, "mean_token_accuracy": 0.8852732181549072, "num_tokens": 853690763.0, "step": 22481 }, { "epoch": 2.8599414832718484, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6584110260009766, "learning_rate": 1e-05, "loss": 0.3863, "mean_token_accuracy": 0.8909097909927368, "num_tokens": 853727437.0, "step": 22482 }, { "epoch": 2.860068693550439, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.802523136138916, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8695836067199707, "num_tokens": 853769179.0, "step": 22483 }, { "epoch": 2.8601959038290294, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6511588096618652, "learning_rate": 1e-05, "loss": 0.3815, "mean_token_accuracy": 0.8922624588012695, "num_tokens": 853803374.0, "step": 22484 }, { "epoch": 2.86032311410762, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8328511714935303, "learning_rate": 1e-05, "loss": 0.4495, "mean_token_accuracy": 0.8716839551925659, "num_tokens": 853834286.0, "step": 22485 }, { "epoch": 2.8604503243862105, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6454901695251465, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.8845657110214233, "num_tokens": 853872369.0, "step": 22486 }, { "epoch": 2.860577534664801, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.83046555519104, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.868545651435852, "num_tokens": 853916470.0, "step": 22487 }, { "epoch": 2.8607047449433916, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6343700885772705, "learning_rate": 1e-05, "loss": 0.3687, "mean_token_accuracy": 0.8981943130493164, "num_tokens": 853956802.0, "step": 22488 }, { "epoch": 2.860831955221982, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8421709537506104, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8844338655471802, "num_tokens": 853990914.0, "step": 22489 }, { "epoch": 2.8609591655005726, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6420412063598633, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8840929269790649, "num_tokens": 854029707.0, "step": 22490 }, { "epoch": 2.8610863757791627, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.835386037826538, "learning_rate": 1e-05, "loss": 0.364, "mean_token_accuracy": 0.9003087878227234, "num_tokens": 854065584.0, "step": 22491 }, { "epoch": 2.8612135860577537, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6448147296905518, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8778395056724548, "num_tokens": 854105937.0, "step": 22492 }, { "epoch": 2.8613407963363438, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8356404304504395, "learning_rate": 1e-05, "loss": 0.411, "mean_token_accuracy": 0.8831495046615601, "num_tokens": 854146746.0, "step": 22493 }, { "epoch": 2.8614680066149347, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6523876190185547, "learning_rate": 1e-05, "loss": 0.394, "mean_token_accuracy": 0.889983057975769, "num_tokens": 854179953.0, "step": 22494 }, { "epoch": 2.861595216893525, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.831310272216797, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8817000389099121, "num_tokens": 854220765.0, "step": 22495 }, { "epoch": 2.861722427172116, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6464240550994873, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8816908001899719, "num_tokens": 854258544.0, "step": 22496 }, { "epoch": 2.861849637450706, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.841357946395874, "learning_rate": 1e-05, "loss": 0.4381, "mean_token_accuracy": 0.8800745606422424, "num_tokens": 854297250.0, "step": 22497 }, { "epoch": 2.8619768477292964, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6383063793182373, "learning_rate": 1e-05, "loss": 0.4347, "mean_token_accuracy": 0.8768469095230103, "num_tokens": 854340283.0, "step": 22498 }, { "epoch": 2.862104058007887, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8266384601593018, "learning_rate": 1e-05, "loss": 0.3857, "mean_token_accuracy": 0.8942846059799194, "num_tokens": 854380145.0, "step": 22499 }, { "epoch": 2.8622312682864774, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6440353393554688, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8806853294372559, "num_tokens": 854422809.0, "step": 22500 }, { "epoch": 2.862358478565068, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.824759006500244, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8852192759513855, "num_tokens": 854461926.0, "step": 22501 }, { "epoch": 2.8624856888436585, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.660102605819702, "learning_rate": 1e-05, "loss": 0.4585, "mean_token_accuracy": 0.8684550523757935, "num_tokens": 854502431.0, "step": 22502 }, { "epoch": 2.862612899122249, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.81587815284729, "learning_rate": 1e-05, "loss": 0.375, "mean_token_accuracy": 0.8947587013244629, "num_tokens": 854537440.0, "step": 22503 }, { "epoch": 2.8627401094008396, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.869879722595215, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8849499225616455, "num_tokens": 854575121.0, "step": 22504 }, { "epoch": 2.86286731967943, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7336983680725098, "learning_rate": 1e-05, "loss": 0.3808, "mean_token_accuracy": 0.8935656547546387, "num_tokens": 854603624.0, "step": 22505 }, { "epoch": 2.8629945299580206, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.721186637878418, "learning_rate": 1e-05, "loss": 0.4656, "mean_token_accuracy": 0.862256646156311, "num_tokens": 854644616.0, "step": 22506 }, { "epoch": 2.863121740236611, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6754238605499268, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8760970830917358, "num_tokens": 854689772.0, "step": 22507 }, { "epoch": 2.8632489505152017, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.741266965866089, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8891162872314453, "num_tokens": 854725549.0, "step": 22508 }, { "epoch": 2.863376160793792, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6757373809814453, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8848609924316406, "num_tokens": 854768509.0, "step": 22509 }, { "epoch": 2.8635033710723827, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.751909017562866, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.874582052230835, "num_tokens": 854804822.0, "step": 22510 }, { "epoch": 2.8636305813509733, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003681182861328125, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6827383041381836, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8849145174026489, "num_tokens": 854837458.0, "step": 22511 }, { "epoch": 2.863757791629564, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.733339548110962, "learning_rate": 1e-05, "loss": 0.4259, "mean_token_accuracy": 0.87940913438797, "num_tokens": 854877996.0, "step": 22512 }, { "epoch": 2.8638850019081543, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6894071102142334, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8736734390258789, "num_tokens": 854913078.0, "step": 22513 }, { "epoch": 2.864012212186745, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.722414255142212, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8862890005111694, "num_tokens": 854952300.0, "step": 22514 }, { "epoch": 2.8641394224653354, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6751534938812256, "learning_rate": 1e-05, "loss": 0.3982, "mean_token_accuracy": 0.8855005502700806, "num_tokens": 854992657.0, "step": 22515 }, { "epoch": 2.8642666327439255, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7346832752227783, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8805451393127441, "num_tokens": 855028147.0, "step": 22516 }, { "epoch": 2.8643938430225164, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6797494888305664, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8775402307510376, "num_tokens": 855069930.0, "step": 22517 }, { "epoch": 2.8645210533011065, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.739199161529541, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.883387565612793, "num_tokens": 855105224.0, "step": 22518 }, { "epoch": 2.8646482635796975, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.702033758163452, "learning_rate": 1e-05, "loss": 0.4586, "mean_token_accuracy": 0.8671790361404419, "num_tokens": 855139533.0, "step": 22519 }, { "epoch": 2.8647754738582876, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7378499507904053, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8672828674316406, "num_tokens": 855178380.0, "step": 22520 }, { "epoch": 2.864902684136878, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6953232288360596, "learning_rate": 1e-05, "loss": 0.4331, "mean_token_accuracy": 0.877211332321167, "num_tokens": 855212700.0, "step": 22521 }, { "epoch": 2.8650298944154686, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.727381467819214, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.8873614072799683, "num_tokens": 855250340.0, "step": 22522 }, { "epoch": 2.865157104694059, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6938493251800537, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8806294798851013, "num_tokens": 855285341.0, "step": 22523 }, { "epoch": 2.8652843149726497, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7079970836639404, "learning_rate": 1e-05, "loss": 0.4312, "mean_token_accuracy": 0.8772425055503845, "num_tokens": 855331038.0, "step": 22524 }, { "epoch": 2.86541152525124, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7087669372558594, "learning_rate": 1e-05, "loss": 0.3983, "mean_token_accuracy": 0.8847276568412781, "num_tokens": 855363643.0, "step": 22525 }, { "epoch": 2.8655387355298307, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7255802154541016, "learning_rate": 1e-05, "loss": 0.4031, "mean_token_accuracy": 0.8874990344047546, "num_tokens": 855397342.0, "step": 22526 }, { "epoch": 2.8656659458084213, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6922543048858643, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8909999132156372, "num_tokens": 855432867.0, "step": 22527 }, { "epoch": 2.865793156087012, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.808426856994629, "learning_rate": 1e-05, "loss": 0.44, "mean_token_accuracy": 0.8750717043876648, "num_tokens": 855466091.0, "step": 22528 }, { "epoch": 2.8659203663656023, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6711578369140625, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8667393326759338, "num_tokens": 855501200.0, "step": 22529 }, { "epoch": 2.866047576644193, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7468650341033936, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8889061808586121, "num_tokens": 855543625.0, "step": 22530 }, { "epoch": 2.8661747869227834, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6633596420288086, "learning_rate": 1e-05, "loss": 0.4494, "mean_token_accuracy": 0.8730549812316895, "num_tokens": 855579653.0, "step": 22531 }, { "epoch": 2.866301997201374, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7659924030303955, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8802183866500854, "num_tokens": 855611081.0, "step": 22532 }, { "epoch": 2.8664292074799644, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6668474674224854, "learning_rate": 1e-05, "loss": 0.3832, "mean_token_accuracy": 0.8927016854286194, "num_tokens": 855642389.0, "step": 22533 }, { "epoch": 2.866556417758555, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7574386596679688, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.873833417892456, "num_tokens": 855686240.0, "step": 22534 }, { "epoch": 2.8666836280371455, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6563360691070557, "learning_rate": 1e-05, "loss": 0.4341, "mean_token_accuracy": 0.8748467564582825, "num_tokens": 855727979.0, "step": 22535 }, { "epoch": 2.866810838315736, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.770195722579956, "learning_rate": 1e-05, "loss": 0.4566, "mean_token_accuracy": 0.8696044683456421, "num_tokens": 855769119.0, "step": 22536 }, { "epoch": 2.8669380485943265, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6429827213287354, "learning_rate": 1e-05, "loss": 0.3791, "mean_token_accuracy": 0.8911730051040649, "num_tokens": 855811472.0, "step": 22537 }, { "epoch": 2.867065258872917, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.796761989593506, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8812052011489868, "num_tokens": 855842739.0, "step": 22538 }, { "epoch": 2.8671924691515076, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6657843589782715, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8779492378234863, "num_tokens": 855874520.0, "step": 22539 }, { "epoch": 2.867319679430098, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7661337852478027, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.888817548751831, "num_tokens": 855915151.0, "step": 22540 }, { "epoch": 2.867446889708688, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.680060625076294, "learning_rate": 1e-05, "loss": 0.4675, "mean_token_accuracy": 0.8631912469863892, "num_tokens": 855948944.0, "step": 22541 }, { "epoch": 2.867574099987279, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.781261682510376, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8757950067520142, "num_tokens": 855981604.0, "step": 22542 }, { "epoch": 2.8677013102658693, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.672975540161133, "learning_rate": 1e-05, "loss": 0.3902, "mean_token_accuracy": 0.8900499939918518, "num_tokens": 856020407.0, "step": 22543 }, { "epoch": 2.8678285205444602, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7780239582061768, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8675372004508972, "num_tokens": 856056514.0, "step": 22544 }, { "epoch": 2.8679557308230503, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6716840267181396, "learning_rate": 1e-05, "loss": 0.4305, "mean_token_accuracy": 0.8775947093963623, "num_tokens": 856091385.0, "step": 22545 }, { "epoch": 2.868082941101641, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.749826192855835, "learning_rate": 1e-05, "loss": 0.3727, "mean_token_accuracy": 0.896851122379303, "num_tokens": 856133077.0, "step": 22546 }, { "epoch": 2.8682101513802314, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.660388469696045, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8853135108947754, "num_tokens": 856174682.0, "step": 22547 }, { "epoch": 2.868337361658822, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7739648818969727, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8645792007446289, "num_tokens": 856214486.0, "step": 22548 }, { "epoch": 2.8684645719374124, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6759426593780518, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8726786971092224, "num_tokens": 856246816.0, "step": 22549 }, { "epoch": 2.868591782216003, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7710275650024414, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8730688691139221, "num_tokens": 856286232.0, "step": 22550 }, { "epoch": 2.8687189924945935, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6485061645507812, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8859547972679138, "num_tokens": 856326451.0, "step": 22551 }, { "epoch": 2.868846202773184, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7641820907592773, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8748131990432739, "num_tokens": 856367766.0, "step": 22552 }, { "epoch": 2.8689734130517746, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.691295623779297, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8757047057151794, "num_tokens": 856396085.0, "step": 22553 }, { "epoch": 2.869100623330365, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7346086502075195, "learning_rate": 1e-05, "loss": 0.383, "mean_token_accuracy": 0.8924480676651001, "num_tokens": 856439835.0, "step": 22554 }, { "epoch": 2.8692278336089556, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6970596313476562, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.8868693113327026, "num_tokens": 856470890.0, "step": 22555 }, { "epoch": 2.869355043887546, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.737821340560913, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8710409998893738, "num_tokens": 856510281.0, "step": 22556 }, { "epoch": 2.8694822541661367, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6713919639587402, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8666331171989441, "num_tokens": 856558194.0, "step": 22557 }, { "epoch": 2.869609464444727, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7493836879730225, "learning_rate": 1e-05, "loss": 0.419, "mean_token_accuracy": 0.8816092014312744, "num_tokens": 856594001.0, "step": 22558 }, { "epoch": 2.8697366747233177, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.665414810180664, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.8811431527137756, "num_tokens": 856636424.0, "step": 22559 }, { "epoch": 2.8698638850019083, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7615394592285156, "learning_rate": 1e-05, "loss": 0.4511, "mean_token_accuracy": 0.8701265454292297, "num_tokens": 856670561.0, "step": 22560 }, { "epoch": 2.869991095280499, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6787078380584717, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8710033297538757, "num_tokens": 856704852.0, "step": 22561 }, { "epoch": 2.8701183055590893, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7463202476501465, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8862466216087341, "num_tokens": 856743943.0, "step": 22562 }, { "epoch": 2.87024551583768, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.676727533340454, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8708755970001221, "num_tokens": 856782944.0, "step": 22563 }, { "epoch": 2.87037272611627, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.747602939605713, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8786844611167908, "num_tokens": 856821240.0, "step": 22564 }, { "epoch": 2.870499936394861, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6732749938964844, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8763163685798645, "num_tokens": 856860113.0, "step": 22565 }, { "epoch": 2.870627146673451, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7404277324676514, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8802292346954346, "num_tokens": 856902501.0, "step": 22566 }, { "epoch": 2.870754356952042, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6656241416931152, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8867301940917969, "num_tokens": 856940023.0, "step": 22567 }, { "epoch": 2.870881567230632, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7682831287384033, "learning_rate": 1e-05, "loss": 0.4889, "mean_token_accuracy": 0.8575812578201294, "num_tokens": 856980653.0, "step": 22568 }, { "epoch": 2.871008777509223, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6895503997802734, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8868752121925354, "num_tokens": 857008081.0, "step": 22569 }, { "epoch": 2.871135987787813, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7468535900115967, "learning_rate": 1e-05, "loss": 0.3838, "mean_token_accuracy": 0.89173424243927, "num_tokens": 857040884.0, "step": 22570 }, { "epoch": 2.8712631980664036, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6775925159454346, "learning_rate": 1e-05, "loss": 0.4421, "mean_token_accuracy": 0.8755430579185486, "num_tokens": 857078519.0, "step": 22571 }, { "epoch": 2.871390408344994, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7383432388305664, "learning_rate": 1e-05, "loss": 0.4254, "mean_token_accuracy": 0.8802922964096069, "num_tokens": 857119250.0, "step": 22572 }, { "epoch": 2.8715176186235847, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6717259883880615, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8814319372177124, "num_tokens": 857160003.0, "step": 22573 }, { "epoch": 2.871644828902175, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.737363338470459, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8814030289649963, "num_tokens": 857200918.0, "step": 22574 }, { "epoch": 2.8717720391807657, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.668097496032715, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.8889337778091431, "num_tokens": 857241157.0, "step": 22575 }, { "epoch": 2.8718992494593563, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.754056215286255, "learning_rate": 1e-05, "loss": 0.4169, "mean_token_accuracy": 0.878902792930603, "num_tokens": 857276237.0, "step": 22576 }, { "epoch": 2.872026459737947, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.668132781982422, "learning_rate": 1e-05, "loss": 0.3796, "mean_token_accuracy": 0.8938860893249512, "num_tokens": 857311059.0, "step": 22577 }, { "epoch": 2.8721536700165373, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.769045352935791, "learning_rate": 1e-05, "loss": 0.4721, "mean_token_accuracy": 0.8665217161178589, "num_tokens": 857350326.0, "step": 22578 }, { "epoch": 2.872280880295128, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6663706302642822, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8709585666656494, "num_tokens": 857387708.0, "step": 22579 }, { "epoch": 2.8724080905737184, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.763653039932251, "learning_rate": 1e-05, "loss": 0.4135, "mean_token_accuracy": 0.8835479021072388, "num_tokens": 857427658.0, "step": 22580 }, { "epoch": 2.872535300852309, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.650524377822876, "learning_rate": 1e-05, "loss": 0.4413, "mean_token_accuracy": 0.8752239942550659, "num_tokens": 857468529.0, "step": 22581 }, { "epoch": 2.8726625111308994, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.764056921005249, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8818720579147339, "num_tokens": 857509406.0, "step": 22582 }, { "epoch": 2.87278972140949, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6576590538024902, "learning_rate": 1e-05, "loss": 0.426, "mean_token_accuracy": 0.882487416267395, "num_tokens": 857548302.0, "step": 22583 }, { "epoch": 2.8729169316880805, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7648768424987793, "learning_rate": 1e-05, "loss": 0.3708, "mean_token_accuracy": 0.8965392708778381, "num_tokens": 857584022.0, "step": 22584 }, { "epoch": 2.873044141966671, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6638453006744385, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8724005222320557, "num_tokens": 857627334.0, "step": 22585 }, { "epoch": 2.8731713522452615, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7747409343719482, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8810203075408936, "num_tokens": 857663317.0, "step": 22586 }, { "epoch": 2.873298562523852, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6872475147247314, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.864544153213501, "num_tokens": 857697358.0, "step": 22587 }, { "epoch": 2.8734257728024426, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.760509490966797, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8756908178329468, "num_tokens": 857737686.0, "step": 22588 }, { "epoch": 2.8735529830810327, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6694867610931396, "learning_rate": 1e-05, "loss": 0.4541, "mean_token_accuracy": 0.8706923723220825, "num_tokens": 857774730.0, "step": 22589 }, { "epoch": 2.8736801933596237, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7575109004974365, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.874789834022522, "num_tokens": 857811503.0, "step": 22590 }, { "epoch": 2.8738074036382137, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.669848680496216, "learning_rate": 1e-05, "loss": 0.4451, "mean_token_accuracy": 0.8720619082450867, "num_tokens": 857850696.0, "step": 22591 }, { "epoch": 2.8739346139168047, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.740107297897339, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8856512904167175, "num_tokens": 857894849.0, "step": 22592 }, { "epoch": 2.874061824195395, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6551923751831055, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.8833901286125183, "num_tokens": 857938876.0, "step": 22593 }, { "epoch": 2.8741890344739858, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7604315280914307, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8841363191604614, "num_tokens": 857973424.0, "step": 22594 }, { "epoch": 2.874316244752576, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6742923259735107, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8728618025779724, "num_tokens": 858008105.0, "step": 22595 }, { "epoch": 2.8744434550311664, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.757925510406494, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8836159706115723, "num_tokens": 858043898.0, "step": 22596 }, { "epoch": 2.874570665309757, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.675537347793579, "learning_rate": 1e-05, "loss": 0.4695, "mean_token_accuracy": 0.8650995492935181, "num_tokens": 858080671.0, "step": 22597 }, { "epoch": 2.8746978755883474, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7420711517333984, "learning_rate": 1e-05, "loss": 0.3781, "mean_token_accuracy": 0.8943121433258057, "num_tokens": 858118619.0, "step": 22598 }, { "epoch": 2.874825085866938, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6676406860351562, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8772150278091431, "num_tokens": 858160207.0, "step": 22599 }, { "epoch": 2.8749522961455285, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.749181032180786, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8720666170120239, "num_tokens": 858201634.0, "step": 22600 }, { "epoch": 2.875079506424119, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6766207218170166, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8708901405334473, "num_tokens": 858241618.0, "step": 22601 }, { "epoch": 2.8752067167027096, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7594919204711914, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8784526586532593, "num_tokens": 858275393.0, "step": 22602 }, { "epoch": 2.8753339269813, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.672988176345825, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8744222521781921, "num_tokens": 858316470.0, "step": 22603 }, { "epoch": 2.8754611372598906, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7657580375671387, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8860776424407959, "num_tokens": 858350567.0, "step": 22604 }, { "epoch": 2.875588347538481, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6582565307617188, "learning_rate": 1e-05, "loss": 0.4383, "mean_token_accuracy": 0.8782662153244019, "num_tokens": 858393998.0, "step": 22605 }, { "epoch": 2.8757155578170717, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.758626937866211, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8916153907775879, "num_tokens": 858430987.0, "step": 22606 }, { "epoch": 2.875842768095662, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6718034744262695, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8752743005752563, "num_tokens": 858464230.0, "step": 22607 }, { "epoch": 2.8759699783742527, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.763786792755127, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8740338087081909, "num_tokens": 858505965.0, "step": 22608 }, { "epoch": 2.8760971886528433, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.655017614364624, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8824318647384644, "num_tokens": 858542629.0, "step": 22609 }, { "epoch": 2.876224398931434, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7561933994293213, "learning_rate": 1e-05, "loss": 0.3893, "mean_token_accuracy": 0.8898863196372986, "num_tokens": 858583377.0, "step": 22610 }, { "epoch": 2.8763516092100243, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6526992321014404, "learning_rate": 1e-05, "loss": 0.3876, "mean_token_accuracy": 0.8924791216850281, "num_tokens": 858620546.0, "step": 22611 }, { "epoch": 2.876478819488615, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.774420976638794, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.8724510073661804, "num_tokens": 858659469.0, "step": 22612 }, { "epoch": 2.8766060297672054, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6673059463500977, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.879069447517395, "num_tokens": 858694519.0, "step": 22613 }, { "epoch": 2.8767332400457954, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7680773735046387, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8874594569206238, "num_tokens": 858728371.0, "step": 22614 }, { "epoch": 2.8768604503243864, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.666945457458496, "learning_rate": 1e-05, "loss": 0.3993, "mean_token_accuracy": 0.8863211870193481, "num_tokens": 858765195.0, "step": 22615 }, { "epoch": 2.8769876606029765, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7610552310943604, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8883577585220337, "num_tokens": 858804129.0, "step": 22616 }, { "epoch": 2.8771148708815675, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.658618450164795, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8877862691879272, "num_tokens": 858841267.0, "step": 22617 }, { "epoch": 2.8772420811601576, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.766792058944702, "learning_rate": 1e-05, "loss": 0.4238, "mean_token_accuracy": 0.8791740536689758, "num_tokens": 858882405.0, "step": 22618 }, { "epoch": 2.877369291438748, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.654920816421509, "learning_rate": 1e-05, "loss": 0.4105, "mean_token_accuracy": 0.883058488368988, "num_tokens": 858920176.0, "step": 22619 }, { "epoch": 2.8774965017173386, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7789342403411865, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8747686743736267, "num_tokens": 858959065.0, "step": 22620 }, { "epoch": 2.877623711995929, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.644383430480957, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8882266283035278, "num_tokens": 858997497.0, "step": 22621 }, { "epoch": 2.8777509222745197, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7750391960144043, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8851093649864197, "num_tokens": 859036425.0, "step": 22622 }, { "epoch": 2.87787813255311, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.666208505630493, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.8783425688743591, "num_tokens": 859073602.0, "step": 22623 }, { "epoch": 2.8780053428317007, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7855169773101807, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8751442432403564, "num_tokens": 859107576.0, "step": 22624 }, { "epoch": 2.8781325531102913, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.671450614929199, "learning_rate": 1e-05, "loss": 0.4488, "mean_token_accuracy": 0.8715382814407349, "num_tokens": 859141866.0, "step": 22625 }, { "epoch": 2.878259763388882, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7839832305908203, "learning_rate": 1e-05, "loss": 0.3935, "mean_token_accuracy": 0.8900636434555054, "num_tokens": 859180592.0, "step": 22626 }, { "epoch": 2.8783869736674723, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.668776273727417, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.874993085861206, "num_tokens": 859216365.0, "step": 22627 }, { "epoch": 2.878514183946063, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7817137241363525, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8780561685562134, "num_tokens": 859254167.0, "step": 22628 }, { "epoch": 2.8786413942246534, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6566014289855957, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8862171173095703, "num_tokens": 859288980.0, "step": 22629 }, { "epoch": 2.878768604503244, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7720484733581543, "learning_rate": 1e-05, "loss": 0.3823, "mean_token_accuracy": 0.8942626714706421, "num_tokens": 859326741.0, "step": 22630 }, { "epoch": 2.8788958147818344, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6598668098449707, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8739929795265198, "num_tokens": 859369530.0, "step": 22631 }, { "epoch": 2.879023025060425, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.778512716293335, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.8632960915565491, "num_tokens": 859409870.0, "step": 22632 }, { "epoch": 2.8791502353390155, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.8107783794403076, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8653690218925476, "num_tokens": 859443419.0, "step": 22633 }, { "epoch": 2.879277445617606, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7253406047821045, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8853870034217834, "num_tokens": 859474832.0, "step": 22634 }, { "epoch": 2.8794046558961965, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.69966197013855, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8777250051498413, "num_tokens": 859517668.0, "step": 22635 }, { "epoch": 2.879531866174787, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7035863399505615, "learning_rate": 1e-05, "loss": 0.4168, "mean_token_accuracy": 0.8845679759979248, "num_tokens": 859555954.0, "step": 22636 }, { "epoch": 2.8796590764533776, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.712174415588379, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.874592661857605, "num_tokens": 859591221.0, "step": 22637 }, { "epoch": 2.879786286731968, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6959774494171143, "learning_rate": 1e-05, "loss": 0.3733, "mean_token_accuracy": 0.8931381702423096, "num_tokens": 859631879.0, "step": 22638 }, { "epoch": 2.879913497010558, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.721632957458496, "learning_rate": 1e-05, "loss": 0.4592, "mean_token_accuracy": 0.8732521533966064, "num_tokens": 859669259.0, "step": 22639 }, { "epoch": 2.880040707289149, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.689877510070801, "learning_rate": 1e-05, "loss": 0.3967, "mean_token_accuracy": 0.8860125541687012, "num_tokens": 859706168.0, "step": 22640 }, { "epoch": 2.8801679175677393, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.721118450164795, "learning_rate": 1e-05, "loss": 0.4147, "mean_token_accuracy": 0.8801454901695251, "num_tokens": 859743182.0, "step": 22641 }, { "epoch": 2.8802951278463302, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.679422378540039, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8827677965164185, "num_tokens": 859786120.0, "step": 22642 }, { "epoch": 2.8804223381249203, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7259886264801025, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8855459690093994, "num_tokens": 859824097.0, "step": 22643 }, { "epoch": 2.880549548403511, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7032670974731445, "learning_rate": 1e-05, "loss": 0.4535, "mean_token_accuracy": 0.8682603240013123, "num_tokens": 859860081.0, "step": 22644 }, { "epoch": 2.8806767586821014, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.715651273727417, "learning_rate": 1e-05, "loss": 0.4414, "mean_token_accuracy": 0.8729324340820312, "num_tokens": 859898731.0, "step": 22645 }, { "epoch": 2.880803968960692, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6931798458099365, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8799751996994019, "num_tokens": 859938406.0, "step": 22646 }, { "epoch": 2.8809311792392824, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7130866050720215, "learning_rate": 1e-05, "loss": 0.3834, "mean_token_accuracy": 0.8929039239883423, "num_tokens": 859974270.0, "step": 22647 }, { "epoch": 2.881058389517873, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.701211929321289, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.882999062538147, "num_tokens": 860007261.0, "step": 22648 }, { "epoch": 2.8811855997964635, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6974592208862305, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.8793057799339294, "num_tokens": 860051170.0, "step": 22649 }, { "epoch": 2.881312810075054, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7258660793304443, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8844714164733887, "num_tokens": 860082771.0, "step": 22650 }, { "epoch": 2.8814400203536445, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6939656734466553, "learning_rate": 1e-05, "loss": 0.4681, "mean_token_accuracy": 0.8671050071716309, "num_tokens": 860124184.0, "step": 22651 }, { "epoch": 2.881567230632235, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.73511004447937, "learning_rate": 1e-05, "loss": 0.4582, "mean_token_accuracy": 0.8675447702407837, "num_tokens": 860158453.0, "step": 22652 }, { "epoch": 2.8816944409108256, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.686230182647705, "learning_rate": 1e-05, "loss": 0.4062, "mean_token_accuracy": 0.8863953351974487, "num_tokens": 860195403.0, "step": 22653 }, { "epoch": 2.881821651189416, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.727285385131836, "learning_rate": 1e-05, "loss": 0.4437, "mean_token_accuracy": 0.8704532384872437, "num_tokens": 860232813.0, "step": 22654 }, { "epoch": 2.8819488614680067, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.684346914291382, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8766276240348816, "num_tokens": 860269542.0, "step": 22655 }, { "epoch": 2.882076071746597, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.747225522994995, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8734560012817383, "num_tokens": 860305386.0, "step": 22656 }, { "epoch": 2.8822032820251877, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6724743843078613, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8799307942390442, "num_tokens": 860343137.0, "step": 22657 }, { "epoch": 2.8823304923037782, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.745784282684326, "learning_rate": 1e-05, "loss": 0.4407, "mean_token_accuracy": 0.8738366365432739, "num_tokens": 860382968.0, "step": 22658 }, { "epoch": 2.8824577025823688, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.665877103805542, "learning_rate": 1e-05, "loss": 0.4467, "mean_token_accuracy": 0.8750172853469849, "num_tokens": 860426985.0, "step": 22659 }, { "epoch": 2.8825849128609593, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7599942684173584, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8702445030212402, "num_tokens": 860460950.0, "step": 22660 }, { "epoch": 2.88271212313955, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.661966562271118, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8852328062057495, "num_tokens": 860499328.0, "step": 22661 }, { "epoch": 2.88283933341814, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.775083303451538, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8789485692977905, "num_tokens": 860532411.0, "step": 22662 }, { "epoch": 2.882966543696731, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.657623291015625, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8908381462097168, "num_tokens": 860567565.0, "step": 22663 }, { "epoch": 2.883093753975321, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7596778869628906, "learning_rate": 1e-05, "loss": 0.444, "mean_token_accuracy": 0.8733222484588623, "num_tokens": 860611687.0, "step": 22664 }, { "epoch": 2.883220964253912, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6603214740753174, "learning_rate": 1e-05, "loss": 0.4737, "mean_token_accuracy": 0.8638754487037659, "num_tokens": 860652220.0, "step": 22665 }, { "epoch": 2.883348174532502, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.760563611984253, "learning_rate": 1e-05, "loss": 0.4044, "mean_token_accuracy": 0.8866554498672485, "num_tokens": 860689641.0, "step": 22666 }, { "epoch": 2.883475384811093, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.661921501159668, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8833123445510864, "num_tokens": 860727490.0, "step": 22667 }, { "epoch": 2.883602595089683, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7673771381378174, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8744021654129028, "num_tokens": 860768625.0, "step": 22668 }, { "epoch": 2.8837298053682736, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.652350425720215, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8859201669692993, "num_tokens": 860806294.0, "step": 22669 }, { "epoch": 2.883857015646864, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7877109050750732, "learning_rate": 1e-05, "loss": 0.4701, "mean_token_accuracy": 0.8636096119880676, "num_tokens": 860842672.0, "step": 22670 }, { "epoch": 2.8839842259254547, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6565351486206055, "learning_rate": 1e-05, "loss": 0.4398, "mean_token_accuracy": 0.8709077835083008, "num_tokens": 860879882.0, "step": 22671 }, { "epoch": 2.884111436204045, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7785189151763916, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8797087669372559, "num_tokens": 860916099.0, "step": 22672 }, { "epoch": 2.8842386464826357, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6494927406311035, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8795844316482544, "num_tokens": 860957982.0, "step": 22673 }, { "epoch": 2.8843658567612263, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7862207889556885, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8884545564651489, "num_tokens": 860995528.0, "step": 22674 }, { "epoch": 2.884493067039817, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.646254777908325, "learning_rate": 1e-05, "loss": 0.4328, "mean_token_accuracy": 0.8774709701538086, "num_tokens": 861034676.0, "step": 22675 }, { "epoch": 2.8846202773184073, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.810884714126587, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8713421821594238, "num_tokens": 861069767.0, "step": 22676 }, { "epoch": 2.884747487596998, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.637592077255249, "learning_rate": 1e-05, "loss": 0.4134, "mean_token_accuracy": 0.8812024593353271, "num_tokens": 861109787.0, "step": 22677 }, { "epoch": 2.8848746978755884, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.810858726501465, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.8830894231796265, "num_tokens": 861144603.0, "step": 22678 }, { "epoch": 2.885001908154179, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6285674571990967, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8871645927429199, "num_tokens": 861184750.0, "step": 22679 }, { "epoch": 2.8851291184327694, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8228189945220947, "learning_rate": 1e-05, "loss": 0.4302, "mean_token_accuracy": 0.8771803975105286, "num_tokens": 861223591.0, "step": 22680 }, { "epoch": 2.88525632871136, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.64269757270813, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8750946521759033, "num_tokens": 861260456.0, "step": 22681 }, { "epoch": 2.8853835389899505, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8330132961273193, "learning_rate": 1e-05, "loss": 0.4315, "mean_token_accuracy": 0.8770119547843933, "num_tokens": 861298266.0, "step": 22682 }, { "epoch": 2.885510749268541, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6493887901306152, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8847521543502808, "num_tokens": 861334646.0, "step": 22683 }, { "epoch": 2.8856379595471315, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.831815481185913, "learning_rate": 1e-05, "loss": 0.4471, "mean_token_accuracy": 0.873166561126709, "num_tokens": 861370734.0, "step": 22684 }, { "epoch": 2.885765169825722, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.648481607437134, "learning_rate": 1e-05, "loss": 0.3749, "mean_token_accuracy": 0.8962479829788208, "num_tokens": 861403111.0, "step": 22685 }, { "epoch": 2.8858923801043126, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.82853102684021, "learning_rate": 1e-05, "loss": 0.4107, "mean_token_accuracy": 0.8831547498703003, "num_tokens": 861440668.0, "step": 22686 }, { "epoch": 2.8860195903829027, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6581356525421143, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8698570132255554, "num_tokens": 861478586.0, "step": 22687 }, { "epoch": 2.8861468006614936, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.826481580734253, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8784971237182617, "num_tokens": 861511637.0, "step": 22688 }, { "epoch": 2.8862740109400837, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6624703407287598, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8794009685516357, "num_tokens": 861545421.0, "step": 22689 }, { "epoch": 2.8864012212186747, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8216121196746826, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8743820786476135, "num_tokens": 861586230.0, "step": 22690 }, { "epoch": 2.886528431497265, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6562416553497314, "learning_rate": 1e-05, "loss": 0.4009, "mean_token_accuracy": 0.8876358270645142, "num_tokens": 861620556.0, "step": 22691 }, { "epoch": 2.8866556417758558, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8539035320281982, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8775936365127563, "num_tokens": 861655826.0, "step": 22692 }, { "epoch": 2.886782852054446, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6276636123657227, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8828266859054565, "num_tokens": 861699160.0, "step": 22693 }, { "epoch": 2.8869100623330364, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8501110076904297, "learning_rate": 1e-05, "loss": 0.4712, "mean_token_accuracy": 0.8710655570030212, "num_tokens": 861736784.0, "step": 22694 }, { "epoch": 2.887037272611627, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6409897804260254, "learning_rate": 1e-05, "loss": 0.3919, "mean_token_accuracy": 0.8905472159385681, "num_tokens": 861775359.0, "step": 22695 }, { "epoch": 2.8871644828902174, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.850773334503174, "learning_rate": 1e-05, "loss": 0.4761, "mean_token_accuracy": 0.8672725558280945, "num_tokens": 861813852.0, "step": 22696 }, { "epoch": 2.887291693168808, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.644709348678589, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8787280321121216, "num_tokens": 861853730.0, "step": 22697 }, { "epoch": 2.8874189034473985, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.847951650619507, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8772855997085571, "num_tokens": 861891405.0, "step": 22698 }, { "epoch": 2.887546113725989, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6481504440307617, "learning_rate": 1e-05, "loss": 0.392, "mean_token_accuracy": 0.8885543346405029, "num_tokens": 861929595.0, "step": 22699 }, { "epoch": 2.8876733240045795, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.859241008758545, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8679596185684204, "num_tokens": 861963263.0, "step": 22700 }, { "epoch": 2.88780053428317, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.653095006942749, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8720036149024963, "num_tokens": 862002259.0, "step": 22701 }, { "epoch": 2.8879277445617606, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8560657501220703, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8635676503181458, "num_tokens": 862039869.0, "step": 22702 }, { "epoch": 2.888054954840351, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6747238636016846, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8801071643829346, "num_tokens": 862068979.0, "step": 22703 }, { "epoch": 2.8881821651189417, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8184444904327393, "learning_rate": 1e-05, "loss": 0.4683, "mean_token_accuracy": 0.8659103512763977, "num_tokens": 862117384.0, "step": 22704 }, { "epoch": 2.888309375397532, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.666672706604004, "learning_rate": 1e-05, "loss": 0.4561, "mean_token_accuracy": 0.8691257238388062, "num_tokens": 862157961.0, "step": 22705 }, { "epoch": 2.8884365856761227, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8087596893310547, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8810974955558777, "num_tokens": 862199023.0, "step": 22706 }, { "epoch": 2.8885637959547132, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.662184953689575, "learning_rate": 1e-05, "loss": 0.4215, "mean_token_accuracy": 0.8804947733879089, "num_tokens": 862237839.0, "step": 22707 }, { "epoch": 2.8886910062333038, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8139901161193848, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8754749298095703, "num_tokens": 862273060.0, "step": 22708 }, { "epoch": 2.8888182165118943, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6585144996643066, "learning_rate": 1e-05, "loss": 0.4499, "mean_token_accuracy": 0.8699694871902466, "num_tokens": 862316469.0, "step": 22709 }, { "epoch": 2.888945426790485, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.817305564880371, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8714720010757446, "num_tokens": 862359367.0, "step": 22710 }, { "epoch": 2.8890726370690754, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6507327556610107, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8856064677238464, "num_tokens": 862397113.0, "step": 22711 }, { "epoch": 2.8891998473476654, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8195464611053467, "learning_rate": 1e-05, "loss": 0.4469, "mean_token_accuracy": 0.8746660351753235, "num_tokens": 862435627.0, "step": 22712 }, { "epoch": 2.8893270576262564, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6518008708953857, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.880980372428894, "num_tokens": 862476948.0, "step": 22713 }, { "epoch": 2.8894542679048465, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8189728260040283, "learning_rate": 1e-05, "loss": 0.4384, "mean_token_accuracy": 0.8768548965454102, "num_tokens": 862513429.0, "step": 22714 }, { "epoch": 2.8895814781834375, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.671625852584839, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8678798675537109, "num_tokens": 862548570.0, "step": 22715 }, { "epoch": 2.8897086884620276, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8112287521362305, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8781170845031738, "num_tokens": 862586265.0, "step": 22716 }, { "epoch": 2.889835898740618, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.674488067626953, "learning_rate": 1e-05, "loss": 0.4658, "mean_token_accuracy": 0.8646560907363892, "num_tokens": 862620552.0, "step": 22717 }, { "epoch": 2.8899631090192086, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7894670963287354, "learning_rate": 1e-05, "loss": 0.3894, "mean_token_accuracy": 0.8912234902381897, "num_tokens": 862656449.0, "step": 22718 }, { "epoch": 2.890090319297799, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.663846731185913, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8706588745117188, "num_tokens": 862697612.0, "step": 22719 }, { "epoch": 2.8902175295763897, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7907872200012207, "learning_rate": 1e-05, "loss": 0.4606, "mean_token_accuracy": 0.8683339953422546, "num_tokens": 862742304.0, "step": 22720 }, { "epoch": 2.89034473985498, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.658864974975586, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8810518383979797, "num_tokens": 862781738.0, "step": 22721 }, { "epoch": 2.8904719501335707, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.799513578414917, "learning_rate": 1e-05, "loss": 0.4976, "mean_token_accuracy": 0.852039635181427, "num_tokens": 862823124.0, "step": 22722 }, { "epoch": 2.8905991604121613, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.676842451095581, "learning_rate": 1e-05, "loss": 0.4637, "mean_token_accuracy": 0.8669270277023315, "num_tokens": 862857706.0, "step": 22723 }, { "epoch": 2.890726370690752, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.780744791030884, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.861664354801178, "num_tokens": 862899966.0, "step": 22724 }, { "epoch": 2.8908535809693423, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6790239810943604, "learning_rate": 1e-05, "loss": 0.3959, "mean_token_accuracy": 0.8795802593231201, "num_tokens": 862931090.0, "step": 22725 }, { "epoch": 2.890980791247933, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 3.0884556770324707, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8853760957717896, "num_tokens": 862962833.0, "step": 22726 }, { "epoch": 2.8911080015265234, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.809213161468506, "learning_rate": 1e-05, "loss": 0.435, "mean_token_accuracy": 0.8737577199935913, "num_tokens": 863003769.0, "step": 22727 }, { "epoch": 2.891235211805114, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.792480707168579, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8797824382781982, "num_tokens": 863038130.0, "step": 22728 }, { "epoch": 2.8913624220837044, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6274406909942627, "learning_rate": 1e-05, "loss": 0.3835, "mean_token_accuracy": 0.8925859928131104, "num_tokens": 863077943.0, "step": 22729 }, { "epoch": 2.891489632362295, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.805793046951294, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8760311603546143, "num_tokens": 863113853.0, "step": 22730 }, { "epoch": 2.8916168426408855, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.6600019931793213, "learning_rate": 1e-05, "loss": 0.413, "mean_token_accuracy": 0.8856710195541382, "num_tokens": 863149407.0, "step": 22731 }, { "epoch": 2.891744052919476, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.80747389793396, "learning_rate": 1e-05, "loss": 0.4475, "mean_token_accuracy": 0.8735051155090332, "num_tokens": 863183324.0, "step": 22732 }, { "epoch": 2.8918712631980665, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.6537954807281494, "learning_rate": 1e-05, "loss": 0.3807, "mean_token_accuracy": 0.8933039307594299, "num_tokens": 863220400.0, "step": 22733 }, { "epoch": 2.891998473476657, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8071067333221436, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8862442970275879, "num_tokens": 863259565.0, "step": 22734 }, { "epoch": 2.8921256837552476, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.672865390777588, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8719604015350342, "num_tokens": 863293480.0, "step": 22735 }, { "epoch": 2.892252894033838, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.809267282485962, "learning_rate": 1e-05, "loss": 0.4442, "mean_token_accuracy": 0.8776934742927551, "num_tokens": 863331702.0, "step": 22736 }, { "epoch": 2.892380104312428, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.6781890392303467, "learning_rate": 1e-05, "loss": 0.4501, "mean_token_accuracy": 0.8735082745552063, "num_tokens": 863366195.0, "step": 22737 }, { "epoch": 2.892507314591019, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8013758659362793, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8743342161178589, "num_tokens": 863406674.0, "step": 22738 }, { "epoch": 2.8926345248696093, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.682420492172241, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8775017261505127, "num_tokens": 863441837.0, "step": 22739 }, { "epoch": 2.8927617351482002, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.796603202819824, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.8860499858856201, "num_tokens": 863476461.0, "step": 22740 }, { "epoch": 2.8928889454267903, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.669102191925049, "learning_rate": 1e-05, "loss": 0.4463, "mean_token_accuracy": 0.8704018592834473, "num_tokens": 863513715.0, "step": 22741 }, { "epoch": 2.893016155705381, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.78416109085083, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8832749724388123, "num_tokens": 863553172.0, "step": 22742 }, { "epoch": 2.8931433659839714, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.6867287158966064, "learning_rate": 1e-05, "loss": 0.4609, "mean_token_accuracy": 0.8667259812355042, "num_tokens": 863585353.0, "step": 22743 }, { "epoch": 2.893270576262562, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7678143978118896, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8722496032714844, "num_tokens": 863627510.0, "step": 22744 }, { "epoch": 2.8933977865411524, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.684124708175659, "learning_rate": 1e-05, "loss": 0.4526, "mean_token_accuracy": 0.8716621994972229, "num_tokens": 863670683.0, "step": 22745 }, { "epoch": 2.893524996819743, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.747307777404785, "learning_rate": 1e-05, "loss": 0.4438, "mean_token_accuracy": 0.8733850717544556, "num_tokens": 863710793.0, "step": 22746 }, { "epoch": 2.8936522070983335, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.696059226989746, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.8849369883537292, "num_tokens": 863745963.0, "step": 22747 }, { "epoch": 2.893779417376924, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.737461566925049, "learning_rate": 1e-05, "loss": 0.3901, "mean_token_accuracy": 0.8895018696784973, "num_tokens": 863783729.0, "step": 22748 }, { "epoch": 2.8939066276555145, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6861369609832764, "learning_rate": 1e-05, "loss": 0.4174, "mean_token_accuracy": 0.8796460628509521, "num_tokens": 863820070.0, "step": 22749 }, { "epoch": 2.894033837934105, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7442426681518555, "learning_rate": 1e-05, "loss": 0.4027, "mean_token_accuracy": 0.8825825452804565, "num_tokens": 863853438.0, "step": 22750 }, { "epoch": 2.8941610482126956, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.69244384765625, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.8812606334686279, "num_tokens": 863894151.0, "step": 22751 }, { "epoch": 2.894288258491286, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7437453269958496, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8857403993606567, "num_tokens": 863929712.0, "step": 22752 }, { "epoch": 2.8944154687698767, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.690892219543457, "learning_rate": 1e-05, "loss": 0.4782, "mean_token_accuracy": 0.8616044521331787, "num_tokens": 863970236.0, "step": 22753 }, { "epoch": 2.894542679048467, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.76025652885437, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8693885803222656, "num_tokens": 864009164.0, "step": 22754 }, { "epoch": 2.8946698893270577, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6735668182373047, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8740795850753784, "num_tokens": 864048941.0, "step": 22755 }, { "epoch": 2.8947970996056482, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7691776752471924, "learning_rate": 1e-05, "loss": 0.4572, "mean_token_accuracy": 0.8743369579315186, "num_tokens": 864084041.0, "step": 22756 }, { "epoch": 2.8949243098842388, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.667527675628662, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8821625709533691, "num_tokens": 864118370.0, "step": 22757 }, { "epoch": 2.8950515201628293, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7774689197540283, "learning_rate": 1e-05, "loss": 0.3854, "mean_token_accuracy": 0.8909348249435425, "num_tokens": 864148888.0, "step": 22758 }, { "epoch": 2.89517873044142, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6564137935638428, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8826565742492676, "num_tokens": 864185850.0, "step": 22759 }, { "epoch": 2.89530594072001, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.793154239654541, "learning_rate": 1e-05, "loss": 0.4736, "mean_token_accuracy": 0.862682580947876, "num_tokens": 864224914.0, "step": 22760 }, { "epoch": 2.895433150998601, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6445720195770264, "learning_rate": 1e-05, "loss": 0.4293, "mean_token_accuracy": 0.8748045563697815, "num_tokens": 864266430.0, "step": 22761 }, { "epoch": 2.895560361277191, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8007025718688965, "learning_rate": 1e-05, "loss": 0.4229, "mean_token_accuracy": 0.8753247857093811, "num_tokens": 864299769.0, "step": 22762 }, { "epoch": 2.895687571555782, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6438395977020264, "learning_rate": 1e-05, "loss": 0.4382, "mean_token_accuracy": 0.8727779984474182, "num_tokens": 864340713.0, "step": 22763 }, { "epoch": 2.895814781834372, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8112263679504395, "learning_rate": 1e-05, "loss": 0.4615, "mean_token_accuracy": 0.8683099746704102, "num_tokens": 864382930.0, "step": 22764 }, { "epoch": 2.895941992112963, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.631453514099121, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8877263069152832, "num_tokens": 864424632.0, "step": 22765 }, { "epoch": 2.896069202391553, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.809338331222534, "learning_rate": 1e-05, "loss": 0.3998, "mean_token_accuracy": 0.8907830119132996, "num_tokens": 864463444.0, "step": 22766 }, { "epoch": 2.8961964126701436, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6503570079803467, "learning_rate": 1e-05, "loss": 0.484, "mean_token_accuracy": 0.8602049350738525, "num_tokens": 864506651.0, "step": 22767 }, { "epoch": 2.896323622948734, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.811920166015625, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8837564587593079, "num_tokens": 864544075.0, "step": 22768 }, { "epoch": 2.8964508332273247, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.652217388153076, "learning_rate": 1e-05, "loss": 0.3761, "mean_token_accuracy": 0.8952181339263916, "num_tokens": 864579814.0, "step": 22769 }, { "epoch": 2.896578043505915, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.817558765411377, "learning_rate": 1e-05, "loss": 0.4862, "mean_token_accuracy": 0.8603610992431641, "num_tokens": 864616690.0, "step": 22770 }, { "epoch": 2.8967052537845057, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.630885362625122, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8837347030639648, "num_tokens": 864662035.0, "step": 22771 }, { "epoch": 2.8968324640630962, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.839148759841919, "learning_rate": 1e-05, "loss": 0.4654, "mean_token_accuracy": 0.8680397272109985, "num_tokens": 864695219.0, "step": 22772 }, { "epoch": 2.8969596743416868, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.657637119293213, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8809688091278076, "num_tokens": 864731990.0, "step": 22773 }, { "epoch": 2.8970868846202773, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.824528932571411, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8816578388214111, "num_tokens": 864769642.0, "step": 22774 }, { "epoch": 2.897214094898868, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.656606912612915, "learning_rate": 1e-05, "loss": 0.4434, "mean_token_accuracy": 0.8744441866874695, "num_tokens": 864805329.0, "step": 22775 }, { "epoch": 2.8973413051774584, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.806976079940796, "learning_rate": 1e-05, "loss": 0.4036, "mean_token_accuracy": 0.8885157108306885, "num_tokens": 864847748.0, "step": 22776 }, { "epoch": 2.897468515456049, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6665635108947754, "learning_rate": 1e-05, "loss": 0.4619, "mean_token_accuracy": 0.8665061593055725, "num_tokens": 864884621.0, "step": 22777 }, { "epoch": 2.8975957257346394, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.79430890083313, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8866769075393677, "num_tokens": 864927398.0, "step": 22778 }, { "epoch": 2.89772293601323, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.671438455581665, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8890335559844971, "num_tokens": 864961038.0, "step": 22779 }, { "epoch": 2.8978501462918205, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8017144203186035, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8747683763504028, "num_tokens": 865000307.0, "step": 22780 }, { "epoch": 2.897977356570411, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6629509925842285, "learning_rate": 1e-05, "loss": 0.3893, "mean_token_accuracy": 0.8882737159729004, "num_tokens": 865035898.0, "step": 22781 }, { "epoch": 2.8981045668490015, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7838892936706543, "learning_rate": 1e-05, "loss": 0.3701, "mean_token_accuracy": 0.8987631797790527, "num_tokens": 865076325.0, "step": 22782 }, { "epoch": 2.898231777127592, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6623103618621826, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8733198046684265, "num_tokens": 865117290.0, "step": 22783 }, { "epoch": 2.8983589874061826, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7895572185516357, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8842965364456177, "num_tokens": 865153003.0, "step": 22784 }, { "epoch": 2.8984861976847727, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.649514675140381, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8745505809783936, "num_tokens": 865195113.0, "step": 22785 }, { "epoch": 2.8986134079633636, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.805058479309082, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8861612677574158, "num_tokens": 865231319.0, "step": 22786 }, { "epoch": 2.8987406182419537, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.661268711090088, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8794554471969604, "num_tokens": 865265862.0, "step": 22787 }, { "epoch": 2.8988678285205447, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8102285861968994, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.8749855160713196, "num_tokens": 865303873.0, "step": 22788 }, { "epoch": 2.898995038799135, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6619696617126465, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8882031440734863, "num_tokens": 865335163.0, "step": 22789 }, { "epoch": 2.8991222490777258, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.799248456954956, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8778131008148193, "num_tokens": 865373498.0, "step": 22790 }, { "epoch": 2.899249459356316, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.660189151763916, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8782875537872314, "num_tokens": 865411042.0, "step": 22791 }, { "epoch": 2.8993766696349064, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7911245822906494, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8803795576095581, "num_tokens": 865447676.0, "step": 22792 }, { "epoch": 2.899503879913497, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.649986982345581, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8747049570083618, "num_tokens": 865490905.0, "step": 22793 }, { "epoch": 2.8996310901920874, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.805318832397461, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.881507158279419, "num_tokens": 865530572.0, "step": 22794 }, { "epoch": 2.899758300470678, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6430814266204834, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.8823612928390503, "num_tokens": 865565999.0, "step": 22795 }, { "epoch": 2.8998855107492685, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.816650390625, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8722575902938843, "num_tokens": 865602455.0, "step": 22796 }, { "epoch": 2.900012721027859, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.638293504714966, "learning_rate": 1e-05, "loss": 0.3662, "mean_token_accuracy": 0.8991696238517761, "num_tokens": 865637723.0, "step": 22797 }, { "epoch": 2.9001399313064495, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8406765460968018, "learning_rate": 1e-05, "loss": 0.4458, "mean_token_accuracy": 0.8715031147003174, "num_tokens": 865669668.0, "step": 22798 }, { "epoch": 2.90026714158504, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6378750801086426, "learning_rate": 1e-05, "loss": 0.4227, "mean_token_accuracy": 0.8791317343711853, "num_tokens": 865708255.0, "step": 22799 }, { "epoch": 2.9003943518636306, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.826303482055664, "learning_rate": 1e-05, "loss": 0.4536, "mean_token_accuracy": 0.867908239364624, "num_tokens": 865749607.0, "step": 22800 }, { "epoch": 2.900521562142221, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.651451826095581, "learning_rate": 1e-05, "loss": 0.4565, "mean_token_accuracy": 0.8708252310752869, "num_tokens": 865786372.0, "step": 22801 }, { "epoch": 2.9006487724208116, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.826291084289551, "learning_rate": 1e-05, "loss": 0.4132, "mean_token_accuracy": 0.882125735282898, "num_tokens": 865822939.0, "step": 22802 }, { "epoch": 2.900775982699402, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.648704767227173, "learning_rate": 1e-05, "loss": 0.3871, "mean_token_accuracy": 0.8885061144828796, "num_tokens": 865859012.0, "step": 22803 }, { "epoch": 2.9009031929779927, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.821221113204956, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.8844360113143921, "num_tokens": 865897641.0, "step": 22804 }, { "epoch": 2.9010304032565832, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6404473781585693, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8768978118896484, "num_tokens": 865938270.0, "step": 22805 }, { "epoch": 2.9011576135351738, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.831134080886841, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8824794292449951, "num_tokens": 865971066.0, "step": 22806 }, { "epoch": 2.9012848238137643, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6468212604522705, "learning_rate": 1e-05, "loss": 0.3978, "mean_token_accuracy": 0.8884334564208984, "num_tokens": 866008595.0, "step": 22807 }, { "epoch": 2.901412034092355, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8593249320983887, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8821673393249512, "num_tokens": 866046650.0, "step": 22808 }, { "epoch": 2.9015392443709453, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6506333351135254, "learning_rate": 1e-05, "loss": 0.4498, "mean_token_accuracy": 0.8705543875694275, "num_tokens": 866086305.0, "step": 22809 }, { "epoch": 2.9016664546495354, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8352811336517334, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.8765659332275391, "num_tokens": 866125770.0, "step": 22810 }, { "epoch": 2.9017936649281264, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6492347717285156, "learning_rate": 1e-05, "loss": 0.3961, "mean_token_accuracy": 0.8883756399154663, "num_tokens": 866160777.0, "step": 22811 }, { "epoch": 2.9019208752067165, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8423984050750732, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8870347142219543, "num_tokens": 866195212.0, "step": 22812 }, { "epoch": 2.9020480854853075, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6422157287597656, "learning_rate": 1e-05, "loss": 0.4051, "mean_token_accuracy": 0.8847185969352722, "num_tokens": 866236252.0, "step": 22813 }, { "epoch": 2.9021752957638975, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8437275886535645, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.890313982963562, "num_tokens": 866275705.0, "step": 22814 }, { "epoch": 2.902302506042488, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.653010368347168, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8768805265426636, "num_tokens": 866316458.0, "step": 22815 }, { "epoch": 2.9024297163210786, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8418610095977783, "learning_rate": 1e-05, "loss": 0.4424, "mean_token_accuracy": 0.8733974695205688, "num_tokens": 866351107.0, "step": 22816 }, { "epoch": 2.902556926599669, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.655852794647217, "learning_rate": 1e-05, "loss": 0.378, "mean_token_accuracy": 0.895368218421936, "num_tokens": 866385381.0, "step": 22817 }, { "epoch": 2.9026841368782597, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8415744304656982, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.8866759538650513, "num_tokens": 866427018.0, "step": 22818 }, { "epoch": 2.90281134715685, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6368651390075684, "learning_rate": 1e-05, "loss": 0.376, "mean_token_accuracy": 0.8947681188583374, "num_tokens": 866463973.0, "step": 22819 }, { "epoch": 2.9029385574354407, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.846806764602661, "learning_rate": 1e-05, "loss": 0.4436, "mean_token_accuracy": 0.875519335269928, "num_tokens": 866500900.0, "step": 22820 }, { "epoch": 2.9030657677140312, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6355292797088623, "learning_rate": 1e-05, "loss": 0.3787, "mean_token_accuracy": 0.8955177068710327, "num_tokens": 866539757.0, "step": 22821 }, { "epoch": 2.9031929779926218, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8509886264801025, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8793240189552307, "num_tokens": 866577454.0, "step": 22822 }, { "epoch": 2.9033201882712123, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.645308017730713, "learning_rate": 1e-05, "loss": 0.3876, "mean_token_accuracy": 0.8912183046340942, "num_tokens": 866615032.0, "step": 22823 }, { "epoch": 2.903447398549803, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.792213439941406e-05, "grad_norm": 2.8664374351501465, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8738003969192505, "num_tokens": 866649700.0, "step": 22824 }, { "epoch": 2.9035746088283934, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.657149314880371, "learning_rate": 1e-05, "loss": 0.4339, "mean_token_accuracy": 0.8766718506813049, "num_tokens": 866684720.0, "step": 22825 }, { "epoch": 2.903701819106984, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.864412546157837, "learning_rate": 1e-05, "loss": 0.4359, "mean_token_accuracy": 0.8784485459327698, "num_tokens": 866722127.0, "step": 22826 }, { "epoch": 2.9038290293855744, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.6643552780151367, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8823871612548828, "num_tokens": 866757093.0, "step": 22827 }, { "epoch": 2.903956239664165, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8397316932678223, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8724728226661682, "num_tokens": 866796229.0, "step": 22828 }, { "epoch": 2.9040834499427555, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.66296124458313, "learning_rate": 1e-05, "loss": 0.4137, "mean_token_accuracy": 0.8826140761375427, "num_tokens": 866831867.0, "step": 22829 }, { "epoch": 2.904210660221346, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8133342266082764, "learning_rate": 1e-05, "loss": 0.449, "mean_token_accuracy": 0.871698260307312, "num_tokens": 866876246.0, "step": 22830 }, { "epoch": 2.9043378704999365, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.686345338821411, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8791148066520691, "num_tokens": 866909555.0, "step": 22831 }, { "epoch": 2.904465080778527, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7973389625549316, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8758918046951294, "num_tokens": 866945208.0, "step": 22832 }, { "epoch": 2.9045922910571176, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.686145305633545, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.883550763130188, "num_tokens": 866976832.0, "step": 22833 }, { "epoch": 2.904719501335708, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7759525775909424, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8765844702720642, "num_tokens": 867013964.0, "step": 22834 }, { "epoch": 2.904846711614298, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.8175125122070312, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.8779586553573608, "num_tokens": 867054361.0, "step": 22835 }, { "epoch": 2.904973921892889, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6858115196228027, "learning_rate": 1e-05, "loss": 0.3707, "mean_token_accuracy": 0.8954349756240845, "num_tokens": 867098732.0, "step": 22836 }, { "epoch": 2.9051011321714793, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7713518142700195, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.8640080690383911, "num_tokens": 867133839.0, "step": 22837 }, { "epoch": 2.9052283424500702, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6693549156188965, "learning_rate": 1e-05, "loss": 0.4851, "mean_token_accuracy": 0.8579219579696655, "num_tokens": 867175419.0, "step": 22838 }, { "epoch": 2.9053555527286603, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.76765513420105, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8737907409667969, "num_tokens": 867216585.0, "step": 22839 }, { "epoch": 2.905482763007251, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.660202741622925, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.885107159614563, "num_tokens": 867257228.0, "step": 22840 }, { "epoch": 2.9056099732858414, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.77107310295105, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.877653956413269, "num_tokens": 867297504.0, "step": 22841 }, { "epoch": 2.905737183564432, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6635661125183105, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8806633949279785, "num_tokens": 867336799.0, "step": 22842 }, { "epoch": 2.9058643938430224, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.781541109085083, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8789801597595215, "num_tokens": 867371337.0, "step": 22843 }, { "epoch": 2.905991604121613, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6544301509857178, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8826447129249573, "num_tokens": 867411870.0, "step": 22844 }, { "epoch": 2.9061188144002035, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.819265365600586, "learning_rate": 1e-05, "loss": 0.4558, "mean_token_accuracy": 0.8686227202415466, "num_tokens": 867450278.0, "step": 22845 }, { "epoch": 2.906246024678794, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6440746784210205, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8734607696533203, "num_tokens": 867490316.0, "step": 22846 }, { "epoch": 2.9063732349573845, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.806894063949585, "learning_rate": 1e-05, "loss": 0.402, "mean_token_accuracy": 0.8868656754493713, "num_tokens": 867526243.0, "step": 22847 }, { "epoch": 2.906500445235975, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6448261737823486, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8767998218536377, "num_tokens": 867568757.0, "step": 22848 }, { "epoch": 2.9066276555145656, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.824671506881714, "learning_rate": 1e-05, "loss": 0.409, "mean_token_accuracy": 0.8844395875930786, "num_tokens": 867602164.0, "step": 22849 }, { "epoch": 2.906754865793156, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6454310417175293, "learning_rate": 1e-05, "loss": 0.4409, "mean_token_accuracy": 0.8728295564651489, "num_tokens": 867640722.0, "step": 22850 }, { "epoch": 2.9068820760717466, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8217215538024902, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8818125128746033, "num_tokens": 867679568.0, "step": 22851 }, { "epoch": 2.907009286350337, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6482436656951904, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8735002279281616, "num_tokens": 867714991.0, "step": 22852 }, { "epoch": 2.9071364966289277, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8204619884490967, "learning_rate": 1e-05, "loss": 0.4417, "mean_token_accuracy": 0.8737049102783203, "num_tokens": 867753180.0, "step": 22853 }, { "epoch": 2.9072637069075182, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.651155948638916, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8804396986961365, "num_tokens": 867789535.0, "step": 22854 }, { "epoch": 2.9073909171861088, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.828242301940918, "learning_rate": 1e-05, "loss": 0.4076, "mean_token_accuracy": 0.8829326629638672, "num_tokens": 867821830.0, "step": 22855 }, { "epoch": 2.9075181274646993, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.642759084701538, "learning_rate": 1e-05, "loss": 0.4128, "mean_token_accuracy": 0.8833078145980835, "num_tokens": 867859878.0, "step": 22856 }, { "epoch": 2.90764533774329, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8377411365509033, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8742913603782654, "num_tokens": 867897212.0, "step": 22857 }, { "epoch": 2.90777254802188, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.667180061340332, "learning_rate": 1e-05, "loss": 0.4584, "mean_token_accuracy": 0.8661661744117737, "num_tokens": 867929899.0, "step": 22858 }, { "epoch": 2.907899758300471, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.828763961791992, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8854935765266418, "num_tokens": 867962313.0, "step": 22859 }, { "epoch": 2.908026968579061, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.671546459197998, "learning_rate": 1e-05, "loss": 0.5002, "mean_token_accuracy": 0.8565402030944824, "num_tokens": 867998886.0, "step": 22860 }, { "epoch": 2.908154178857652, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8072845935821533, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8786860704421997, "num_tokens": 868039151.0, "step": 22861 }, { "epoch": 2.908281389136242, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6722283363342285, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8820763230323792, "num_tokens": 868071021.0, "step": 22862 }, { "epoch": 2.908408599414833, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.797168493270874, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8827507495880127, "num_tokens": 868109872.0, "step": 22863 }, { "epoch": 2.908535809693423, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.663097381591797, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8803085684776306, "num_tokens": 868150258.0, "step": 22864 }, { "epoch": 2.9086630199720136, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7845022678375244, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8680222630500793, "num_tokens": 868195202.0, "step": 22865 }, { "epoch": 2.908790230250604, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6900434494018555, "learning_rate": 1e-05, "loss": 0.3976, "mean_token_accuracy": 0.8896924257278442, "num_tokens": 868226458.0, "step": 22866 }, { "epoch": 2.9089174405291947, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7825777530670166, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8783535361289978, "num_tokens": 868262908.0, "step": 22867 }, { "epoch": 2.909044650807785, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6760661602020264, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8697304725646973, "num_tokens": 868300013.0, "step": 22868 }, { "epoch": 2.9091718610863757, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7661354541778564, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8812787532806396, "num_tokens": 868342794.0, "step": 22869 }, { "epoch": 2.9092990713649662, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6794369220733643, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8751266002655029, "num_tokens": 868382009.0, "step": 22870 }, { "epoch": 2.9094262816435568, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7827260494232178, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8669987916946411, "num_tokens": 868417770.0, "step": 22871 }, { "epoch": 2.9095534919221473, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6666176319122314, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8845332264900208, "num_tokens": 868455447.0, "step": 22872 }, { "epoch": 2.909680702200738, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.769357442855835, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8742642402648926, "num_tokens": 868500388.0, "step": 22873 }, { "epoch": 2.9098079124793284, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6688997745513916, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.87838214635849, "num_tokens": 868535707.0, "step": 22874 }, { "epoch": 2.909935122757919, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7864184379577637, "learning_rate": 1e-05, "loss": 0.4886, "mean_token_accuracy": 0.861242413520813, "num_tokens": 868576634.0, "step": 22875 }, { "epoch": 2.9100623330365094, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.650299549102783, "learning_rate": 1e-05, "loss": 0.4023, "mean_token_accuracy": 0.8859597444534302, "num_tokens": 868619914.0, "step": 22876 }, { "epoch": 2.9101895433151, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8009326457977295, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8823341131210327, "num_tokens": 868652279.0, "step": 22877 }, { "epoch": 2.9103167535936905, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6545779705047607, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8713640570640564, "num_tokens": 868689418.0, "step": 22878 }, { "epoch": 2.910443963872281, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7873075008392334, "learning_rate": 1e-05, "loss": 0.3796, "mean_token_accuracy": 0.8941025733947754, "num_tokens": 868726695.0, "step": 22879 }, { "epoch": 2.9105711741508715, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6552796363830566, "learning_rate": 1e-05, "loss": 0.4317, "mean_token_accuracy": 0.874064028263092, "num_tokens": 868766120.0, "step": 22880 }, { "epoch": 2.910698384429462, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.788066864013672, "learning_rate": 1e-05, "loss": 0.3906, "mean_token_accuracy": 0.8913388252258301, "num_tokens": 868806041.0, "step": 22881 }, { "epoch": 2.9108255947080526, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6459403038024902, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8783103227615356, "num_tokens": 868852980.0, "step": 22882 }, { "epoch": 2.9109528049866427, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.805955410003662, "learning_rate": 1e-05, "loss": 0.4617, "mean_token_accuracy": 0.8685792684555054, "num_tokens": 868890951.0, "step": 22883 }, { "epoch": 2.9110800152652336, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.659287214279175, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8713348507881165, "num_tokens": 868926412.0, "step": 22884 }, { "epoch": 2.9112072255438237, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.812319755554199, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8714316487312317, "num_tokens": 868965181.0, "step": 22885 }, { "epoch": 2.9113344358224147, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6427032947540283, "learning_rate": 1e-05, "loss": 0.4141, "mean_token_accuracy": 0.8825410604476929, "num_tokens": 869004635.0, "step": 22886 }, { "epoch": 2.9114616461010048, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.808828830718994, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8806114792823792, "num_tokens": 869043833.0, "step": 22887 }, { "epoch": 2.9115888563795957, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6657888889312744, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.870685338973999, "num_tokens": 869078868.0, "step": 22888 }, { "epoch": 2.911716066658186, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8071095943450928, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8837504386901855, "num_tokens": 869113444.0, "step": 22889 }, { "epoch": 2.9118432769367764, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6567718982696533, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8839821219444275, "num_tokens": 869147741.0, "step": 22890 }, { "epoch": 2.911970487215367, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8137741088867188, "learning_rate": 1e-05, "loss": 0.4357, "mean_token_accuracy": 0.8755521178245544, "num_tokens": 869180247.0, "step": 22891 }, { "epoch": 2.9120976974939574, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.633800983428955, "learning_rate": 1e-05, "loss": 0.3791, "mean_token_accuracy": 0.8945996761322021, "num_tokens": 869221602.0, "step": 22892 }, { "epoch": 2.912224907772548, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8158369064331055, "learning_rate": 1e-05, "loss": 0.4173, "mean_token_accuracy": 0.8793842196464539, "num_tokens": 869259601.0, "step": 22893 }, { "epoch": 2.9123521180511385, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6311748027801514, "learning_rate": 1e-05, "loss": 0.3824, "mean_token_accuracy": 0.8921979069709778, "num_tokens": 869303366.0, "step": 22894 }, { "epoch": 2.912479328329729, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8410561084747314, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8629665970802307, "num_tokens": 869338579.0, "step": 22895 }, { "epoch": 2.9126065386083195, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6409144401550293, "learning_rate": 1e-05, "loss": 0.4644, "mean_token_accuracy": 0.8668007850646973, "num_tokens": 869384129.0, "step": 22896 }, { "epoch": 2.91273374888691, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8411405086517334, "learning_rate": 1e-05, "loss": 0.4285, "mean_token_accuracy": 0.8787956237792969, "num_tokens": 869421752.0, "step": 22897 }, { "epoch": 2.9128609591655006, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6486618518829346, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.874153733253479, "num_tokens": 869459309.0, "step": 22898 }, { "epoch": 2.912988169444091, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.832357168197632, "learning_rate": 1e-05, "loss": 0.3969, "mean_token_accuracy": 0.8896914720535278, "num_tokens": 869497718.0, "step": 22899 }, { "epoch": 2.9131153797226816, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6428678035736084, "learning_rate": 1e-05, "loss": 0.452, "mean_token_accuracy": 0.8710962533950806, "num_tokens": 869540522.0, "step": 22900 }, { "epoch": 2.913242590001272, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.8416244983673096, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8706223964691162, "num_tokens": 869579610.0, "step": 22901 }, { "epoch": 2.9133698002798627, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6460816860198975, "learning_rate": 1e-05, "loss": 0.3985, "mean_token_accuracy": 0.8882383108139038, "num_tokens": 869619186.0, "step": 22902 }, { "epoch": 2.9134970105584532, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8440206050872803, "learning_rate": 1e-05, "loss": 0.445, "mean_token_accuracy": 0.8717288970947266, "num_tokens": 869657415.0, "step": 22903 }, { "epoch": 2.9136242208370438, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6399176120758057, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8759928941726685, "num_tokens": 869698319.0, "step": 22904 }, { "epoch": 2.9137514311156343, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.831416368484497, "learning_rate": 1e-05, "loss": 0.3996, "mean_token_accuracy": 0.8883123397827148, "num_tokens": 869738630.0, "step": 22905 }, { "epoch": 2.913878641394225, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6500532627105713, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8667703866958618, "num_tokens": 869780713.0, "step": 22906 }, { "epoch": 2.9140058516728153, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8321261405944824, "learning_rate": 1e-05, "loss": 0.3914, "mean_token_accuracy": 0.8916757702827454, "num_tokens": 869817046.0, "step": 22907 }, { "epoch": 2.9141330619514054, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.652700424194336, "learning_rate": 1e-05, "loss": 0.4108, "mean_token_accuracy": 0.8873864412307739, "num_tokens": 869855099.0, "step": 22908 }, { "epoch": 2.9142602722299964, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.843432664871216, "learning_rate": 1e-05, "loss": 0.442, "mean_token_accuracy": 0.8778457045555115, "num_tokens": 869893207.0, "step": 22909 }, { "epoch": 2.9143874825085865, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.659395933151245, "learning_rate": 1e-05, "loss": 0.4226, "mean_token_accuracy": 0.879914402961731, "num_tokens": 869929925.0, "step": 22910 }, { "epoch": 2.9145146927871775, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8182461261749268, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8837099075317383, "num_tokens": 869974607.0, "step": 22911 }, { "epoch": 2.9146419030657675, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.656609296798706, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8773440718650818, "num_tokens": 870013056.0, "step": 22912 }, { "epoch": 2.914769113344358, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8338615894317627, "learning_rate": 1e-05, "loss": 0.4531, "mean_token_accuracy": 0.8705950975418091, "num_tokens": 870053735.0, "step": 22913 }, { "epoch": 2.9148963236229486, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.658461809158325, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8864659667015076, "num_tokens": 870093607.0, "step": 22914 }, { "epoch": 2.915023533901539, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.819823741912842, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8789443969726562, "num_tokens": 870133118.0, "step": 22915 }, { "epoch": 2.9151507441801296, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6838877201080322, "learning_rate": 1e-05, "loss": 0.4172, "mean_token_accuracy": 0.8811535835266113, "num_tokens": 870169782.0, "step": 22916 }, { "epoch": 2.91527795445872, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7947657108306885, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8788274526596069, "num_tokens": 870206891.0, "step": 22917 }, { "epoch": 2.9154051647373107, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6690518856048584, "learning_rate": 1e-05, "loss": 0.4518, "mean_token_accuracy": 0.8715864419937134, "num_tokens": 870253035.0, "step": 22918 }, { "epoch": 2.9155323750159012, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7959396839141846, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.8791248202323914, "num_tokens": 870292286.0, "step": 22919 }, { "epoch": 2.9156595852944918, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.657700538635254, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8701601624488831, "num_tokens": 870335594.0, "step": 22920 }, { "epoch": 2.9157867955730823, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7841126918792725, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8816781044006348, "num_tokens": 870379465.0, "step": 22921 }, { "epoch": 2.915914005851673, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6743743419647217, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8890317678451538, "num_tokens": 870418060.0, "step": 22922 }, { "epoch": 2.9160412161302633, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7742459774017334, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.885866641998291, "num_tokens": 870455565.0, "step": 22923 }, { "epoch": 2.916168426408854, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6724562644958496, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8768868446350098, "num_tokens": 870492609.0, "step": 22924 }, { "epoch": 2.9162956366874444, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7776901721954346, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8784292936325073, "num_tokens": 870528150.0, "step": 22925 }, { "epoch": 2.916422846966035, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6647188663482666, "learning_rate": 1e-05, "loss": 0.4065, "mean_token_accuracy": 0.8841670751571655, "num_tokens": 870568188.0, "step": 22926 }, { "epoch": 2.9165500572446255, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7871923446655273, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8818597793579102, "num_tokens": 870601459.0, "step": 22927 }, { "epoch": 2.916677267523216, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6663260459899902, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8827427625656128, "num_tokens": 870638976.0, "step": 22928 }, { "epoch": 2.9168044778018065, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.796114206314087, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8654797077178955, "num_tokens": 870671163.0, "step": 22929 }, { "epoch": 2.916931688080397, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6682448387145996, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8795681595802307, "num_tokens": 870705806.0, "step": 22930 }, { "epoch": 2.917058898358987, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7787179946899414, "learning_rate": 1e-05, "loss": 0.4374, "mean_token_accuracy": 0.8745652437210083, "num_tokens": 870743045.0, "step": 22931 }, { "epoch": 2.917186108637578, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6701436042785645, "learning_rate": 1e-05, "loss": 0.3927, "mean_token_accuracy": 0.8879633545875549, "num_tokens": 870780965.0, "step": 22932 }, { "epoch": 2.917313318916168, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7791430950164795, "learning_rate": 1e-05, "loss": 0.3885, "mean_token_accuracy": 0.8918476104736328, "num_tokens": 870814694.0, "step": 22933 }, { "epoch": 2.917440529194759, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.681145191192627, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8733077645301819, "num_tokens": 870845654.0, "step": 22934 }, { "epoch": 2.9175677394733492, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.769226551055908, "learning_rate": 1e-05, "loss": 0.4209, "mean_token_accuracy": 0.8794231414794922, "num_tokens": 870881720.0, "step": 22935 }, { "epoch": 2.91769494975194, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6486151218414307, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8839619755744934, "num_tokens": 870927480.0, "step": 22936 }, { "epoch": 2.9178221600305303, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7713146209716797, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8838033080101013, "num_tokens": 870962407.0, "step": 22937 }, { "epoch": 2.917949370309121, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.670672655105591, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.887786865234375, "num_tokens": 870994638.0, "step": 22938 }, { "epoch": 2.9180765805877114, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7649168968200684, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8838587999343872, "num_tokens": 871032638.0, "step": 22939 }, { "epoch": 2.918203790866302, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.663083791732788, "learning_rate": 1e-05, "loss": 0.4233, "mean_token_accuracy": 0.8776594400405884, "num_tokens": 871070955.0, "step": 22940 }, { "epoch": 2.9183310011448924, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7585995197296143, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8787086009979248, "num_tokens": 871109282.0, "step": 22941 }, { "epoch": 2.918458211423483, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6599643230438232, "learning_rate": 1e-05, "loss": 0.4389, "mean_token_accuracy": 0.8710291385650635, "num_tokens": 871151214.0, "step": 22942 }, { "epoch": 2.9185854217020735, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7726516723632812, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8704983592033386, "num_tokens": 871192909.0, "step": 22943 }, { "epoch": 2.918712631980664, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6686549186706543, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8762651681900024, "num_tokens": 871234952.0, "step": 22944 }, { "epoch": 2.9188398422592545, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7751715183258057, "learning_rate": 1e-05, "loss": 0.4272, "mean_token_accuracy": 0.879605233669281, "num_tokens": 871270789.0, "step": 22945 }, { "epoch": 2.918967052537845, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6696712970733643, "learning_rate": 1e-05, "loss": 0.4545, "mean_token_accuracy": 0.8722212314605713, "num_tokens": 871312425.0, "step": 22946 }, { "epoch": 2.9190942628164356, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7693097591400146, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.8919914960861206, "num_tokens": 871347685.0, "step": 22947 }, { "epoch": 2.919221473095026, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.670599937438965, "learning_rate": 1e-05, "loss": 0.4189, "mean_token_accuracy": 0.8794553875923157, "num_tokens": 871381789.0, "step": 22948 }, { "epoch": 2.9193486833736166, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7710180282592773, "learning_rate": 1e-05, "loss": 0.389, "mean_token_accuracy": 0.8905333876609802, "num_tokens": 871417578.0, "step": 22949 }, { "epoch": 2.919475893652207, "ewc_loss": 0.0791015625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.220008850097656e-05, "grad_norm": 2.6632044315338135, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8773813247680664, "num_tokens": 871454816.0, "step": 22950 }, { "epoch": 2.9196031039307977, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.772325038909912, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8800857067108154, "num_tokens": 871488898.0, "step": 22951 }, { "epoch": 2.9197303142093882, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.64636492729187, "learning_rate": 1e-05, "loss": 0.4084, "mean_token_accuracy": 0.8861149549484253, "num_tokens": 871531895.0, "step": 22952 }, { "epoch": 2.9198575244879788, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.79044508934021, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8870955109596252, "num_tokens": 871566151.0, "step": 22953 }, { "epoch": 2.9199847347665693, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6637182235717773, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8772775530815125, "num_tokens": 871603777.0, "step": 22954 }, { "epoch": 2.92011194504516, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.79135799407959, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8834618926048279, "num_tokens": 871636118.0, "step": 22955 }, { "epoch": 2.92023915532375, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6616740226745605, "learning_rate": 1e-05, "loss": 0.4223, "mean_token_accuracy": 0.876844048500061, "num_tokens": 871670948.0, "step": 22956 }, { "epoch": 2.920366365602341, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7972943782806396, "learning_rate": 1e-05, "loss": 0.4547, "mean_token_accuracy": 0.8734147548675537, "num_tokens": 871707987.0, "step": 22957 }, { "epoch": 2.920493575880931, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6331162452697754, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8863102197647095, "num_tokens": 871753861.0, "step": 22958 }, { "epoch": 2.920620786159522, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8155057430267334, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8743980526924133, "num_tokens": 871792667.0, "step": 22959 }, { "epoch": 2.920747996438112, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.658426523208618, "learning_rate": 1e-05, "loss": 0.4804, "mean_token_accuracy": 0.8626662492752075, "num_tokens": 871830994.0, "step": 22960 }, { "epoch": 2.920875206716703, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.823436737060547, "learning_rate": 1e-05, "loss": 0.471, "mean_token_accuracy": 0.8655142784118652, "num_tokens": 871869984.0, "step": 22961 }, { "epoch": 2.921002416995293, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6573517322540283, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8677428364753723, "num_tokens": 871906757.0, "step": 22962 }, { "epoch": 2.9211296272738836, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8086602687835693, "learning_rate": 1e-05, "loss": 0.4196, "mean_token_accuracy": 0.8857898712158203, "num_tokens": 871945387.0, "step": 22963 }, { "epoch": 2.921256837552474, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6439552307128906, "learning_rate": 1e-05, "loss": 0.4202, "mean_token_accuracy": 0.8800750970840454, "num_tokens": 871981456.0, "step": 22964 }, { "epoch": 2.9213840478310646, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8065788745880127, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8857410550117493, "num_tokens": 872022953.0, "step": 22965 }, { "epoch": 2.921511258109655, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.645979166030884, "learning_rate": 1e-05, "loss": 0.4171, "mean_token_accuracy": 0.8793973326683044, "num_tokens": 872059103.0, "step": 22966 }, { "epoch": 2.9216384683882457, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8216779232025146, "learning_rate": 1e-05, "loss": 0.4103, "mean_token_accuracy": 0.8808051347732544, "num_tokens": 872091910.0, "step": 22967 }, { "epoch": 2.9217656786668362, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6455788612365723, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8783992528915405, "num_tokens": 872130863.0, "step": 22968 }, { "epoch": 2.9218928889454268, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8311336040496826, "learning_rate": 1e-05, "loss": 0.4648, "mean_token_accuracy": 0.8688489198684692, "num_tokens": 872168104.0, "step": 22969 }, { "epoch": 2.9220200992240173, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6735072135925293, "learning_rate": 1e-05, "loss": 0.4403, "mean_token_accuracy": 0.8741490244865417, "num_tokens": 872197554.0, "step": 22970 }, { "epoch": 2.922147309502608, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.809948205947876, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.879181444644928, "num_tokens": 872232283.0, "step": 22971 }, { "epoch": 2.9222745197811983, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6719250679016113, "learning_rate": 1e-05, "loss": 0.4085, "mean_token_accuracy": 0.8806860446929932, "num_tokens": 872263459.0, "step": 22972 }, { "epoch": 2.922401730059789, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.791963577270508, "learning_rate": 1e-05, "loss": 0.3933, "mean_token_accuracy": 0.8910081386566162, "num_tokens": 872303056.0, "step": 22973 }, { "epoch": 2.9225289403383794, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6697354316711426, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8759474158287048, "num_tokens": 872341030.0, "step": 22974 }, { "epoch": 2.92265615061697, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7817013263702393, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8731606602668762, "num_tokens": 872384006.0, "step": 22975 }, { "epoch": 2.9227833608955605, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.669137954711914, "learning_rate": 1e-05, "loss": 0.4533, "mean_token_accuracy": 0.8687878847122192, "num_tokens": 872425575.0, "step": 22976 }, { "epoch": 2.922910571174151, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7783923149108887, "learning_rate": 1e-05, "loss": 0.4123, "mean_token_accuracy": 0.8841206431388855, "num_tokens": 872463483.0, "step": 22977 }, { "epoch": 2.9230377814527415, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.67927622795105, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8782680630683899, "num_tokens": 872496066.0, "step": 22978 }, { "epoch": 2.923164991731332, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.781102418899536, "learning_rate": 1e-05, "loss": 0.4176, "mean_token_accuracy": 0.8821820020675659, "num_tokens": 872529330.0, "step": 22979 }, { "epoch": 2.9232922020099226, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6801955699920654, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8829470872879028, "num_tokens": 872563989.0, "step": 22980 }, { "epoch": 2.9234194122885127, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7639341354370117, "learning_rate": 1e-05, "loss": 0.4003, "mean_token_accuracy": 0.885648250579834, "num_tokens": 872601506.0, "step": 22981 }, { "epoch": 2.9235466225671036, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6862828731536865, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8830001354217529, "num_tokens": 872633258.0, "step": 22982 }, { "epoch": 2.9236738328456937, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7807133197784424, "learning_rate": 1e-05, "loss": 0.46, "mean_token_accuracy": 0.8670082092285156, "num_tokens": 872665803.0, "step": 22983 }, { "epoch": 2.9238010431242847, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6750400066375732, "learning_rate": 1e-05, "loss": 0.4262, "mean_token_accuracy": 0.877713143825531, "num_tokens": 872703607.0, "step": 22984 }, { "epoch": 2.9239282534028748, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.764758586883545, "learning_rate": 1e-05, "loss": 0.4367, "mean_token_accuracy": 0.8772467374801636, "num_tokens": 872741966.0, "step": 22985 }, { "epoch": 2.9240554636814657, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.693061351776123, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8821141123771667, "num_tokens": 872773052.0, "step": 22986 }, { "epoch": 2.924182673960056, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7539050579071045, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.869835615158081, "num_tokens": 872813667.0, "step": 22987 }, { "epoch": 2.9243098842386464, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6810779571533203, "learning_rate": 1e-05, "loss": 0.4693, "mean_token_accuracy": 0.868056058883667, "num_tokens": 872854320.0, "step": 22988 }, { "epoch": 2.924437094517237, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.748805284500122, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8797909617424011, "num_tokens": 872895190.0, "step": 22989 }, { "epoch": 2.9245643047958274, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.689406394958496, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8813902735710144, "num_tokens": 872932288.0, "step": 22990 }, { "epoch": 2.924691515074418, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.765026330947876, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8829456567764282, "num_tokens": 872963617.0, "step": 22991 }, { "epoch": 2.9248187253530085, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.688606023788452, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8726515769958496, "num_tokens": 872997635.0, "step": 22992 }, { "epoch": 2.924945935631599, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.752530097961426, "learning_rate": 1e-05, "loss": 0.4742, "mean_token_accuracy": 0.8638365268707275, "num_tokens": 873037275.0, "step": 22993 }, { "epoch": 2.9250731459101895, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6854875087738037, "learning_rate": 1e-05, "loss": 0.4449, "mean_token_accuracy": 0.8718616366386414, "num_tokens": 873074791.0, "step": 22994 }, { "epoch": 2.92520035618878, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.747122049331665, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8851710557937622, "num_tokens": 873111521.0, "step": 22995 }, { "epoch": 2.9253275664673706, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6936230659484863, "learning_rate": 1e-05, "loss": 0.4815, "mean_token_accuracy": 0.8598369359970093, "num_tokens": 873150007.0, "step": 22996 }, { "epoch": 2.925454776745961, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7431132793426514, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8796371817588806, "num_tokens": 873188360.0, "step": 22997 }, { "epoch": 2.9255819870245516, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6831247806549072, "learning_rate": 1e-05, "loss": 0.3865, "mean_token_accuracy": 0.8870751261711121, "num_tokens": 873222962.0, "step": 22998 }, { "epoch": 2.925709197303142, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7520720958709717, "learning_rate": 1e-05, "loss": 0.3994, "mean_token_accuracy": 0.886776328086853, "num_tokens": 873260579.0, "step": 22999 }, { "epoch": 2.9258364075817327, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.6932907104492188, "learning_rate": 1e-05, "loss": 0.3949, "mean_token_accuracy": 0.8879972100257874, "num_tokens": 873292862.0, "step": 23000 }, { "epoch": 2.925963617860323, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.747594118118286, "learning_rate": 1e-05, "loss": 0.4988, "mean_token_accuracy": 0.8550156354904175, "num_tokens": 873331824.0, "step": 23001 }, { "epoch": 2.9260908281389137, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6916089057922363, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.8782668113708496, "num_tokens": 873364902.0, "step": 23002 }, { "epoch": 2.9262180384175043, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.731448173522949, "learning_rate": 1e-05, "loss": 0.3812, "mean_token_accuracy": 0.8934429287910461, "num_tokens": 873398523.0, "step": 23003 }, { "epoch": 2.926345248696095, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6828060150146484, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8824092745780945, "num_tokens": 873435779.0, "step": 23004 }, { "epoch": 2.9264724589746853, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7237792015075684, "learning_rate": 1e-05, "loss": 0.4021, "mean_token_accuracy": 0.8896318674087524, "num_tokens": 873476706.0, "step": 23005 }, { "epoch": 2.9265996692532754, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7013564109802246, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8805966377258301, "num_tokens": 873507762.0, "step": 23006 }, { "epoch": 2.9267268795318664, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7237510681152344, "learning_rate": 1e-05, "loss": 0.3909, "mean_token_accuracy": 0.8876752257347107, "num_tokens": 873545692.0, "step": 23007 }, { "epoch": 2.9268540898104565, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6856637001037598, "learning_rate": 1e-05, "loss": 0.4158, "mean_token_accuracy": 0.8819693326950073, "num_tokens": 873589131.0, "step": 23008 }, { "epoch": 2.9269813000890474, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7376809120178223, "learning_rate": 1e-05, "loss": 0.477, "mean_token_accuracy": 0.8631591796875, "num_tokens": 873628593.0, "step": 23009 }, { "epoch": 2.9271085103676375, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6803386211395264, "learning_rate": 1e-05, "loss": 0.4466, "mean_token_accuracy": 0.8728703856468201, "num_tokens": 873671525.0, "step": 23010 }, { "epoch": 2.927235720646228, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7371115684509277, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8798438906669617, "num_tokens": 873710618.0, "step": 23011 }, { "epoch": 2.9273629309248186, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7061703205108643, "learning_rate": 1e-05, "loss": 0.4813, "mean_token_accuracy": 0.8639202117919922, "num_tokens": 873751731.0, "step": 23012 }, { "epoch": 2.927490141203409, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7336690425872803, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8845140933990479, "num_tokens": 873791948.0, "step": 23013 }, { "epoch": 2.9276173514819996, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6953859329223633, "learning_rate": 1e-05, "loss": 0.4433, "mean_token_accuracy": 0.8722851276397705, "num_tokens": 873834193.0, "step": 23014 }, { "epoch": 2.92774456176059, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7440743446350098, "learning_rate": 1e-05, "loss": 0.4465, "mean_token_accuracy": 0.8730393648147583, "num_tokens": 873870035.0, "step": 23015 }, { "epoch": 2.9278717720391807, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.701693534851074, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8848953247070312, "num_tokens": 873902007.0, "step": 23016 }, { "epoch": 2.9279989823177712, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.727849245071411, "learning_rate": 1e-05, "loss": 0.4835, "mean_token_accuracy": 0.860262930393219, "num_tokens": 873947063.0, "step": 23017 }, { "epoch": 2.9281261925963618, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.699106454849243, "learning_rate": 1e-05, "loss": 0.4603, "mean_token_accuracy": 0.869231641292572, "num_tokens": 873987542.0, "step": 23018 }, { "epoch": 2.9282534028749523, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7366654872894287, "learning_rate": 1e-05, "loss": 0.4253, "mean_token_accuracy": 0.8795064091682434, "num_tokens": 874020492.0, "step": 23019 }, { "epoch": 2.928380613153543, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.686643600463867, "learning_rate": 1e-05, "loss": 0.3892, "mean_token_accuracy": 0.8880075216293335, "num_tokens": 874061902.0, "step": 23020 }, { "epoch": 2.9285078234321333, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.739513874053955, "learning_rate": 1e-05, "loss": 0.3806, "mean_token_accuracy": 0.894227921962738, "num_tokens": 874093277.0, "step": 23021 }, { "epoch": 2.928635033710724, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.664198637008667, "learning_rate": 1e-05, "loss": 0.3946, "mean_token_accuracy": 0.8873453140258789, "num_tokens": 874136543.0, "step": 23022 }, { "epoch": 2.9287622439893144, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7412819862365723, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.8893927335739136, "num_tokens": 874180933.0, "step": 23023 }, { "epoch": 2.928889454267905, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6636409759521484, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8835424184799194, "num_tokens": 874221980.0, "step": 23024 }, { "epoch": 2.9290166645464955, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7762019634246826, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8820807337760925, "num_tokens": 874255814.0, "step": 23025 }, { "epoch": 2.929143874825086, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6628758907318115, "learning_rate": 1e-05, "loss": 0.4514, "mean_token_accuracy": 0.8717399835586548, "num_tokens": 874296230.0, "step": 23026 }, { "epoch": 2.9292710851036765, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.78175950050354, "learning_rate": 1e-05, "loss": 0.4616, "mean_token_accuracy": 0.8669491410255432, "num_tokens": 874331854.0, "step": 23027 }, { "epoch": 2.929398295382267, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6434192657470703, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8840012550354004, "num_tokens": 874373348.0, "step": 23028 }, { "epoch": 2.929525505660857, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.779280185699463, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8828567266464233, "num_tokens": 874415226.0, "step": 23029 }, { "epoch": 2.929652715939448, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6635522842407227, "learning_rate": 1e-05, "loss": 0.4852, "mean_token_accuracy": 0.8628253936767578, "num_tokens": 874451894.0, "step": 23030 }, { "epoch": 2.929779926218038, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.781256914138794, "learning_rate": 1e-05, "loss": 0.4152, "mean_token_accuracy": 0.8836101293563843, "num_tokens": 874490248.0, "step": 23031 }, { "epoch": 2.929907136496629, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6617164611816406, "learning_rate": 1e-05, "loss": 0.4191, "mean_token_accuracy": 0.882648229598999, "num_tokens": 874527421.0, "step": 23032 }, { "epoch": 2.9300343467752192, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7684671878814697, "learning_rate": 1e-05, "loss": 0.4491, "mean_token_accuracy": 0.8734908103942871, "num_tokens": 874570834.0, "step": 23033 }, { "epoch": 2.93016155705381, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6675302982330322, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8694623708724976, "num_tokens": 874607956.0, "step": 23034 }, { "epoch": 2.9302887673324003, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.775946855545044, "learning_rate": 1e-05, "loss": 0.4391, "mean_token_accuracy": 0.8738429546356201, "num_tokens": 874644599.0, "step": 23035 }, { "epoch": 2.930415977610991, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6664228439331055, "learning_rate": 1e-05, "loss": 0.4506, "mean_token_accuracy": 0.8728612661361694, "num_tokens": 874682214.0, "step": 23036 }, { "epoch": 2.9305431878895813, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.774589776992798, "learning_rate": 1e-05, "loss": 0.4595, "mean_token_accuracy": 0.8664601445198059, "num_tokens": 874719298.0, "step": 23037 }, { "epoch": 2.930670398168172, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6575043201446533, "learning_rate": 1e-05, "loss": 0.4005, "mean_token_accuracy": 0.8882627487182617, "num_tokens": 874757391.0, "step": 23038 }, { "epoch": 2.9307976084467624, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.795400381088257, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8842140436172485, "num_tokens": 874788307.0, "step": 23039 }, { "epoch": 2.930924818725353, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6551353931427, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8777170181274414, "num_tokens": 874825648.0, "step": 23040 }, { "epoch": 2.9310520290039435, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7942750453948975, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8754309415817261, "num_tokens": 874862275.0, "step": 23041 }, { "epoch": 2.931179239282534, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6630680561065674, "learning_rate": 1e-05, "loss": 0.4492, "mean_token_accuracy": 0.873497486114502, "num_tokens": 874898411.0, "step": 23042 }, { "epoch": 2.9313064495611245, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.814647674560547, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8801321387290955, "num_tokens": 874935599.0, "step": 23043 }, { "epoch": 2.931433659839715, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6606576442718506, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8762781023979187, "num_tokens": 874966302.0, "step": 23044 }, { "epoch": 2.9315608701183056, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.796034336090088, "learning_rate": 1e-05, "loss": 0.3849, "mean_token_accuracy": 0.8937585353851318, "num_tokens": 875003158.0, "step": 23045 }, { "epoch": 2.931688080396896, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6518115997314453, "learning_rate": 1e-05, "loss": 0.4213, "mean_token_accuracy": 0.8797479271888733, "num_tokens": 875042163.0, "step": 23046 }, { "epoch": 2.9318152906754866, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.7946996688842773, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8753688335418701, "num_tokens": 875081957.0, "step": 23047 }, { "epoch": 2.931942500954077, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6448445320129395, "learning_rate": 1e-05, "loss": 0.4127, "mean_token_accuracy": 0.8812116384506226, "num_tokens": 875122286.0, "step": 23048 }, { "epoch": 2.9320697112326677, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8043456077575684, "learning_rate": 1e-05, "loss": 0.4778, "mean_token_accuracy": 0.8631969690322876, "num_tokens": 875163320.0, "step": 23049 }, { "epoch": 2.932196921511258, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6366477012634277, "learning_rate": 1e-05, "loss": 0.4021, "mean_token_accuracy": 0.8869556188583374, "num_tokens": 875203819.0, "step": 23050 }, { "epoch": 2.9323241317898487, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.815089225769043, "learning_rate": 1e-05, "loss": 0.4154, "mean_token_accuracy": 0.881070077419281, "num_tokens": 875239028.0, "step": 23051 }, { "epoch": 2.9324513420684393, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6512813568115234, "learning_rate": 1e-05, "loss": 0.4614, "mean_token_accuracy": 0.8702921867370605, "num_tokens": 875276324.0, "step": 23052 }, { "epoch": 2.93257855234703, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8162951469421387, "learning_rate": 1e-05, "loss": 0.4057, "mean_token_accuracy": 0.8847628831863403, "num_tokens": 875316918.0, "step": 23053 }, { "epoch": 2.93270576262562, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.670616388320923, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8717520236968994, "num_tokens": 875351009.0, "step": 23054 }, { "epoch": 2.932832972904211, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.806896924972534, "learning_rate": 1e-05, "loss": 0.3932, "mean_token_accuracy": 0.8903506994247437, "num_tokens": 875386988.0, "step": 23055 }, { "epoch": 2.932960183182801, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6385037899017334, "learning_rate": 1e-05, "loss": 0.4118, "mean_token_accuracy": 0.8828610181808472, "num_tokens": 875429607.0, "step": 23056 }, { "epoch": 2.933087393461392, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8273208141326904, "learning_rate": 1e-05, "loss": 0.4151, "mean_token_accuracy": 0.8840681314468384, "num_tokens": 875461071.0, "step": 23057 }, { "epoch": 2.933214603739982, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6442575454711914, "learning_rate": 1e-05, "loss": 0.4101, "mean_token_accuracy": 0.8838198781013489, "num_tokens": 875498146.0, "step": 23058 }, { "epoch": 2.933341814018573, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.822291374206543, "learning_rate": 1e-05, "loss": 0.4298, "mean_token_accuracy": 0.878874659538269, "num_tokens": 875536155.0, "step": 23059 }, { "epoch": 2.933469024297163, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.665860414505005, "learning_rate": 1e-05, "loss": 0.4334, "mean_token_accuracy": 0.8752305507659912, "num_tokens": 875569868.0, "step": 23060 }, { "epoch": 2.9335962345757536, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8112878799438477, "learning_rate": 1e-05, "loss": 0.4292, "mean_token_accuracy": 0.8773990869522095, "num_tokens": 875609298.0, "step": 23061 }, { "epoch": 2.933723444854344, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6511566638946533, "learning_rate": 1e-05, "loss": 0.38, "mean_token_accuracy": 0.8905892968177795, "num_tokens": 875642837.0, "step": 23062 }, { "epoch": 2.9338506551329346, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8176729679107666, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.874590277671814, "num_tokens": 875683273.0, "step": 23063 }, { "epoch": 2.933977865411525, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6746180057525635, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.8640875816345215, "num_tokens": 875717461.0, "step": 23064 }, { "epoch": 2.9341050756901157, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.802677869796753, "learning_rate": 1e-05, "loss": 0.4538, "mean_token_accuracy": 0.8695593476295471, "num_tokens": 875756975.0, "step": 23065 }, { "epoch": 2.934232285968706, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6638638973236084, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8832379579544067, "num_tokens": 875793414.0, "step": 23066 }, { "epoch": 2.9343594962472968, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.838522434234619, "learning_rate": 1e-05, "loss": 0.4791, "mean_token_accuracy": 0.8600853085517883, "num_tokens": 875822477.0, "step": 23067 }, { "epoch": 2.9344867065258873, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.663296699523926, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8799916505813599, "num_tokens": 875856220.0, "step": 23068 }, { "epoch": 2.934613916804478, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8172836303710938, "learning_rate": 1e-05, "loss": 0.4242, "mean_token_accuracy": 0.8802050352096558, "num_tokens": 875890572.0, "step": 23069 }, { "epoch": 2.9347411270830683, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.64469051361084, "learning_rate": 1e-05, "loss": 0.4122, "mean_token_accuracy": 0.8878225088119507, "num_tokens": 875928231.0, "step": 23070 }, { "epoch": 2.934868337361659, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8437249660491943, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.871570348739624, "num_tokens": 875966211.0, "step": 23071 }, { "epoch": 2.9349955476402494, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6421968936920166, "learning_rate": 1e-05, "loss": 0.4186, "mean_token_accuracy": 0.8801386952400208, "num_tokens": 876004857.0, "step": 23072 }, { "epoch": 2.93512275791884, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8523733615875244, "learning_rate": 1e-05, "loss": 0.4705, "mean_token_accuracy": 0.8643532991409302, "num_tokens": 876041966.0, "step": 23073 }, { "epoch": 2.9352499681974304, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.656020402908325, "learning_rate": 1e-05, "loss": 0.4058, "mean_token_accuracy": 0.8851103782653809, "num_tokens": 876075732.0, "step": 23074 }, { "epoch": 2.935377178476021, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.8435964584350586, "learning_rate": 1e-05, "loss": 0.4476, "mean_token_accuracy": 0.8738738298416138, "num_tokens": 876116582.0, "step": 23075 }, { "epoch": 2.9355043887546115, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6455373764038086, "learning_rate": 1e-05, "loss": 0.3759, "mean_token_accuracy": 0.8928593993186951, "num_tokens": 876150122.0, "step": 23076 }, { "epoch": 2.935631599033202, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.832732677459717, "learning_rate": 1e-05, "loss": 0.4015, "mean_token_accuracy": 0.8889824748039246, "num_tokens": 876187902.0, "step": 23077 }, { "epoch": 2.9357588093117926, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.663316488265991, "learning_rate": 1e-05, "loss": 0.436, "mean_token_accuracy": 0.8765396475791931, "num_tokens": 876224304.0, "step": 23078 }, { "epoch": 2.9358860195903826, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.840562343597412, "learning_rate": 1e-05, "loss": 0.4803, "mean_token_accuracy": 0.8631263375282288, "num_tokens": 876261436.0, "step": 23079 }, { "epoch": 2.9360132298689736, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.657625198364258, "learning_rate": 1e-05, "loss": 0.4441, "mean_token_accuracy": 0.871554434299469, "num_tokens": 876299996.0, "step": 23080 }, { "epoch": 2.9361404401475637, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8321757316589355, "learning_rate": 1e-05, "loss": 0.4521, "mean_token_accuracy": 0.8691285848617554, "num_tokens": 876337793.0, "step": 23081 }, { "epoch": 2.9362676504261547, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.661770820617676, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8820043802261353, "num_tokens": 876371837.0, "step": 23082 }, { "epoch": 2.9363948607047448, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.8063442707061768, "learning_rate": 1e-05, "loss": 0.4092, "mean_token_accuracy": 0.8870968222618103, "num_tokens": 876414402.0, "step": 23083 }, { "epoch": 2.9365220709833357, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.659900188446045, "learning_rate": 1e-05, "loss": 0.3746, "mean_token_accuracy": 0.8955379128456116, "num_tokens": 876451790.0, "step": 23084 }, { "epoch": 2.936649281261926, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8292129039764404, "learning_rate": 1e-05, "loss": 0.4634, "mean_token_accuracy": 0.8634771704673767, "num_tokens": 876486891.0, "step": 23085 }, { "epoch": 2.9367764915405163, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.651195764541626, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8870247602462769, "num_tokens": 876525560.0, "step": 23086 }, { "epoch": 2.936903701819107, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.831594228744507, "learning_rate": 1e-05, "loss": 0.4551, "mean_token_accuracy": 0.8725674748420715, "num_tokens": 876569037.0, "step": 23087 }, { "epoch": 2.9370309120976974, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6560723781585693, "learning_rate": 1e-05, "loss": 0.4337, "mean_token_accuracy": 0.8733984231948853, "num_tokens": 876605354.0, "step": 23088 }, { "epoch": 2.937158122376288, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.81620717048645, "learning_rate": 1e-05, "loss": 0.4018, "mean_token_accuracy": 0.8864338397979736, "num_tokens": 876644251.0, "step": 23089 }, { "epoch": 2.9372853326548785, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6461188793182373, "learning_rate": 1e-05, "loss": 0.4001, "mean_token_accuracy": 0.8877177238464355, "num_tokens": 876686558.0, "step": 23090 }, { "epoch": 2.937412542933469, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.834864616394043, "learning_rate": 1e-05, "loss": 0.42, "mean_token_accuracy": 0.8788758516311646, "num_tokens": 876718629.0, "step": 23091 }, { "epoch": 2.9375397532120595, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.654587507247925, "learning_rate": 1e-05, "loss": 0.4557, "mean_token_accuracy": 0.8657550811767578, "num_tokens": 876758037.0, "step": 23092 }, { "epoch": 2.93766696349065, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8218445777893066, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8803566694259644, "num_tokens": 876795201.0, "step": 23093 }, { "epoch": 2.9377941737692406, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6474685668945312, "learning_rate": 1e-05, "loss": 0.387, "mean_token_accuracy": 0.8910811543464661, "num_tokens": 876833465.0, "step": 23094 }, { "epoch": 2.937921384047831, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.825410842895508, "learning_rate": 1e-05, "loss": 0.4343, "mean_token_accuracy": 0.8778660297393799, "num_tokens": 876869915.0, "step": 23095 }, { "epoch": 2.9380485943264216, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.653554916381836, "learning_rate": 1e-05, "loss": 0.4048, "mean_token_accuracy": 0.8852962255477905, "num_tokens": 876906811.0, "step": 23096 }, { "epoch": 2.938175804605012, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8325746059417725, "learning_rate": 1e-05, "loss": 0.4512, "mean_token_accuracy": 0.8704185485839844, "num_tokens": 876944247.0, "step": 23097 }, { "epoch": 2.9383030148836027, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6586108207702637, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8824946284294128, "num_tokens": 876978557.0, "step": 23098 }, { "epoch": 2.938430225162193, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.826371908187866, "learning_rate": 1e-05, "loss": 0.4746, "mean_token_accuracy": 0.8629416823387146, "num_tokens": 877021241.0, "step": 23099 }, { "epoch": 2.9385574354407837, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6599490642547607, "learning_rate": 1e-05, "loss": 0.447, "mean_token_accuracy": 0.8726221323013306, "num_tokens": 877060777.0, "step": 23100 }, { "epoch": 2.9386846457193743, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8294029235839844, "learning_rate": 1e-05, "loss": 0.4635, "mean_token_accuracy": 0.8695929050445557, "num_tokens": 877096955.0, "step": 23101 }, { "epoch": 2.938811855997965, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6671009063720703, "learning_rate": 1e-05, "loss": 0.4435, "mean_token_accuracy": 0.8761398196220398, "num_tokens": 877130202.0, "step": 23102 }, { "epoch": 2.9389390662765553, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.818535089492798, "learning_rate": 1e-05, "loss": 0.4459, "mean_token_accuracy": 0.8723185062408447, "num_tokens": 877168777.0, "step": 23103 }, { "epoch": 2.9390662765551454, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.651474952697754, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8769930601119995, "num_tokens": 877209737.0, "step": 23104 }, { "epoch": 2.9391934868337364, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8136298656463623, "learning_rate": 1e-05, "loss": 0.4236, "mean_token_accuracy": 0.880077600479126, "num_tokens": 877245826.0, "step": 23105 }, { "epoch": 2.9393206971123265, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6546099185943604, "learning_rate": 1e-05, "loss": 0.4245, "mean_token_accuracy": 0.884558379650116, "num_tokens": 877285249.0, "step": 23106 }, { "epoch": 2.9394479073909174, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8065760135650635, "learning_rate": 1e-05, "loss": 0.3766, "mean_token_accuracy": 0.8950914740562439, "num_tokens": 877321971.0, "step": 23107 }, { "epoch": 2.9395751176695075, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6614441871643066, "learning_rate": 1e-05, "loss": 0.4649, "mean_token_accuracy": 0.8749775886535645, "num_tokens": 877361479.0, "step": 23108 }, { "epoch": 2.939702327948098, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.801158905029297, "learning_rate": 1e-05, "loss": 0.427, "mean_token_accuracy": 0.8768699765205383, "num_tokens": 877400166.0, "step": 23109 }, { "epoch": 2.9398295382266886, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6690988540649414, "learning_rate": 1e-05, "loss": 0.4206, "mean_token_accuracy": 0.8811209797859192, "num_tokens": 877437755.0, "step": 23110 }, { "epoch": 2.939956748505279, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.807579517364502, "learning_rate": 1e-05, "loss": 0.4739, "mean_token_accuracy": 0.8653247356414795, "num_tokens": 877471496.0, "step": 23111 }, { "epoch": 2.9400839587838696, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6697094440460205, "learning_rate": 1e-05, "loss": 0.3863, "mean_token_accuracy": 0.8926119804382324, "num_tokens": 877504576.0, "step": 23112 }, { "epoch": 2.94021116906246, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7995331287384033, "learning_rate": 1e-05, "loss": 0.4473, "mean_token_accuracy": 0.8732919692993164, "num_tokens": 877542795.0, "step": 23113 }, { "epoch": 2.9403383793410507, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6584267616271973, "learning_rate": 1e-05, "loss": 0.4415, "mean_token_accuracy": 0.8741123676300049, "num_tokens": 877581731.0, "step": 23114 }, { "epoch": 2.940465589619641, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8033347129821777, "learning_rate": 1e-05, "loss": 0.4502, "mean_token_accuracy": 0.8729111552238464, "num_tokens": 877618405.0, "step": 23115 }, { "epoch": 2.9405927998982317, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6577038764953613, "learning_rate": 1e-05, "loss": 0.4378, "mean_token_accuracy": 0.8806016445159912, "num_tokens": 877657792.0, "step": 23116 }, { "epoch": 2.9407200101768223, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.808441638946533, "learning_rate": 1e-05, "loss": 0.395, "mean_token_accuracy": 0.8889418244361877, "num_tokens": 877688965.0, "step": 23117 }, { "epoch": 2.940847220455413, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6494827270507812, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8884814977645874, "num_tokens": 877726595.0, "step": 23118 }, { "epoch": 2.9409744307340033, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8003294467926025, "learning_rate": 1e-05, "loss": 0.4278, "mean_token_accuracy": 0.8804410696029663, "num_tokens": 877771713.0, "step": 23119 }, { "epoch": 2.941101641012594, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6623833179473877, "learning_rate": 1e-05, "loss": 0.4095, "mean_token_accuracy": 0.8839088678359985, "num_tokens": 877805048.0, "step": 23120 }, { "epoch": 2.9412288512911844, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.800966501235962, "learning_rate": 1e-05, "loss": 0.4042, "mean_token_accuracy": 0.8842228651046753, "num_tokens": 877840469.0, "step": 23121 }, { "epoch": 2.941356061569775, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6725451946258545, "learning_rate": 1e-05, "loss": 0.4394, "mean_token_accuracy": 0.8712159395217896, "num_tokens": 877876747.0, "step": 23122 }, { "epoch": 2.9414832718483654, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.788266181945801, "learning_rate": 1e-05, "loss": 0.4142, "mean_token_accuracy": 0.8830888271331787, "num_tokens": 877913871.0, "step": 23123 }, { "epoch": 2.941610482126956, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.681666374206543, "learning_rate": 1e-05, "loss": 0.5098, "mean_token_accuracy": 0.865433394908905, "num_tokens": 877954195.0, "step": 23124 }, { "epoch": 2.9417376924055465, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.788714647293091, "learning_rate": 1e-05, "loss": 0.4581, "mean_token_accuracy": 0.8649784326553345, "num_tokens": 877989124.0, "step": 23125 }, { "epoch": 2.941864902684137, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6721503734588623, "learning_rate": 1e-05, "loss": 0.4314, "mean_token_accuracy": 0.8780889511108398, "num_tokens": 878030188.0, "step": 23126 }, { "epoch": 2.941992112962727, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7555389404296875, "learning_rate": 1e-05, "loss": 0.3689, "mean_token_accuracy": 0.8981955051422119, "num_tokens": 878073839.0, "step": 23127 }, { "epoch": 2.942119323241318, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.673905611038208, "learning_rate": 1e-05, "loss": 0.414, "mean_token_accuracy": 0.8818212747573853, "num_tokens": 878111446.0, "step": 23128 }, { "epoch": 2.942246533519908, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7613437175750732, "learning_rate": 1e-05, "loss": 0.4821, "mean_token_accuracy": 0.8597055077552795, "num_tokens": 878154543.0, "step": 23129 }, { "epoch": 2.942373743798499, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6917903423309326, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8854507803916931, "num_tokens": 878194265.0, "step": 23130 }, { "epoch": 2.9425009540770892, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.769364595413208, "learning_rate": 1e-05, "loss": 0.4805, "mean_token_accuracy": 0.863675594329834, "num_tokens": 878231475.0, "step": 23131 }, { "epoch": 2.94262816435568, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.674246072769165, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8814835548400879, "num_tokens": 878269492.0, "step": 23132 }, { "epoch": 2.9427553746342703, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7635185718536377, "learning_rate": 1e-05, "loss": 0.468, "mean_token_accuracy": 0.866305410861969, "num_tokens": 878307047.0, "step": 23133 }, { "epoch": 2.942882584912861, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.672971725463867, "learning_rate": 1e-05, "loss": 0.438, "mean_token_accuracy": 0.8751343488693237, "num_tokens": 878343971.0, "step": 23134 }, { "epoch": 2.9430097951914513, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7722508907318115, "learning_rate": 1e-05, "loss": 0.4106, "mean_token_accuracy": 0.8840331435203552, "num_tokens": 878380858.0, "step": 23135 }, { "epoch": 2.943137005470042, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6618990898132324, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.8863098621368408, "num_tokens": 878420854.0, "step": 23136 }, { "epoch": 2.9432642157486324, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7754509449005127, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8683957457542419, "num_tokens": 878455465.0, "step": 23137 }, { "epoch": 2.943391426027223, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.672065258026123, "learning_rate": 1e-05, "loss": 0.3953, "mean_token_accuracy": 0.8903210163116455, "num_tokens": 878487122.0, "step": 23138 }, { "epoch": 2.9435186363058135, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7791900634765625, "learning_rate": 1e-05, "loss": 0.4722, "mean_token_accuracy": 0.8649004697799683, "num_tokens": 878523881.0, "step": 23139 }, { "epoch": 2.943645846584404, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6724698543548584, "learning_rate": 1e-05, "loss": 0.4493, "mean_token_accuracy": 0.8688064813613892, "num_tokens": 878564351.0, "step": 23140 }, { "epoch": 2.9437730568629945, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7581372261047363, "learning_rate": 1e-05, "loss": 0.3891, "mean_token_accuracy": 0.8903939127922058, "num_tokens": 878605759.0, "step": 23141 }, { "epoch": 2.943900267141585, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.666288137435913, "learning_rate": 1e-05, "loss": 0.4002, "mean_token_accuracy": 0.8881412744522095, "num_tokens": 878644975.0, "step": 23142 }, { "epoch": 2.9440274774201756, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.774630069732666, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8862448930740356, "num_tokens": 878679882.0, "step": 23143 }, { "epoch": 2.944154687698766, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6502223014831543, "learning_rate": 1e-05, "loss": 0.3786, "mean_token_accuracy": 0.8937429189682007, "num_tokens": 878723539.0, "step": 23144 }, { "epoch": 2.9442818979773566, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7761287689208984, "learning_rate": 1e-05, "loss": 0.398, "mean_token_accuracy": 0.8874440789222717, "num_tokens": 878761993.0, "step": 23145 }, { "epoch": 2.944409108255947, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.667144536972046, "learning_rate": 1e-05, "loss": 0.4183, "mean_token_accuracy": 0.8793020248413086, "num_tokens": 878794355.0, "step": 23146 }, { "epoch": 2.9445363185345377, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.799206256866455, "learning_rate": 1e-05, "loss": 0.4857, "mean_token_accuracy": 0.862486720085144, "num_tokens": 878828477.0, "step": 23147 }, { "epoch": 2.944663528813128, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6561267375946045, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8690130710601807, "num_tokens": 878868134.0, "step": 23148 }, { "epoch": 2.9447907390917187, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.789029359817505, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.883210301399231, "num_tokens": 878908912.0, "step": 23149 }, { "epoch": 2.9449179493703093, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6494839191436768, "learning_rate": 1e-05, "loss": 0.4258, "mean_token_accuracy": 0.8781816959381104, "num_tokens": 878947354.0, "step": 23150 }, { "epoch": 2.9450451596489, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7990431785583496, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8837459087371826, "num_tokens": 878985538.0, "step": 23151 }, { "epoch": 2.94517236992749, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6449520587921143, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8708063960075378, "num_tokens": 879028884.0, "step": 23152 }, { "epoch": 2.945299580206081, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7935478687286377, "learning_rate": 1e-05, "loss": 0.4097, "mean_token_accuracy": 0.8830651640892029, "num_tokens": 879064388.0, "step": 23153 }, { "epoch": 2.945426790484671, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.650395631790161, "learning_rate": 1e-05, "loss": 0.3992, "mean_token_accuracy": 0.8853144645690918, "num_tokens": 879100353.0, "step": 23154 }, { "epoch": 2.945554000763262, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.796750545501709, "learning_rate": 1e-05, "loss": 0.4088, "mean_token_accuracy": 0.8856315612792969, "num_tokens": 879138447.0, "step": 23155 }, { "epoch": 2.945681211041852, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6661369800567627, "learning_rate": 1e-05, "loss": 0.4688, "mean_token_accuracy": 0.8661215305328369, "num_tokens": 879172825.0, "step": 23156 }, { "epoch": 2.945808421320443, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7983579635620117, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8730087280273438, "num_tokens": 879212086.0, "step": 23157 }, { "epoch": 2.945935631599033, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6645381450653076, "learning_rate": 1e-05, "loss": 0.4157, "mean_token_accuracy": 0.8822149038314819, "num_tokens": 879247689.0, "step": 23158 }, { "epoch": 2.9460628418776236, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.790085792541504, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8888633251190186, "num_tokens": 879283510.0, "step": 23159 }, { "epoch": 2.946190052156214, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7240567207336426, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8787572383880615, "num_tokens": 879322037.0, "step": 23160 }, { "epoch": 2.9463172624348046, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.753890037536621, "learning_rate": 1e-05, "loss": 0.4477, "mean_token_accuracy": 0.8703601360321045, "num_tokens": 879366400.0, "step": 23161 }, { "epoch": 2.946444472713395, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.703240394592285, "learning_rate": 1e-05, "loss": 0.4899, "mean_token_accuracy": 0.8598169088363647, "num_tokens": 879397768.0, "step": 23162 }, { "epoch": 2.9465716829919857, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.742976427078247, "learning_rate": 1e-05, "loss": 0.4267, "mean_token_accuracy": 0.878031313419342, "num_tokens": 879436751.0, "step": 23163 }, { "epoch": 2.946698893270576, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.69374942779541, "learning_rate": 1e-05, "loss": 0.4759, "mean_token_accuracy": 0.8632785081863403, "num_tokens": 879477750.0, "step": 23164 }, { "epoch": 2.9468261035491667, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.730698823928833, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.876700758934021, "num_tokens": 879516049.0, "step": 23165 }, { "epoch": 2.9469533138277573, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7135326862335205, "learning_rate": 1e-05, "loss": 0.4698, "mean_token_accuracy": 0.8649682402610779, "num_tokens": 879553176.0, "step": 23166 }, { "epoch": 2.947080524106348, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7237839698791504, "learning_rate": 1e-05, "loss": 0.3813, "mean_token_accuracy": 0.8920469284057617, "num_tokens": 879588435.0, "step": 23167 }, { "epoch": 2.9472077343849383, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7122466564178467, "learning_rate": 1e-05, "loss": 0.3884, "mean_token_accuracy": 0.8917052745819092, "num_tokens": 879623232.0, "step": 23168 }, { "epoch": 2.947334944663529, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.717294931411743, "learning_rate": 1e-05, "loss": 0.4605, "mean_token_accuracy": 0.8686520457267761, "num_tokens": 879669470.0, "step": 23169 }, { "epoch": 2.9474621549421194, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.708557367324829, "learning_rate": 1e-05, "loss": 0.4028, "mean_token_accuracy": 0.8910210132598877, "num_tokens": 879708173.0, "step": 23170 }, { "epoch": 2.94758936522071, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.713369846343994, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8722310066223145, "num_tokens": 879749086.0, "step": 23171 }, { "epoch": 2.9477165754993004, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7205417156219482, "learning_rate": 1e-05, "loss": 0.4439, "mean_token_accuracy": 0.8793656826019287, "num_tokens": 879786521.0, "step": 23172 }, { "epoch": 2.947843785777891, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.7095324993133545, "learning_rate": 1e-05, "loss": 0.4066, "mean_token_accuracy": 0.8866103887557983, "num_tokens": 879820332.0, "step": 23173 }, { "epoch": 2.9479709960564815, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.774345636367798, "learning_rate": 1e-05, "loss": 0.4306, "mean_token_accuracy": 0.8763744235038757, "num_tokens": 879859224.0, "step": 23174 }, { "epoch": 2.948098206335072, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.685950517654419, "learning_rate": 1e-05, "loss": 0.403, "mean_token_accuracy": 0.883080005645752, "num_tokens": 879892177.0, "step": 23175 }, { "epoch": 2.9482254166136626, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.754175901412964, "learning_rate": 1e-05, "loss": 0.3757, "mean_token_accuracy": 0.8981587290763855, "num_tokens": 879924195.0, "step": 23176 }, { "epoch": 2.9483526268922526, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6649720668792725, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8761454820632935, "num_tokens": 879965679.0, "step": 23177 }, { "epoch": 2.9484798371708436, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7573373317718506, "learning_rate": 1e-05, "loss": 0.4034, "mean_token_accuracy": 0.8846322298049927, "num_tokens": 880006441.0, "step": 23178 }, { "epoch": 2.9486070474494337, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6762874126434326, "learning_rate": 1e-05, "loss": 0.4631, "mean_token_accuracy": 0.8673496842384338, "num_tokens": 880042823.0, "step": 23179 }, { "epoch": 2.9487342577280247, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.779496431350708, "learning_rate": 1e-05, "loss": 0.4217, "mean_token_accuracy": 0.8790104389190674, "num_tokens": 880073949.0, "step": 23180 }, { "epoch": 2.9488614680066147, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6543567180633545, "learning_rate": 1e-05, "loss": 0.3717, "mean_token_accuracy": 0.8972655534744263, "num_tokens": 880112921.0, "step": 23181 }, { "epoch": 2.9489886782852053, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7739808559417725, "learning_rate": 1e-05, "loss": 0.4554, "mean_token_accuracy": 0.870500385761261, "num_tokens": 880152384.0, "step": 23182 }, { "epoch": 2.949115888563796, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.679612159729004, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8780909776687622, "num_tokens": 880187107.0, "step": 23183 }, { "epoch": 2.9492430988423863, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7652320861816406, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8815944194793701, "num_tokens": 880226072.0, "step": 23184 }, { "epoch": 2.949370309120977, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.677039623260498, "learning_rate": 1e-05, "loss": 0.4166, "mean_token_accuracy": 0.8809620141983032, "num_tokens": 880259607.0, "step": 23185 }, { "epoch": 2.9494975193995674, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.756814956665039, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.871275782585144, "num_tokens": 880301238.0, "step": 23186 }, { "epoch": 2.949624729678158, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.712393045425415, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8718316555023193, "num_tokens": 880336449.0, "step": 23187 }, { "epoch": 2.9497519399567484, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7563812732696533, "learning_rate": 1e-05, "loss": 0.4387, "mean_token_accuracy": 0.8744776248931885, "num_tokens": 880373527.0, "step": 23188 }, { "epoch": 2.949879150235339, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6811881065368652, "learning_rate": 1e-05, "loss": 0.4208, "mean_token_accuracy": 0.879981517791748, "num_tokens": 880412676.0, "step": 23189 }, { "epoch": 2.9500063605139295, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.74898099899292, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8826820850372314, "num_tokens": 880448171.0, "step": 23190 }, { "epoch": 2.95013357079252, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.687413215637207, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8753840923309326, "num_tokens": 880487056.0, "step": 23191 }, { "epoch": 2.9502607810711106, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7377896308898926, "learning_rate": 1e-05, "loss": 0.3725, "mean_token_accuracy": 0.8979377746582031, "num_tokens": 880520990.0, "step": 23192 }, { "epoch": 2.950387991349701, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.699998378753662, "learning_rate": 1e-05, "loss": 0.4594, "mean_token_accuracy": 0.8667482733726501, "num_tokens": 880557919.0, "step": 23193 }, { "epoch": 2.9505152016282916, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.751410961151123, "learning_rate": 1e-05, "loss": 0.3939, "mean_token_accuracy": 0.889662504196167, "num_tokens": 880590578.0, "step": 23194 }, { "epoch": 2.950642411906882, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.67809796333313, "learning_rate": 1e-05, "loss": 0.4156, "mean_token_accuracy": 0.8807417154312134, "num_tokens": 880631607.0, "step": 23195 }, { "epoch": 2.9507696221854727, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.737072467803955, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8831799626350403, "num_tokens": 880674301.0, "step": 23196 }, { "epoch": 2.950896832464063, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6815879344940186, "learning_rate": 1e-05, "loss": 0.4346, "mean_token_accuracy": 0.8761754035949707, "num_tokens": 880714254.0, "step": 23197 }, { "epoch": 2.9510240427426537, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.738661766052246, "learning_rate": 1e-05, "loss": 0.3666, "mean_token_accuracy": 0.8983567357063293, "num_tokens": 880751562.0, "step": 23198 }, { "epoch": 2.9511512530212443, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7078800201416016, "learning_rate": 1e-05, "loss": 0.4574, "mean_token_accuracy": 0.8672170042991638, "num_tokens": 880785062.0, "step": 23199 }, { "epoch": 2.951278463299835, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.737677574157715, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8754662275314331, "num_tokens": 880820673.0, "step": 23200 }, { "epoch": 2.9514056735784253, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7030577659606934, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.874183177947998, "num_tokens": 880860707.0, "step": 23201 }, { "epoch": 2.9515328838570154, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7135868072509766, "learning_rate": 1e-05, "loss": 0.4116, "mean_token_accuracy": 0.8823487758636475, "num_tokens": 880903504.0, "step": 23202 }, { "epoch": 2.9516600941356064, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7036726474761963, "learning_rate": 1e-05, "loss": 0.3878, "mean_token_accuracy": 0.8900697827339172, "num_tokens": 880943414.0, "step": 23203 }, { "epoch": 2.9517873044141965, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.735705852508545, "learning_rate": 1e-05, "loss": 0.4121, "mean_token_accuracy": 0.886444628238678, "num_tokens": 880977736.0, "step": 23204 }, { "epoch": 2.9519145146927874, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.684048652648926, "learning_rate": 1e-05, "loss": 0.3782, "mean_token_accuracy": 0.893947958946228, "num_tokens": 881019793.0, "step": 23205 }, { "epoch": 2.9520417249713775, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7420766353607178, "learning_rate": 1e-05, "loss": 0.3932, "mean_token_accuracy": 0.8905131816864014, "num_tokens": 881057703.0, "step": 23206 }, { "epoch": 2.952168935249968, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6966793537139893, "learning_rate": 1e-05, "loss": 0.404, "mean_token_accuracy": 0.8854379057884216, "num_tokens": 881093649.0, "step": 23207 }, { "epoch": 2.9522961455285586, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7724132537841797, "learning_rate": 1e-05, "loss": 0.4775, "mean_token_accuracy": 0.8613916039466858, "num_tokens": 881125620.0, "step": 23208 }, { "epoch": 2.952423355807149, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6718170642852783, "learning_rate": 1e-05, "loss": 0.3863, "mean_token_accuracy": 0.8912982940673828, "num_tokens": 881165313.0, "step": 23209 }, { "epoch": 2.9525505660857396, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7768445014953613, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.88190758228302, "num_tokens": 881195390.0, "step": 23210 }, { "epoch": 2.95267777636433, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.65411639213562, "learning_rate": 1e-05, "loss": 0.4484, "mean_token_accuracy": 0.8724832534790039, "num_tokens": 881242005.0, "step": 23211 }, { "epoch": 2.9528049866429207, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.775123119354248, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8819580078125, "num_tokens": 881284587.0, "step": 23212 }, { "epoch": 2.952932196921511, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6531906127929688, "learning_rate": 1e-05, "loss": 0.4136, "mean_token_accuracy": 0.8844380378723145, "num_tokens": 881324174.0, "step": 23213 }, { "epoch": 2.9530594072001017, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8061883449554443, "learning_rate": 1e-05, "loss": 0.4802, "mean_token_accuracy": 0.8597249388694763, "num_tokens": 881360727.0, "step": 23214 }, { "epoch": 2.9531866174786923, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6440036296844482, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8795862197875977, "num_tokens": 881400482.0, "step": 23215 }, { "epoch": 2.953313827757283, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7994918823242188, "learning_rate": 1e-05, "loss": 0.4261, "mean_token_accuracy": 0.8793210983276367, "num_tokens": 881440157.0, "step": 23216 }, { "epoch": 2.9534410380358733, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6263387203216553, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.887749433517456, "num_tokens": 881486814.0, "step": 23217 }, { "epoch": 2.953568248314464, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.816134452819824, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8745303153991699, "num_tokens": 881526997.0, "step": 23218 }, { "epoch": 2.9536954585930544, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.636166572570801, "learning_rate": 1e-05, "loss": 0.4175, "mean_token_accuracy": 0.8802328705787659, "num_tokens": 881569103.0, "step": 23219 }, { "epoch": 2.953822668871645, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.824319362640381, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8842795491218567, "num_tokens": 881604734.0, "step": 23220 }, { "epoch": 2.9539498791502354, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.6385552883148193, "learning_rate": 1e-05, "loss": 0.4165, "mean_token_accuracy": 0.8812109231948853, "num_tokens": 881645788.0, "step": 23221 }, { "epoch": 2.954077089428826, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8470070362091064, "learning_rate": 1e-05, "loss": 0.3844, "mean_token_accuracy": 0.8907265663146973, "num_tokens": 881674738.0, "step": 23222 }, { "epoch": 2.9542042997074165, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.648454427719116, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8789669275283813, "num_tokens": 881711685.0, "step": 23223 }, { "epoch": 2.954331509986007, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.8299877643585205, "learning_rate": 1e-05, "loss": 0.3831, "mean_token_accuracy": 0.8931484222412109, "num_tokens": 881748165.0, "step": 23224 }, { "epoch": 2.954458720264597, "ewc_loss": 0.07958984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.2438507080078125e-05, "grad_norm": 2.648207187652588, "learning_rate": 1e-05, "loss": 0.4073, "mean_token_accuracy": 0.8844116926193237, "num_tokens": 881781332.0, "step": 23225 }, { "epoch": 2.954585930543188, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8356850147247314, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.880898118019104, "num_tokens": 881814526.0, "step": 23226 }, { "epoch": 2.954713140821778, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.645582437515259, "learning_rate": 1e-05, "loss": 0.4641, "mean_token_accuracy": 0.8674758076667786, "num_tokens": 881856625.0, "step": 23227 }, { "epoch": 2.954840351100369, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.833045244216919, "learning_rate": 1e-05, "loss": 0.415, "mean_token_accuracy": 0.8862466812133789, "num_tokens": 881895092.0, "step": 23228 }, { "epoch": 2.954967561378959, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.638671636581421, "learning_rate": 1e-05, "loss": 0.4198, "mean_token_accuracy": 0.8816558122634888, "num_tokens": 881936789.0, "step": 23229 }, { "epoch": 2.95509477165755, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.843992233276367, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8893102407455444, "num_tokens": 881973471.0, "step": 23230 }, { "epoch": 2.9552219819361403, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.650163412094116, "learning_rate": 1e-05, "loss": 0.4358, "mean_token_accuracy": 0.8750035762786865, "num_tokens": 882011039.0, "step": 23231 }, { "epoch": 2.955349192214731, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8340983390808105, "learning_rate": 1e-05, "loss": 0.4193, "mean_token_accuracy": 0.8804007768630981, "num_tokens": 882049221.0, "step": 23232 }, { "epoch": 2.9554764024933213, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6527371406555176, "learning_rate": 1e-05, "loss": 0.4054, "mean_token_accuracy": 0.8847395777702332, "num_tokens": 882086527.0, "step": 23233 }, { "epoch": 2.955603612771912, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8390109539031982, "learning_rate": 1e-05, "loss": 0.3599, "mean_token_accuracy": 0.8996292352676392, "num_tokens": 882118375.0, "step": 23234 }, { "epoch": 2.9557308230505024, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.669020414352417, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8779541254043579, "num_tokens": 882150604.0, "step": 23235 }, { "epoch": 2.955858033329093, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8204901218414307, "learning_rate": 1e-05, "loss": 0.4231, "mean_token_accuracy": 0.879156231880188, "num_tokens": 882185822.0, "step": 23236 }, { "epoch": 2.9559852436076834, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.659989833831787, "learning_rate": 1e-05, "loss": 0.4199, "mean_token_accuracy": 0.882698655128479, "num_tokens": 882224829.0, "step": 23237 }, { "epoch": 2.956112453886274, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8229241371154785, "learning_rate": 1e-05, "loss": 0.4692, "mean_token_accuracy": 0.8624043464660645, "num_tokens": 882263107.0, "step": 23238 }, { "epoch": 2.9562396641648645, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.67942214012146, "learning_rate": 1e-05, "loss": 0.4218, "mean_token_accuracy": 0.8795303702354431, "num_tokens": 882294302.0, "step": 23239 }, { "epoch": 2.956366874443455, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8108065128326416, "learning_rate": 1e-05, "loss": 0.4074, "mean_token_accuracy": 0.8842259645462036, "num_tokens": 882327081.0, "step": 23240 }, { "epoch": 2.9564940847220456, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.671220302581787, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.8842345476150513, "num_tokens": 882361793.0, "step": 23241 }, { "epoch": 2.956621295000636, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.790043592453003, "learning_rate": 1e-05, "loss": 0.3685, "mean_token_accuracy": 0.8963519334793091, "num_tokens": 882395255.0, "step": 23242 }, { "epoch": 2.9567485052792266, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6689319610595703, "learning_rate": 1e-05, "loss": 0.4228, "mean_token_accuracy": 0.8806968927383423, "num_tokens": 882431915.0, "step": 23243 }, { "epoch": 2.956875715557817, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7814362049102783, "learning_rate": 1e-05, "loss": 0.4263, "mean_token_accuracy": 0.8772932291030884, "num_tokens": 882472688.0, "step": 23244 }, { "epoch": 2.9570029258364077, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.681087017059326, "learning_rate": 1e-05, "loss": 0.4145, "mean_token_accuracy": 0.8815819621086121, "num_tokens": 882506851.0, "step": 23245 }, { "epoch": 2.957130136114998, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7868001461029053, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8720254302024841, "num_tokens": 882541935.0, "step": 23246 }, { "epoch": 2.9572573463935887, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.665355682373047, "learning_rate": 1e-05, "loss": 0.401, "mean_token_accuracy": 0.885158121585846, "num_tokens": 882577860.0, "step": 23247 }, { "epoch": 2.9573845566721793, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.793534278869629, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8658612966537476, "num_tokens": 882616182.0, "step": 23248 }, { "epoch": 2.95751176695077, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.671527862548828, "learning_rate": 1e-05, "loss": 0.482, "mean_token_accuracy": 0.8607522249221802, "num_tokens": 882658847.0, "step": 23249 }, { "epoch": 2.95763897722936, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7728428840637207, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.8757933974266052, "num_tokens": 882699842.0, "step": 23250 }, { "epoch": 2.957766187507951, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6821205615997314, "learning_rate": 1e-05, "loss": 0.3962, "mean_token_accuracy": 0.8864709734916687, "num_tokens": 882733143.0, "step": 23251 }, { "epoch": 2.957893397786541, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7861897945404053, "learning_rate": 1e-05, "loss": 0.4309, "mean_token_accuracy": 0.8752455115318298, "num_tokens": 882763233.0, "step": 23252 }, { "epoch": 2.958020608065132, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6575961112976074, "learning_rate": 1e-05, "loss": 0.4204, "mean_token_accuracy": 0.8811806440353394, "num_tokens": 882807612.0, "step": 23253 }, { "epoch": 2.958147818343722, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7767937183380127, "learning_rate": 1e-05, "loss": 0.3596, "mean_token_accuracy": 0.9006060361862183, "num_tokens": 882842135.0, "step": 23254 }, { "epoch": 2.958275028622313, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.665757656097412, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8836921453475952, "num_tokens": 882876687.0, "step": 23255 }, { "epoch": 2.958402238900903, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.780141592025757, "learning_rate": 1e-05, "loss": 0.4353, "mean_token_accuracy": 0.8757642507553101, "num_tokens": 882919142.0, "step": 23256 }, { "epoch": 2.9585294491794936, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.664954423904419, "learning_rate": 1e-05, "loss": 0.4052, "mean_token_accuracy": 0.8879984021186829, "num_tokens": 882955301.0, "step": 23257 }, { "epoch": 2.958656659458084, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7898318767547607, "learning_rate": 1e-05, "loss": 0.3833, "mean_token_accuracy": 0.8945857882499695, "num_tokens": 882985436.0, "step": 23258 }, { "epoch": 2.9587838697366746, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.658879518508911, "learning_rate": 1e-05, "loss": 0.4444, "mean_token_accuracy": 0.8759510517120361, "num_tokens": 883028082.0, "step": 23259 }, { "epoch": 2.958911080015265, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7944388389587402, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8785815238952637, "num_tokens": 883063771.0, "step": 23260 }, { "epoch": 2.9590382902938557, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6641111373901367, "learning_rate": 1e-05, "loss": 0.4024, "mean_token_accuracy": 0.8830692768096924, "num_tokens": 883096742.0, "step": 23261 }, { "epoch": 2.959165500572446, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7888262271881104, "learning_rate": 1e-05, "loss": 0.4408, "mean_token_accuracy": 0.8756555914878845, "num_tokens": 883137916.0, "step": 23262 }, { "epoch": 2.9592927108510367, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6713201999664307, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8747249841690063, "num_tokens": 883172730.0, "step": 23263 }, { "epoch": 2.9594199211296273, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7912769317626953, "learning_rate": 1e-05, "loss": 0.4319, "mean_token_accuracy": 0.8779585957527161, "num_tokens": 883212139.0, "step": 23264 }, { "epoch": 2.959547131408218, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.665947198867798, "learning_rate": 1e-05, "loss": 0.3997, "mean_token_accuracy": 0.8883289694786072, "num_tokens": 883244866.0, "step": 23265 }, { "epoch": 2.9596743416868083, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.795342206954956, "learning_rate": 1e-05, "loss": 0.4274, "mean_token_accuracy": 0.8835124373435974, "num_tokens": 883277595.0, "step": 23266 }, { "epoch": 2.959801551965399, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.669802188873291, "learning_rate": 1e-05, "loss": 0.4281, "mean_token_accuracy": 0.877936601638794, "num_tokens": 883311246.0, "step": 23267 }, { "epoch": 2.9599287622439894, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.783651828765869, "learning_rate": 1e-05, "loss": 0.4308, "mean_token_accuracy": 0.8785592317581177, "num_tokens": 883353351.0, "step": 23268 }, { "epoch": 2.96005597252258, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6619651317596436, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8850141763687134, "num_tokens": 883392245.0, "step": 23269 }, { "epoch": 2.9601831828011704, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.783393144607544, "learning_rate": 1e-05, "loss": 0.4049, "mean_token_accuracy": 0.8850992918014526, "num_tokens": 883430404.0, "step": 23270 }, { "epoch": 2.960310393079761, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6621763706207275, "learning_rate": 1e-05, "loss": 0.4918, "mean_token_accuracy": 0.8599170446395874, "num_tokens": 883477894.0, "step": 23271 }, { "epoch": 2.9604376033583515, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7893354892730713, "learning_rate": 1e-05, "loss": 0.4709, "mean_token_accuracy": 0.8639684915542603, "num_tokens": 883520396.0, "step": 23272 }, { "epoch": 2.960564813636942, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6679491996765137, "learning_rate": 1e-05, "loss": 0.4365, "mean_token_accuracy": 0.8777995109558105, "num_tokens": 883559120.0, "step": 23273 }, { "epoch": 2.9606920239155325, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7744860649108887, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8737051486968994, "num_tokens": 883603061.0, "step": 23274 }, { "epoch": 2.9608192341941226, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.681281089782715, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8776159286499023, "num_tokens": 883637674.0, "step": 23275 }, { "epoch": 2.9609464444727136, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.770789861679077, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8832723498344421, "num_tokens": 883674631.0, "step": 23276 }, { "epoch": 2.9610736547513037, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6867752075195312, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8764461874961853, "num_tokens": 883716823.0, "step": 23277 }, { "epoch": 2.9612008650298947, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.788379669189453, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8766855001449585, "num_tokens": 883751849.0, "step": 23278 }, { "epoch": 2.9613280753084847, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6598832607269287, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8884105682373047, "num_tokens": 883790698.0, "step": 23279 }, { "epoch": 2.9614552855870753, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7767906188964844, "learning_rate": 1e-05, "loss": 0.382, "mean_token_accuracy": 0.8930966854095459, "num_tokens": 883823238.0, "step": 23280 }, { "epoch": 2.961582495865666, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6618690490722656, "learning_rate": 1e-05, "loss": 0.4159, "mean_token_accuracy": 0.8814376592636108, "num_tokens": 883863036.0, "step": 23281 }, { "epoch": 2.9617097061442563, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.783999443054199, "learning_rate": 1e-05, "loss": 0.4478, "mean_token_accuracy": 0.8686572313308716, "num_tokens": 883902556.0, "step": 23282 }, { "epoch": 2.961836916422847, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.742358922958374, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8654404878616333, "num_tokens": 883937039.0, "step": 23283 }, { "epoch": 2.9619641267014374, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7573769092559814, "learning_rate": 1e-05, "loss": 0.425, "mean_token_accuracy": 0.8774280548095703, "num_tokens": 883969047.0, "step": 23284 }, { "epoch": 2.962091336980028, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6849257946014404, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8736466765403748, "num_tokens": 884013431.0, "step": 23285 }, { "epoch": 2.9622185472586184, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7554800510406494, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8766420483589172, "num_tokens": 884053714.0, "step": 23286 }, { "epoch": 2.962345757537209, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6862001419067383, "learning_rate": 1e-05, "loss": 0.4446, "mean_token_accuracy": 0.8718240261077881, "num_tokens": 884094624.0, "step": 23287 }, { "epoch": 2.9624729678157995, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7763984203338623, "learning_rate": 1e-05, "loss": 0.4322, "mean_token_accuracy": 0.876360297203064, "num_tokens": 884131759.0, "step": 23288 }, { "epoch": 2.96260017809439, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.668571710586548, "learning_rate": 1e-05, "loss": 0.4093, "mean_token_accuracy": 0.884028971195221, "num_tokens": 884175333.0, "step": 23289 }, { "epoch": 2.9627273883729806, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7733259201049805, "learning_rate": 1e-05, "loss": 0.3829, "mean_token_accuracy": 0.8934030532836914, "num_tokens": 884209778.0, "step": 23290 }, { "epoch": 2.962854598651571, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6741371154785156, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8806532621383667, "num_tokens": 884245581.0, "step": 23291 }, { "epoch": 2.9629818089301616, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7696332931518555, "learning_rate": 1e-05, "loss": 0.3958, "mean_token_accuracy": 0.8875517249107361, "num_tokens": 884284925.0, "step": 23292 }, { "epoch": 2.963109019208752, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6656086444854736, "learning_rate": 1e-05, "loss": 0.4053, "mean_token_accuracy": 0.8864036798477173, "num_tokens": 884326056.0, "step": 23293 }, { "epoch": 2.9632362294873427, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.7669310569763184, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8798630237579346, "num_tokens": 884368169.0, "step": 23294 }, { "epoch": 2.963363439765933, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.678478240966797, "learning_rate": 1e-05, "loss": 0.4369, "mean_token_accuracy": 0.8792464137077332, "num_tokens": 884402785.0, "step": 23295 }, { "epoch": 2.9634906500445237, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7629992961883545, "learning_rate": 1e-05, "loss": 0.4096, "mean_token_accuracy": 0.882117509841919, "num_tokens": 884444755.0, "step": 23296 }, { "epoch": 2.9636178603231143, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6601905822753906, "learning_rate": 1e-05, "loss": 0.4124, "mean_token_accuracy": 0.8841053247451782, "num_tokens": 884486299.0, "step": 23297 }, { "epoch": 2.963745070601705, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7712748050689697, "learning_rate": 1e-05, "loss": 0.4861, "mean_token_accuracy": 0.8590834736824036, "num_tokens": 884532076.0, "step": 23298 }, { "epoch": 2.9638722808802953, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.670444965362549, "learning_rate": 1e-05, "loss": 0.4295, "mean_token_accuracy": 0.8804250955581665, "num_tokens": 884569518.0, "step": 23299 }, { "epoch": 2.9639994911588854, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7853734493255615, "learning_rate": 1e-05, "loss": 0.4139, "mean_token_accuracy": 0.8820789456367493, "num_tokens": 884601173.0, "step": 23300 }, { "epoch": 2.9641267014374764, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6631805896759033, "learning_rate": 1e-05, "loss": 0.4161, "mean_token_accuracy": 0.8829200267791748, "num_tokens": 884637558.0, "step": 23301 }, { "epoch": 2.9642539117160664, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.782296657562256, "learning_rate": 1e-05, "loss": 0.4222, "mean_token_accuracy": 0.8811061382293701, "num_tokens": 884673377.0, "step": 23302 }, { "epoch": 2.9643811219946574, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.653881788253784, "learning_rate": 1e-05, "loss": 0.3828, "mean_token_accuracy": 0.8948106169700623, "num_tokens": 884710334.0, "step": 23303 }, { "epoch": 2.9645083322732475, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7909905910491943, "learning_rate": 1e-05, "loss": 0.3923, "mean_token_accuracy": 0.8912197947502136, "num_tokens": 884741515.0, "step": 23304 }, { "epoch": 2.964635542551838, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6553452014923096, "learning_rate": 1e-05, "loss": 0.4452, "mean_token_accuracy": 0.8724386692047119, "num_tokens": 884782204.0, "step": 23305 }, { "epoch": 2.9647627528304286, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7823469638824463, "learning_rate": 1e-05, "loss": 0.3979, "mean_token_accuracy": 0.8891664743423462, "num_tokens": 884819613.0, "step": 23306 }, { "epoch": 2.964889963109019, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6584110260009766, "learning_rate": 1e-05, "loss": 0.4324, "mean_token_accuracy": 0.8778692483901978, "num_tokens": 884857381.0, "step": 23307 }, { "epoch": 2.9650171733876096, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8054583072662354, "learning_rate": 1e-05, "loss": 0.4893, "mean_token_accuracy": 0.8570457696914673, "num_tokens": 884893562.0, "step": 23308 }, { "epoch": 2.9651443836662, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6470963954925537, "learning_rate": 1e-05, "loss": 0.4299, "mean_token_accuracy": 0.8785036206245422, "num_tokens": 884934359.0, "step": 23309 }, { "epoch": 2.9652715939447907, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.799110174179077, "learning_rate": 1e-05, "loss": 0.3947, "mean_token_accuracy": 0.8912686109542847, "num_tokens": 884974446.0, "step": 23310 }, { "epoch": 2.965398804223381, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6483571529388428, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8726450204849243, "num_tokens": 885016447.0, "step": 23311 }, { "epoch": 2.9655260145019717, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.818692684173584, "learning_rate": 1e-05, "loss": 0.4427, "mean_token_accuracy": 0.8730719685554504, "num_tokens": 885055064.0, "step": 23312 }, { "epoch": 2.9656532247805623, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6570072174072266, "learning_rate": 1e-05, "loss": 0.4185, "mean_token_accuracy": 0.8839066624641418, "num_tokens": 885092380.0, "step": 23313 }, { "epoch": 2.965780435059153, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8021066188812256, "learning_rate": 1e-05, "loss": 0.3998, "mean_token_accuracy": 0.8860986232757568, "num_tokens": 885130717.0, "step": 23314 }, { "epoch": 2.9659076453377433, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.654153823852539, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8815548419952393, "num_tokens": 885168291.0, "step": 23315 }, { "epoch": 2.966034855616334, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8204092979431152, "learning_rate": 1e-05, "loss": 0.483, "mean_token_accuracy": 0.8623560070991516, "num_tokens": 885203839.0, "step": 23316 }, { "epoch": 2.9661620658949244, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6559369564056396, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8704362511634827, "num_tokens": 885239336.0, "step": 23317 }, { "epoch": 2.966289276173515, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8165106773376465, "learning_rate": 1e-05, "loss": 0.4679, "mean_token_accuracy": 0.8614472150802612, "num_tokens": 885275964.0, "step": 23318 }, { "epoch": 2.9664164864521054, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.656012773513794, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8787060379981995, "num_tokens": 885307444.0, "step": 23319 }, { "epoch": 2.966543696730696, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8069615364074707, "learning_rate": 1e-05, "loss": 0.396, "mean_token_accuracy": 0.8897043466567993, "num_tokens": 885345379.0, "step": 23320 }, { "epoch": 2.9666709070092865, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.647469997406006, "learning_rate": 1e-05, "loss": 0.4038, "mean_token_accuracy": 0.8852570056915283, "num_tokens": 885384105.0, "step": 23321 }, { "epoch": 2.966798117287877, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.000370025634765625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.833134412765503, "learning_rate": 1e-05, "loss": 0.4291, "mean_token_accuracy": 0.8792855143547058, "num_tokens": 885420374.0, "step": 23322 }, { "epoch": 2.966925327566467, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6595795154571533, "learning_rate": 1e-05, "loss": 0.4211, "mean_token_accuracy": 0.8832480907440186, "num_tokens": 885453450.0, "step": 23323 }, { "epoch": 2.967052537845058, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8322789669036865, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8829236626625061, "num_tokens": 885486331.0, "step": 23324 }, { "epoch": 2.967179748123648, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6518144607543945, "learning_rate": 1e-05, "loss": 0.3958, "mean_token_accuracy": 0.8922283053398132, "num_tokens": 885520474.0, "step": 23325 }, { "epoch": 2.967306958402239, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8325014114379883, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.875993013381958, "num_tokens": 885563128.0, "step": 23326 }, { "epoch": 2.967434168680829, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.637080192565918, "learning_rate": 1e-05, "loss": 0.3674, "mean_token_accuracy": 0.900297224521637, "num_tokens": 885604219.0, "step": 23327 }, { "epoch": 2.96756137895942, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.841771125793457, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8849526643753052, "num_tokens": 885639932.0, "step": 23328 }, { "epoch": 2.9676885892380103, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6645519733428955, "learning_rate": 1e-05, "loss": 0.4593, "mean_token_accuracy": 0.8660721778869629, "num_tokens": 885675914.0, "step": 23329 }, { "epoch": 2.967815799516601, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 3.069157361984253, "learning_rate": 1e-05, "loss": 0.4081, "mean_token_accuracy": 0.8862844109535217, "num_tokens": 885711910.0, "step": 23330 }, { "epoch": 2.9679430097951913, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.6233601570129395, "learning_rate": 1e-05, "loss": 0.3923, "mean_token_accuracy": 0.8883607983589172, "num_tokens": 885751615.0, "step": 23331 }, { "epoch": 2.968070220073782, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 5.0067901611328125e-05, "grad_norm": 2.931278944015503, "learning_rate": 1e-05, "loss": 0.4144, "mean_token_accuracy": 0.885323703289032, "num_tokens": 885788394.0, "step": 23332 }, { "epoch": 2.9681974303523724, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.6433072090148926, "learning_rate": 1e-05, "loss": 0.3995, "mean_token_accuracy": 0.891042947769165, "num_tokens": 885820462.0, "step": 23333 }, { "epoch": 2.968324640630963, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 2.9451165199279785, "learning_rate": 1e-05, "loss": 0.4099, "mean_token_accuracy": 0.8894073963165283, "num_tokens": 885853714.0, "step": 23334 }, { "epoch": 2.9684518509095534, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.669701337814331, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8796133995056152, "num_tokens": 885887508.0, "step": 23335 }, { "epoch": 2.968579061188144, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 5.0067901611328125e-05, "grad_norm": 2.9278578758239746, "learning_rate": 1e-05, "loss": 0.4379, "mean_token_accuracy": 0.8755713701248169, "num_tokens": 885923314.0, "step": 23336 }, { "epoch": 2.9687062714667345, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.673464298248291, "learning_rate": 1e-05, "loss": 0.3936, "mean_token_accuracy": 0.8902438282966614, "num_tokens": 885964466.0, "step": 23337 }, { "epoch": 2.968833481745325, "ewc_loss": 0.087890625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 2.907133102416992, "learning_rate": 1e-05, "loss": 0.4321, "mean_token_accuracy": 0.877787709236145, "num_tokens": 886001483.0, "step": 23338 }, { "epoch": 2.9689606920239155, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.6823432445526123, "learning_rate": 1e-05, "loss": 0.4235, "mean_token_accuracy": 0.8818745017051697, "num_tokens": 886040130.0, "step": 23339 }, { "epoch": 2.969087902302506, "ewc_loss": 0.08642578125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.935264587402344e-05, "grad_norm": 2.88881778717041, "learning_rate": 1e-05, "loss": 0.4271, "mean_token_accuracy": 0.8805907964706421, "num_tokens": 886076291.0, "step": 23340 }, { "epoch": 2.9692151125810966, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.707059860229492, "learning_rate": 1e-05, "loss": 0.4723, "mean_token_accuracy": 0.8674212694168091, "num_tokens": 886109430.0, "step": 23341 }, { "epoch": 2.969342322859687, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.863739013671875e-05, "grad_norm": 2.8522250652313232, "learning_rate": 1e-05, "loss": 0.4184, "mean_token_accuracy": 0.8837971687316895, "num_tokens": 886147158.0, "step": 23342 }, { "epoch": 2.9694695331382777, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.6901016235351562, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8801802396774292, "num_tokens": 886184810.0, "step": 23343 }, { "epoch": 2.969596743416868, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.822906494140625, "learning_rate": 1e-05, "loss": 0.4126, "mean_token_accuracy": 0.8837613463401794, "num_tokens": 886225511.0, "step": 23344 }, { "epoch": 2.9697239536954587, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.6792190074920654, "learning_rate": 1e-05, "loss": 0.4067, "mean_token_accuracy": 0.8813491463661194, "num_tokens": 886265185.0, "step": 23345 }, { "epoch": 2.9698511639740492, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8110837936401367, "learning_rate": 1e-05, "loss": 0.4332, "mean_token_accuracy": 0.87903892993927, "num_tokens": 886304897.0, "step": 23346 }, { "epoch": 2.9699783742526398, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.689664840698242, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8704258799552917, "num_tokens": 886343097.0, "step": 23347 }, { "epoch": 2.97010558453123, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.8120288848876953, "learning_rate": 1e-05, "loss": 0.4325, "mean_token_accuracy": 0.8775546550750732, "num_tokens": 886377428.0, "step": 23348 }, { "epoch": 2.970232794809821, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.679825782775879, "learning_rate": 1e-05, "loss": 0.4354, "mean_token_accuracy": 0.8738741278648376, "num_tokens": 886418344.0, "step": 23349 }, { "epoch": 2.970360005088411, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.808486223220825, "learning_rate": 1e-05, "loss": 0.4997, "mean_token_accuracy": 0.8613578081130981, "num_tokens": 886458983.0, "step": 23350 }, { "epoch": 2.970487215367002, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.660557985305786, "learning_rate": 1e-05, "loss": 0.3776, "mean_token_accuracy": 0.8931611776351929, "num_tokens": 886492174.0, "step": 23351 }, { "epoch": 2.970614425645592, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.798124074935913, "learning_rate": 1e-05, "loss": 0.462, "mean_token_accuracy": 0.8688957691192627, "num_tokens": 886537145.0, "step": 23352 }, { "epoch": 2.970741635924183, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.689340353012085, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8694846630096436, "num_tokens": 886573420.0, "step": 23353 }, { "epoch": 2.970868846202773, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.783989667892456, "learning_rate": 1e-05, "loss": 0.4192, "mean_token_accuracy": 0.8805615305900574, "num_tokens": 886610489.0, "step": 23354 }, { "epoch": 2.9709960564813636, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6755430698394775, "learning_rate": 1e-05, "loss": 0.3876, "mean_token_accuracy": 0.8883063197135925, "num_tokens": 886646954.0, "step": 23355 }, { "epoch": 2.971123266759954, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7687153816223145, "learning_rate": 1e-05, "loss": 0.3913, "mean_token_accuracy": 0.8900070786476135, "num_tokens": 886686979.0, "step": 23356 }, { "epoch": 2.9712504770385446, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.713106155395508, "learning_rate": 1e-05, "loss": 0.5383, "mean_token_accuracy": 0.8514707684516907, "num_tokens": 886719734.0, "step": 23357 }, { "epoch": 2.971377687317135, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.760287046432495, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8836385011672974, "num_tokens": 886754762.0, "step": 23358 }, { "epoch": 2.9715048975957257, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.687142848968506, "learning_rate": 1e-05, "loss": 0.3965, "mean_token_accuracy": 0.885603129863739, "num_tokens": 886794830.0, "step": 23359 }, { "epoch": 2.971632107874316, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.750541925430298, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8737226724624634, "num_tokens": 886834952.0, "step": 23360 }, { "epoch": 2.9717593181529067, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.687363624572754, "learning_rate": 1e-05, "loss": 0.459, "mean_token_accuracy": 0.866664469242096, "num_tokens": 886874236.0, "step": 23361 }, { "epoch": 2.9718865284314973, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7570700645446777, "learning_rate": 1e-05, "loss": 0.4482, "mean_token_accuracy": 0.8710925579071045, "num_tokens": 886907723.0, "step": 23362 }, { "epoch": 2.972013738710088, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6959540843963623, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8767908215522766, "num_tokens": 886939914.0, "step": 23363 }, { "epoch": 2.9721409489886783, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.772178888320923, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8693608045578003, "num_tokens": 886970533.0, "step": 23364 }, { "epoch": 2.972268159267269, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.680677890777588, "learning_rate": 1e-05, "loss": 0.4643, "mean_token_accuracy": 0.8646857142448425, "num_tokens": 887012103.0, "step": 23365 }, { "epoch": 2.9723953695458594, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.743398666381836, "learning_rate": 1e-05, "loss": 0.4602, "mean_token_accuracy": 0.8679841756820679, "num_tokens": 887055789.0, "step": 23366 }, { "epoch": 2.97252257982445, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6693239212036133, "learning_rate": 1e-05, "loss": 0.3897, "mean_token_accuracy": 0.8879894018173218, "num_tokens": 887094505.0, "step": 23367 }, { "epoch": 2.9726497901030404, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7553441524505615, "learning_rate": 1e-05, "loss": 0.3981, "mean_token_accuracy": 0.8900487422943115, "num_tokens": 887131009.0, "step": 23368 }, { "epoch": 2.972777000381631, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.655421018600464, "learning_rate": 1e-05, "loss": 0.3748, "mean_token_accuracy": 0.8927476406097412, "num_tokens": 887170969.0, "step": 23369 }, { "epoch": 2.9729042106602215, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.763005256652832, "learning_rate": 1e-05, "loss": 0.3941, "mean_token_accuracy": 0.8879103660583496, "num_tokens": 887205123.0, "step": 23370 }, { "epoch": 2.973031420938812, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.678173542022705, "learning_rate": 1e-05, "loss": 0.3875, "mean_token_accuracy": 0.8906548619270325, "num_tokens": 887235908.0, "step": 23371 }, { "epoch": 2.9731586312174025, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7555840015411377, "learning_rate": 1e-05, "loss": 0.397, "mean_token_accuracy": 0.887083888053894, "num_tokens": 887273929.0, "step": 23372 }, { "epoch": 2.9732858414959926, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.670891284942627, "learning_rate": 1e-05, "loss": 0.4075, "mean_token_accuracy": 0.8844586610794067, "num_tokens": 887309379.0, "step": 23373 }, { "epoch": 2.9734130517745836, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.774711847305298, "learning_rate": 1e-05, "loss": 0.4671, "mean_token_accuracy": 0.8720196485519409, "num_tokens": 887346108.0, "step": 23374 }, { "epoch": 2.9735402620531737, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.669616937637329, "learning_rate": 1e-05, "loss": 0.3934, "mean_token_accuracy": 0.8930094242095947, "num_tokens": 887380066.0, "step": 23375 }, { "epoch": 2.9736674723317646, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.77944278717041, "learning_rate": 1e-05, "loss": 0.4329, "mean_token_accuracy": 0.877780556678772, "num_tokens": 887414043.0, "step": 23376 }, { "epoch": 2.9737946826103547, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6744203567504883, "learning_rate": 1e-05, "loss": 0.4336, "mean_token_accuracy": 0.8767549991607666, "num_tokens": 887451921.0, "step": 23377 }, { "epoch": 2.9739218928889453, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8022632598876953, "learning_rate": 1e-05, "loss": 0.434, "mean_token_accuracy": 0.8811929821968079, "num_tokens": 887483977.0, "step": 23378 }, { "epoch": 2.974049103167536, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.654022455215454, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8829086422920227, "num_tokens": 887516795.0, "step": 23379 }, { "epoch": 2.9741763134461263, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.790581226348877, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8745080232620239, "num_tokens": 887554791.0, "step": 23380 }, { "epoch": 2.974303523724717, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.64046573638916, "learning_rate": 1e-05, "loss": 0.4138, "mean_token_accuracy": 0.8839156031608582, "num_tokens": 887598451.0, "step": 23381 }, { "epoch": 2.9744307340033074, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8093979358673096, "learning_rate": 1e-05, "loss": 0.3851, "mean_token_accuracy": 0.8942328691482544, "num_tokens": 887633242.0, "step": 23382 }, { "epoch": 2.974557944281898, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6570887565612793, "learning_rate": 1e-05, "loss": 0.3952, "mean_token_accuracy": 0.886406660079956, "num_tokens": 887665084.0, "step": 23383 }, { "epoch": 2.9746851545604884, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8150243759155273, "learning_rate": 1e-05, "loss": 0.5095, "mean_token_accuracy": 0.854501485824585, "num_tokens": 887706086.0, "step": 23384 }, { "epoch": 2.974812364839079, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6539113521575928, "learning_rate": 1e-05, "loss": 0.4479, "mean_token_accuracy": 0.8717048764228821, "num_tokens": 887747395.0, "step": 23385 }, { "epoch": 2.9749395751176695, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8024580478668213, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8719468116760254, "num_tokens": 887789834.0, "step": 23386 }, { "epoch": 2.97506678539626, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6492254734039307, "learning_rate": 1e-05, "loss": 0.4755, "mean_token_accuracy": 0.8611029982566833, "num_tokens": 887831416.0, "step": 23387 }, { "epoch": 2.9751939956748505, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.808926820755005, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8700053691864014, "num_tokens": 887869781.0, "step": 23388 }, { "epoch": 2.975321205953441, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6479978561401367, "learning_rate": 1e-05, "loss": 0.4237, "mean_token_accuracy": 0.8810186386108398, "num_tokens": 887906844.0, "step": 23389 }, { "epoch": 2.9754484162320316, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.76837158203125e-05, "grad_norm": 2.831005573272705, "learning_rate": 1e-05, "loss": 0.4559, "mean_token_accuracy": 0.8719234466552734, "num_tokens": 887939514.0, "step": 23390 }, { "epoch": 2.975575626510622, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6585206985473633, "learning_rate": 1e-05, "loss": 0.4669, "mean_token_accuracy": 0.8651978969573975, "num_tokens": 887976041.0, "step": 23391 }, { "epoch": 2.9757028367892127, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.828443765640259, "learning_rate": 1e-05, "loss": 0.4715, "mean_token_accuracy": 0.8671820163726807, "num_tokens": 888010007.0, "step": 23392 }, { "epoch": 2.975830047067803, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6504502296447754, "learning_rate": 1e-05, "loss": 0.407, "mean_token_accuracy": 0.8823287487030029, "num_tokens": 888045654.0, "step": 23393 }, { "epoch": 2.9759572573463937, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8189892768859863, "learning_rate": 1e-05, "loss": 0.417, "mean_token_accuracy": 0.8834320902824402, "num_tokens": 888084075.0, "step": 23394 }, { "epoch": 2.9760844676249842, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.66084361076355, "learning_rate": 1e-05, "loss": 0.4361, "mean_token_accuracy": 0.876785397529602, "num_tokens": 888124891.0, "step": 23395 }, { "epoch": 2.9762116779035748, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8106658458709717, "learning_rate": 1e-05, "loss": 0.4109, "mean_token_accuracy": 0.8862994909286499, "num_tokens": 888161078.0, "step": 23396 }, { "epoch": 2.9763388881821653, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6630542278289795, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8840457797050476, "num_tokens": 888197013.0, "step": 23397 }, { "epoch": 2.9764660984607554, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8096418380737305, "learning_rate": 1e-05, "loss": 0.4386, "mean_token_accuracy": 0.8735612630844116, "num_tokens": 888231676.0, "step": 23398 }, { "epoch": 2.9765933087393464, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.655836820602417, "learning_rate": 1e-05, "loss": 0.4044, "mean_token_accuracy": 0.8849688172340393, "num_tokens": 888272676.0, "step": 23399 }, { "epoch": 2.9767205190179364, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.7980663776397705, "learning_rate": 1e-05, "loss": 0.4072, "mean_token_accuracy": 0.8835358619689941, "num_tokens": 888310064.0, "step": 23400 }, { "epoch": 2.9768477292965274, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.6731936931610107, "learning_rate": 1e-05, "loss": 0.4207, "mean_token_accuracy": 0.8807333707809448, "num_tokens": 888344103.0, "step": 23401 }, { "epoch": 2.9769749395751175, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.800628900527954, "learning_rate": 1e-05, "loss": 0.4496, "mean_token_accuracy": 0.8728759288787842, "num_tokens": 888379284.0, "step": 23402 }, { "epoch": 2.977102149853708, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6637985706329346, "learning_rate": 1e-05, "loss": 0.3999, "mean_token_accuracy": 0.8897582292556763, "num_tokens": 888414058.0, "step": 23403 }, { "epoch": 2.9772293601322986, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.8032662868499756, "learning_rate": 1e-05, "loss": 0.4497, "mean_token_accuracy": 0.8700658082962036, "num_tokens": 888454756.0, "step": 23404 }, { "epoch": 2.977356570410889, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6561505794525146, "learning_rate": 1e-05, "loss": 0.4221, "mean_token_accuracy": 0.8810491561889648, "num_tokens": 888491995.0, "step": 23405 }, { "epoch": 2.9774837806894796, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.808566093444824, "learning_rate": 1e-05, "loss": 0.4915, "mean_token_accuracy": 0.8613980412483215, "num_tokens": 888533073.0, "step": 23406 }, { "epoch": 2.97761099096807, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6472322940826416, "learning_rate": 1e-05, "loss": 0.3601, "mean_token_accuracy": 0.8992871642112732, "num_tokens": 888570392.0, "step": 23407 }, { "epoch": 2.9777382012466607, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.8038575649261475, "learning_rate": 1e-05, "loss": 0.4255, "mean_token_accuracy": 0.8806586265563965, "num_tokens": 888612078.0, "step": 23408 }, { "epoch": 2.977865411525251, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6509554386138916, "learning_rate": 1e-05, "loss": 0.4083, "mean_token_accuracy": 0.8863890171051025, "num_tokens": 888653603.0, "step": 23409 }, { "epoch": 2.9779926218038417, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.825021743774414, "learning_rate": 1e-05, "loss": 0.4567, "mean_token_accuracy": 0.8698747158050537, "num_tokens": 888687089.0, "step": 23410 }, { "epoch": 2.9781198320824323, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6454386711120605, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8656221032142639, "num_tokens": 888732221.0, "step": 23411 }, { "epoch": 2.978247042361023, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8177671432495117, "learning_rate": 1e-05, "loss": 0.3835, "mean_token_accuracy": 0.8930171132087708, "num_tokens": 888768992.0, "step": 23412 }, { "epoch": 2.9783742526396133, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6421713829040527, "learning_rate": 1e-05, "loss": 0.3957, "mean_token_accuracy": 0.8895950317382812, "num_tokens": 888809834.0, "step": 23413 }, { "epoch": 2.978501462918204, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8354270458221436, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8835129737854004, "num_tokens": 888851943.0, "step": 23414 }, { "epoch": 2.9786286731967944, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6404099464416504, "learning_rate": 1e-05, "loss": 0.4119, "mean_token_accuracy": 0.8817176222801208, "num_tokens": 888890465.0, "step": 23415 }, { "epoch": 2.978755883475385, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8316333293914795, "learning_rate": 1e-05, "loss": 0.4405, "mean_token_accuracy": 0.8778737783432007, "num_tokens": 888932954.0, "step": 23416 }, { "epoch": 2.9788830937539754, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.637885332107544, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.881060004234314, "num_tokens": 888973916.0, "step": 23417 }, { "epoch": 2.979010304032566, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.842003583908081, "learning_rate": 1e-05, "loss": 0.4847, "mean_token_accuracy": 0.8599532842636108, "num_tokens": 889017202.0, "step": 23418 }, { "epoch": 2.9791375143111565, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6631009578704834, "learning_rate": 1e-05, "loss": 0.4735, "mean_token_accuracy": 0.8657041788101196, "num_tokens": 889053604.0, "step": 23419 }, { "epoch": 2.979264724589747, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.840000629425049, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8784530758857727, "num_tokens": 889087328.0, "step": 23420 }, { "epoch": 2.979391934868337, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.653716802597046, "learning_rate": 1e-05, "loss": 0.4456, "mean_token_accuracy": 0.8737115859985352, "num_tokens": 889125209.0, "step": 23421 }, { "epoch": 2.979519145146928, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.834988594055176, "learning_rate": 1e-05, "loss": 0.4483, "mean_token_accuracy": 0.8737612962722778, "num_tokens": 889160968.0, "step": 23422 }, { "epoch": 2.979646355425518, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.656731128692627, "learning_rate": 1e-05, "loss": 0.4562, "mean_token_accuracy": 0.8731460571289062, "num_tokens": 889201710.0, "step": 23423 }, { "epoch": 2.979773565704109, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8504295349121094, "learning_rate": 1e-05, "loss": 0.488, "mean_token_accuracy": 0.8635913133621216, "num_tokens": 889237003.0, "step": 23424 }, { "epoch": 2.979900775982699, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.656423330307007, "learning_rate": 1e-05, "loss": 0.4573, "mean_token_accuracy": 0.8696698546409607, "num_tokens": 889278775.0, "step": 23425 }, { "epoch": 2.98002798626129, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.8380680084228516, "learning_rate": 1e-05, "loss": 0.4629, "mean_token_accuracy": 0.871353030204773, "num_tokens": 889320888.0, "step": 23426 }, { "epoch": 2.9801551965398803, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.655019521713257, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8811720609664917, "num_tokens": 889356746.0, "step": 23427 }, { "epoch": 2.980282406818471, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.827501058578491, "learning_rate": 1e-05, "loss": 0.4079, "mean_token_accuracy": 0.8872685432434082, "num_tokens": 889400771.0, "step": 23428 }, { "epoch": 2.9804096170970613, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.658202648162842, "learning_rate": 1e-05, "loss": 0.4468, "mean_token_accuracy": 0.8721221685409546, "num_tokens": 889440459.0, "step": 23429 }, { "epoch": 2.980536827375652, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8238935470581055, "learning_rate": 1e-05, "loss": 0.4178, "mean_token_accuracy": 0.8841500282287598, "num_tokens": 889479782.0, "step": 23430 }, { "epoch": 2.9806640376542424, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6469998359680176, "learning_rate": 1e-05, "loss": 0.3698, "mean_token_accuracy": 0.8956089019775391, "num_tokens": 889520167.0, "step": 23431 }, { "epoch": 2.980791247932833, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.820007085800171, "learning_rate": 1e-05, "loss": 0.4276, "mean_token_accuracy": 0.8817597031593323, "num_tokens": 889567589.0, "step": 23432 }, { "epoch": 2.9809184582114234, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6642463207244873, "learning_rate": 1e-05, "loss": 0.4345, "mean_token_accuracy": 0.8782471418380737, "num_tokens": 889607703.0, "step": 23433 }, { "epoch": 2.981045668490014, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.822610378265381, "learning_rate": 1e-05, "loss": 0.4373, "mean_token_accuracy": 0.8734508752822876, "num_tokens": 889645735.0, "step": 23434 }, { "epoch": 2.9811728787686045, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.672475814819336, "learning_rate": 1e-05, "loss": 0.4045, "mean_token_accuracy": 0.8845939040184021, "num_tokens": 889680035.0, "step": 23435 }, { "epoch": 2.981300089047195, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.7978971004486084, "learning_rate": 1e-05, "loss": 0.4296, "mean_token_accuracy": 0.8781365156173706, "num_tokens": 889720234.0, "step": 23436 }, { "epoch": 2.9814272993257855, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.665081739425659, "learning_rate": 1e-05, "loss": 0.3944, "mean_token_accuracy": 0.8879655599594116, "num_tokens": 889756070.0, "step": 23437 }, { "epoch": 2.981554509604376, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.790480136871338, "learning_rate": 1e-05, "loss": 0.424, "mean_token_accuracy": 0.8805800080299377, "num_tokens": 889796499.0, "step": 23438 }, { "epoch": 2.9816817198829666, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.721973180770874, "learning_rate": 1e-05, "loss": 0.5, "mean_token_accuracy": 0.8603729009628296, "num_tokens": 889830804.0, "step": 23439 }, { "epoch": 2.981808930161557, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.77651047706604, "learning_rate": 1e-05, "loss": 0.4604, "mean_token_accuracy": 0.8678834438323975, "num_tokens": 889865228.0, "step": 23440 }, { "epoch": 2.9819361404401477, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6950523853302, "learning_rate": 1e-05, "loss": 0.4284, "mean_token_accuracy": 0.8767198324203491, "num_tokens": 889900780.0, "step": 23441 }, { "epoch": 2.982063350718738, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.745326519012451, "learning_rate": 1e-05, "loss": 0.4349, "mean_token_accuracy": 0.8804506063461304, "num_tokens": 889944366.0, "step": 23442 }, { "epoch": 2.9821905609973287, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.683565855026245, "learning_rate": 1e-05, "loss": 0.3684, "mean_token_accuracy": 0.8972760438919067, "num_tokens": 889979739.0, "step": 23443 }, { "epoch": 2.9823177712759192, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7462894916534424, "learning_rate": 1e-05, "loss": 0.4112, "mean_token_accuracy": 0.8804895281791687, "num_tokens": 890017288.0, "step": 23444 }, { "epoch": 2.9824449815545098, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.696467638015747, "learning_rate": 1e-05, "loss": 0.3977, "mean_token_accuracy": 0.8840094804763794, "num_tokens": 890050848.0, "step": 23445 }, { "epoch": 2.9825721918331, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003719329833984375, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7407736778259277, "learning_rate": 1e-05, "loss": 0.4164, "mean_token_accuracy": 0.8805119395256042, "num_tokens": 890087985.0, "step": 23446 }, { "epoch": 2.982699402111691, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.6958742141723633, "learning_rate": 1e-05, "loss": 0.4523, "mean_token_accuracy": 0.8716189861297607, "num_tokens": 890128292.0, "step": 23447 }, { "epoch": 2.982826612390281, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.719587564468384, "learning_rate": 1e-05, "loss": 0.4195, "mean_token_accuracy": 0.8811931610107422, "num_tokens": 890176130.0, "step": 23448 }, { "epoch": 2.982953822668872, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7026400566101074, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8784412145614624, "num_tokens": 890212547.0, "step": 23449 }, { "epoch": 2.983081032947462, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7236735820770264, "learning_rate": 1e-05, "loss": 0.45, "mean_token_accuracy": 0.8741618394851685, "num_tokens": 890257423.0, "step": 23450 }, { "epoch": 2.983208243226053, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.6962406635284424, "learning_rate": 1e-05, "loss": 0.4303, "mean_token_accuracy": 0.8781243562698364, "num_tokens": 890299465.0, "step": 23451 }, { "epoch": 2.983335453504643, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7230594158172607, "learning_rate": 1e-05, "loss": 0.4022, "mean_token_accuracy": 0.8853447437286377, "num_tokens": 890340010.0, "step": 23452 }, { "epoch": 2.9834626637832335, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.691439390182495, "learning_rate": 1e-05, "loss": 0.4333, "mean_token_accuracy": 0.8776841163635254, "num_tokens": 890380865.0, "step": 23453 }, { "epoch": 2.983589874061824, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7272253036499023, "learning_rate": 1e-05, "loss": 0.3929, "mean_token_accuracy": 0.8913389444351196, "num_tokens": 890416710.0, "step": 23454 }, { "epoch": 2.9837170843404146, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.691030740737915, "learning_rate": 1e-05, "loss": 0.3951, "mean_token_accuracy": 0.8905442953109741, "num_tokens": 890452963.0, "step": 23455 }, { "epoch": 2.983844294619005, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7291367053985596, "learning_rate": 1e-05, "loss": 0.4102, "mean_token_accuracy": 0.8829239010810852, "num_tokens": 890493308.0, "step": 23456 }, { "epoch": 2.9839715048975957, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6966331005096436, "learning_rate": 1e-05, "loss": 0.4447, "mean_token_accuracy": 0.8737630248069763, "num_tokens": 890529279.0, "step": 23457 }, { "epoch": 2.984098715176186, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.737645387649536, "learning_rate": 1e-05, "loss": 0.48, "mean_token_accuracy": 0.8602511882781982, "num_tokens": 890568866.0, "step": 23458 }, { "epoch": 2.9842259254547767, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6802010536193848, "learning_rate": 1e-05, "loss": 0.4163, "mean_token_accuracy": 0.8826935291290283, "num_tokens": 890613077.0, "step": 23459 }, { "epoch": 2.9843531357333672, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.733234167098999, "learning_rate": 1e-05, "loss": 0.423, "mean_token_accuracy": 0.880556583404541, "num_tokens": 890653599.0, "step": 23460 }, { "epoch": 2.9844803460119578, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6848394870758057, "learning_rate": 1e-05, "loss": 0.4224, "mean_token_accuracy": 0.8804892301559448, "num_tokens": 890694581.0, "step": 23461 }, { "epoch": 2.9846075562905483, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7526721954345703, "learning_rate": 1e-05, "loss": 0.3974, "mean_token_accuracy": 0.8854614496231079, "num_tokens": 890725912.0, "step": 23462 }, { "epoch": 2.984734766569139, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.684471845626831, "learning_rate": 1e-05, "loss": 0.4338, "mean_token_accuracy": 0.8766734600067139, "num_tokens": 890763028.0, "step": 23463 }, { "epoch": 2.9848619768477294, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.7398107051849365, "learning_rate": 1e-05, "loss": 0.4286, "mean_token_accuracy": 0.8780001997947693, "num_tokens": 890803318.0, "step": 23464 }, { "epoch": 2.98498918712632, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6781506538391113, "learning_rate": 1e-05, "loss": 0.4393, "mean_token_accuracy": 0.8759332299232483, "num_tokens": 890841405.0, "step": 23465 }, { "epoch": 2.9851163974049104, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.746255397796631, "learning_rate": 1e-05, "loss": 0.4078, "mean_token_accuracy": 0.8830143809318542, "num_tokens": 890884271.0, "step": 23466 }, { "epoch": 2.985243607683501, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6850922107696533, "learning_rate": 1e-05, "loss": 0.3991, "mean_token_accuracy": 0.8870607018470764, "num_tokens": 890917747.0, "step": 23467 }, { "epoch": 2.9853708179620915, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7372236251831055, "learning_rate": 1e-05, "loss": 0.4156, "mean_token_accuracy": 0.8838614225387573, "num_tokens": 890958659.0, "step": 23468 }, { "epoch": 2.985498028240682, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.687960624694824, "learning_rate": 1e-05, "loss": 0.3831, "mean_token_accuracy": 0.8948416709899902, "num_tokens": 890993792.0, "step": 23469 }, { "epoch": 2.9856252385192725, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7485132217407227, "learning_rate": 1e-05, "loss": 0.408, "mean_token_accuracy": 0.8837975263595581, "num_tokens": 891030873.0, "step": 23470 }, { "epoch": 2.9857524487978626, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.671875, "learning_rate": 1e-05, "loss": 0.383, "mean_token_accuracy": 0.8942690491676331, "num_tokens": 891068062.0, "step": 23471 }, { "epoch": 2.9858796590764536, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7560617923736572, "learning_rate": 1e-05, "loss": 0.4591, "mean_token_accuracy": 0.8695783615112305, "num_tokens": 891108749.0, "step": 23472 }, { "epoch": 2.9860068693550437, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6743879318237305, "learning_rate": 1e-05, "loss": 0.4288, "mean_token_accuracy": 0.8770484924316406, "num_tokens": 891144749.0, "step": 23473 }, { "epoch": 2.9861340796336346, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.769660472869873, "learning_rate": 1e-05, "loss": 0.4747, "mean_token_accuracy": 0.8670534491539001, "num_tokens": 891180795.0, "step": 23474 }, { "epoch": 2.9862612899122247, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6654837131500244, "learning_rate": 1e-05, "loss": 0.4395, "mean_token_accuracy": 0.8724944591522217, "num_tokens": 891221333.0, "step": 23475 }, { "epoch": 2.9863885001908153, "ewc_loss": 0.08251953125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.769129753112793, "learning_rate": 1e-05, "loss": 0.4257, "mean_token_accuracy": 0.8808668851852417, "num_tokens": 891258414.0, "step": 23476 }, { "epoch": 2.986515710469406, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.661261558532715, "learning_rate": 1e-05, "loss": 0.453, "mean_token_accuracy": 0.8717446327209473, "num_tokens": 891296864.0, "step": 23477 }, { "epoch": 2.9866429207479963, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7606313228607178, "learning_rate": 1e-05, "loss": 0.3776, "mean_token_accuracy": 0.8953789472579956, "num_tokens": 891338191.0, "step": 23478 }, { "epoch": 2.986770131026587, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.66636323928833, "learning_rate": 1e-05, "loss": 0.4375, "mean_token_accuracy": 0.8751367330551147, "num_tokens": 891377354.0, "step": 23479 }, { "epoch": 2.9868973413051774, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7642171382904053, "learning_rate": 1e-05, "loss": 0.4149, "mean_token_accuracy": 0.8815843462944031, "num_tokens": 891415868.0, "step": 23480 }, { "epoch": 2.987024551583768, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.668285369873047, "learning_rate": 1e-05, "loss": 0.405, "mean_token_accuracy": 0.8824979066848755, "num_tokens": 891449548.0, "step": 23481 }, { "epoch": 2.9871517618623584, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7756619453430176, "learning_rate": 1e-05, "loss": 0.4822, "mean_token_accuracy": 0.8598084449768066, "num_tokens": 891485937.0, "step": 23482 }, { "epoch": 2.987278972140949, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.6664180755615234, "learning_rate": 1e-05, "loss": 0.4056, "mean_token_accuracy": 0.8845909237861633, "num_tokens": 891522458.0, "step": 23483 }, { "epoch": 2.9874061824195395, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7762181758880615, "learning_rate": 1e-05, "loss": 0.4633, "mean_token_accuracy": 0.8687415719032288, "num_tokens": 891560804.0, "step": 23484 }, { "epoch": 2.98753339269813, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.663740634918213, "learning_rate": 1e-05, "loss": 0.3919, "mean_token_accuracy": 0.8904783725738525, "num_tokens": 891595692.0, "step": 23485 }, { "epoch": 2.9876606029767205, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.776529550552368, "learning_rate": 1e-05, "loss": 0.4037, "mean_token_accuracy": 0.8880518674850464, "num_tokens": 891632007.0, "step": 23486 }, { "epoch": 2.987787813255311, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.656782627105713, "learning_rate": 1e-05, "loss": 0.4059, "mean_token_accuracy": 0.8856253623962402, "num_tokens": 891671605.0, "step": 23487 }, { "epoch": 2.9879150235339016, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.792391061782837, "learning_rate": 1e-05, "loss": 0.4611, "mean_token_accuracy": 0.868245005607605, "num_tokens": 891708369.0, "step": 23488 }, { "epoch": 2.988042233812492, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.660320520401001, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.886258065700531, "num_tokens": 891744673.0, "step": 23489 }, { "epoch": 2.9881694440910826, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.00037384033203125, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7936131954193115, "learning_rate": 1e-05, "loss": 0.4342, "mean_token_accuracy": 0.8774408102035522, "num_tokens": 891781971.0, "step": 23490 }, { "epoch": 2.988296654369673, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.00037384033203125, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.678806781768799, "learning_rate": 1e-05, "loss": 0.3984, "mean_token_accuracy": 0.8889045715332031, "num_tokens": 891814079.0, "step": 23491 }, { "epoch": 2.9884238646482637, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.787564516067505, "learning_rate": 1e-05, "loss": 0.4114, "mean_token_accuracy": 0.8861355781555176, "num_tokens": 891852188.0, "step": 23492 }, { "epoch": 2.9885510749268542, "ewc_loss": 0.080078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.657912015914917, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8770773410797119, "num_tokens": 891894048.0, "step": 23493 }, { "epoch": 2.9886782852054448, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7872092723846436, "learning_rate": 1e-05, "loss": 0.4304, "mean_token_accuracy": 0.8791987299919128, "num_tokens": 891935199.0, "step": 23494 }, { "epoch": 2.9888054954840353, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6581692695617676, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8808033466339111, "num_tokens": 891977686.0, "step": 23495 }, { "epoch": 2.9889327057626254, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.789961576461792, "learning_rate": 1e-05, "loss": 0.4539, "mean_token_accuracy": 0.8703975081443787, "num_tokens": 892017501.0, "step": 23496 }, { "epoch": 2.9890599160412163, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.660228729248047, "learning_rate": 1e-05, "loss": 0.4516, "mean_token_accuracy": 0.8686432838439941, "num_tokens": 892057369.0, "step": 23497 }, { "epoch": 2.9891871263198064, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7973740100860596, "learning_rate": 1e-05, "loss": 0.4268, "mean_token_accuracy": 0.8782127499580383, "num_tokens": 892093998.0, "step": 23498 }, { "epoch": 2.9893143365983974, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6647603511810303, "learning_rate": 1e-05, "loss": 0.3931, "mean_token_accuracy": 0.8923251032829285, "num_tokens": 892125835.0, "step": 23499 }, { "epoch": 2.9894415468769875, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.8169524669647217, "learning_rate": 1e-05, "loss": 0.4351, "mean_token_accuracy": 0.8802492022514343, "num_tokens": 892160340.0, "step": 23500 }, { "epoch": 2.989568757155578, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6547412872314453, "learning_rate": 1e-05, "loss": 0.4282, "mean_token_accuracy": 0.8762969374656677, "num_tokens": 892202096.0, "step": 23501 }, { "epoch": 2.9896959674341685, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8010454177856445, "learning_rate": 1e-05, "loss": 0.4035, "mean_token_accuracy": 0.8883275389671326, "num_tokens": 892241942.0, "step": 23502 }, { "epoch": 2.989823177712759, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.660456657409668, "learning_rate": 1e-05, "loss": 0.4556, "mean_token_accuracy": 0.8700055480003357, "num_tokens": 892284724.0, "step": 23503 }, { "epoch": 2.9899503879913496, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 3.4621429443359375, "learning_rate": 1e-05, "loss": 0.3955, "mean_token_accuracy": 0.8919483423233032, "num_tokens": 892324761.0, "step": 23504 }, { "epoch": 2.99007759826994, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.6837210655212402, "learning_rate": 1e-05, "loss": 0.3889, "mean_token_accuracy": 0.8903626203536987, "num_tokens": 892358128.0, "step": 23505 }, { "epoch": 2.9902048085485307, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 5.1021575927734375e-05, "grad_norm": 3.024240732192993, "learning_rate": 1e-05, "loss": 0.4534, "mean_token_accuracy": 0.8752188086509705, "num_tokens": 892400380.0, "step": 23506 }, { "epoch": 2.990332018827121, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.458427429199219e-05, "grad_norm": 2.6477699279785156, "learning_rate": 1e-05, "loss": 0.4553, "mean_token_accuracy": 0.8724229335784912, "num_tokens": 892439141.0, "step": 23507 }, { "epoch": 2.9904592291057117, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 5.459785461425781e-05, "grad_norm": 3.0659878253936768, "learning_rate": 1e-05, "loss": 0.4362, "mean_token_accuracy": 0.8773502111434937, "num_tokens": 892480983.0, "step": 23508 }, { "epoch": 2.9905864393843022, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.692415237426758, "learning_rate": 1e-05, "loss": 0.439, "mean_token_accuracy": 0.8796300888061523, "num_tokens": 892518220.0, "step": 23509 }, { "epoch": 2.9907136496628928, "ewc_loss": 0.091796875, "ewc_loss_diag": 0.0003757476806640625, "ewc_loss_parallel": 5.412101745605469e-05, "grad_norm": 3.0449774265289307, "learning_rate": 1e-05, "loss": 0.4273, "mean_token_accuracy": 0.8808438777923584, "num_tokens": 892549308.0, "step": 23510 }, { "epoch": 2.9908408599414833, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.000377655029296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.7150003910064697, "learning_rate": 1e-05, "loss": 0.4388, "mean_token_accuracy": 0.8787402510643005, "num_tokens": 892590154.0, "step": 23511 }, { "epoch": 2.990968070220074, "ewc_loss": 0.09228515625, "ewc_loss_diag": 0.000377655029296875, "ewc_loss_parallel": 5.435943603515625e-05, "grad_norm": 3.002225160598755, "learning_rate": 1e-05, "loss": 0.3823, "mean_token_accuracy": 0.896509051322937, "num_tokens": 892630951.0, "step": 23512 }, { "epoch": 2.9910952804986644, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.000377655029296875, "ewc_loss_parallel": 4.744529724121094e-05, "grad_norm": 2.729358434677124, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8744806051254272, "num_tokens": 892675345.0, "step": 23513 }, { "epoch": 2.991222490777255, "ewc_loss": 0.09033203125, "ewc_loss_diag": 0.000377655029296875, "ewc_loss_parallel": 5.245208740234375e-05, "grad_norm": 2.9657599925994873, "learning_rate": 1e-05, "loss": 0.4578, "mean_token_accuracy": 0.8713920712471008, "num_tokens": 892718129.0, "step": 23514 }, { "epoch": 2.9913497010558454, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.000377655029296875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.7326574325561523, "learning_rate": 1e-05, "loss": 0.4348, "mean_token_accuracy": 0.8802238702774048, "num_tokens": 892757159.0, "step": 23515 }, { "epoch": 2.991476911334436, "ewc_loss": 0.0888671875, "ewc_loss_diag": 0.000377655029296875, "ewc_loss_parallel": 5.125999450683594e-05, "grad_norm": 2.930859088897705, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.881580114364624, "num_tokens": 892797872.0, "step": 23516 }, { "epoch": 2.9916041216130265, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7217836380004883, "learning_rate": 1e-05, "loss": 0.4505, "mean_token_accuracy": 0.8709970712661743, "num_tokens": 892837213.0, "step": 23517 }, { "epoch": 2.991731331891617, "ewc_loss": 0.08837890625, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 5.054473876953125e-05, "grad_norm": 2.892749547958374, "learning_rate": 1e-05, "loss": 0.4412, "mean_token_accuracy": 0.8742258548736572, "num_tokens": 892878651.0, "step": 23518 }, { "epoch": 2.991858542170207, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7159931659698486, "learning_rate": 1e-05, "loss": 0.3882, "mean_token_accuracy": 0.8883779048919678, "num_tokens": 892913739.0, "step": 23519 }, { "epoch": 2.991985752448798, "ewc_loss": 0.0869140625, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.9114227294921875e-05, "grad_norm": 2.8761770725250244, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.8766897916793823, "num_tokens": 892949649.0, "step": 23520 }, { "epoch": 2.992112962727388, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.703629732131958, "learning_rate": 1e-05, "loss": 0.421, "mean_token_accuracy": 0.8830509781837463, "num_tokens": 892988831.0, "step": 23521 }, { "epoch": 2.992240173005979, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.8613781929016113, "learning_rate": 1e-05, "loss": 0.4256, "mean_token_accuracy": 0.8804495334625244, "num_tokens": 893025311.0, "step": 23522 }, { "epoch": 2.992367383284569, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.6802682876586914, "learning_rate": 1e-05, "loss": 0.446, "mean_token_accuracy": 0.8730188012123108, "num_tokens": 893072461.0, "step": 23523 }, { "epoch": 2.99249459356316, "ewc_loss": 0.0859375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.8160552978515625e-05, "grad_norm": 2.859400510787964, "learning_rate": 1e-05, "loss": 0.4527, "mean_token_accuracy": 0.8722110390663147, "num_tokens": 893109268.0, "step": 23524 }, { "epoch": 2.9926218038417502, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7104344367980957, "learning_rate": 1e-05, "loss": 0.457, "mean_token_accuracy": 0.8731670379638672, "num_tokens": 893139033.0, "step": 23525 }, { "epoch": 2.992749014120341, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.818084239959717, "learning_rate": 1e-05, "loss": 0.4046, "mean_token_accuracy": 0.8850101232528687, "num_tokens": 893180056.0, "step": 23526 }, { "epoch": 2.9928762243989313, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7039945125579834, "learning_rate": 1e-05, "loss": 0.4402, "mean_token_accuracy": 0.8759227395057678, "num_tokens": 893214164.0, "step": 23527 }, { "epoch": 2.993003434677522, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.601478576660156e-05, "grad_norm": 2.7869138717651367, "learning_rate": 1e-05, "loss": 0.4239, "mean_token_accuracy": 0.8808282613754272, "num_tokens": 893255030.0, "step": 23528 }, { "epoch": 2.9931306449561124, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.717022180557251, "learning_rate": 1e-05, "loss": 0.4366, "mean_token_accuracy": 0.8729062080383301, "num_tokens": 893287661.0, "step": 23529 }, { "epoch": 2.993257855234703, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7796313762664795, "learning_rate": 1e-05, "loss": 0.4947, "mean_token_accuracy": 0.8568379878997803, "num_tokens": 893328461.0, "step": 23530 }, { "epoch": 2.9933850655132934, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.696089029312134, "learning_rate": 1e-05, "loss": 0.41, "mean_token_accuracy": 0.8839995265007019, "num_tokens": 893367543.0, "step": 23531 }, { "epoch": 2.993512275791884, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.506111145019531e-05, "grad_norm": 2.7661776542663574, "learning_rate": 1e-05, "loss": 0.3966, "mean_token_accuracy": 0.8905767798423767, "num_tokens": 893401180.0, "step": 23532 }, { "epoch": 2.9936394860704745, "ewc_loss": 0.08154296875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.363059997558594e-05, "grad_norm": 2.707613468170166, "learning_rate": 1e-05, "loss": 0.3658, "mean_token_accuracy": 0.8974332809448242, "num_tokens": 893429823.0, "step": 23533 }, { "epoch": 2.993766696349065, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.746155023574829, "learning_rate": 1e-05, "loss": 0.4188, "mean_token_accuracy": 0.8780827522277832, "num_tokens": 893472730.0, "step": 23534 }, { "epoch": 2.9938939066276555, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.688431978225708, "learning_rate": 1e-05, "loss": 0.4111, "mean_token_accuracy": 0.8824954032897949, "num_tokens": 893516506.0, "step": 23535 }, { "epoch": 2.994021116906246, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.482269287109375e-05, "grad_norm": 2.74497652053833, "learning_rate": 1e-05, "loss": 0.4287, "mean_token_accuracy": 0.8787063360214233, "num_tokens": 893560539.0, "step": 23536 }, { "epoch": 2.9941483271848366, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.7031872272491455, "learning_rate": 1e-05, "loss": 0.4232, "mean_token_accuracy": 0.8786683678627014, "num_tokens": 893595986.0, "step": 23537 }, { "epoch": 2.994275537463427, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7464241981506348, "learning_rate": 1e-05, "loss": 0.4, "mean_token_accuracy": 0.8868401646614075, "num_tokens": 893632997.0, "step": 23538 }, { "epoch": 2.9944027477420176, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7208211421966553, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8765446543693542, "num_tokens": 893671889.0, "step": 23539 }, { "epoch": 2.994529958020608, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7453014850616455, "learning_rate": 1e-05, "loss": 0.3973, "mean_token_accuracy": 0.8882970213890076, "num_tokens": 893708781.0, "step": 23540 }, { "epoch": 2.9946571682991987, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7099435329437256, "learning_rate": 1e-05, "loss": 0.4266, "mean_token_accuracy": 0.8753856420516968, "num_tokens": 893745590.0, "step": 23541 }, { "epoch": 2.9947843785777892, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.72265887260437, "learning_rate": 1e-05, "loss": 0.4087, "mean_token_accuracy": 0.884510338306427, "num_tokens": 893783599.0, "step": 23542 }, { "epoch": 2.9949115888563798, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.7018556594848633, "learning_rate": 1e-05, "loss": 0.3685, "mean_token_accuracy": 0.8963913917541504, "num_tokens": 893822603.0, "step": 23543 }, { "epoch": 2.99503879913497, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.7431869506835938, "learning_rate": 1e-05, "loss": 0.4194, "mean_token_accuracy": 0.8835372924804688, "num_tokens": 893857318.0, "step": 23544 }, { "epoch": 2.995166009413561, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6923866271972656, "learning_rate": 1e-05, "loss": 0.4205, "mean_token_accuracy": 0.8758907318115234, "num_tokens": 893896269.0, "step": 23545 }, { "epoch": 2.995293219692151, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.410743713378906e-05, "grad_norm": 2.7446131706237793, "learning_rate": 1e-05, "loss": 0.4628, "mean_token_accuracy": 0.8654323816299438, "num_tokens": 893936877.0, "step": 23546 }, { "epoch": 2.995420429970742, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.702744722366333, "learning_rate": 1e-05, "loss": 0.4517, "mean_token_accuracy": 0.8773846626281738, "num_tokens": 893970014.0, "step": 23547 }, { "epoch": 2.995547640249332, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.7388219833374023, "learning_rate": 1e-05, "loss": 0.4248, "mean_token_accuracy": 0.8798424601554871, "num_tokens": 894005568.0, "step": 23548 }, { "epoch": 2.995674850527923, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.38690185546875e-05, "grad_norm": 2.67934250831604, "learning_rate": 1e-05, "loss": 0.4294, "mean_token_accuracy": 0.8790401220321655, "num_tokens": 894053880.0, "step": 23549 }, { "epoch": 2.995802060806513, "ewc_loss": 0.08203125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.4345855712890625e-05, "grad_norm": 2.792836904525757, "learning_rate": 1e-05, "loss": 0.3807, "mean_token_accuracy": 0.8923429250717163, "num_tokens": 894083981.0, "step": 23550 }, { "epoch": 2.9959292710851035, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.658296585083008, "learning_rate": 1e-05, "loss": 0.4216, "mean_token_accuracy": 0.8824421763420105, "num_tokens": 894125315.0, "step": 23551 }, { "epoch": 2.996056481363694, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.7818374633789062, "learning_rate": 1e-05, "loss": 0.4445, "mean_token_accuracy": 0.8743161559104919, "num_tokens": 894165195.0, "step": 23552 }, { "epoch": 2.9961836916422846, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.656998872756958, "learning_rate": 1e-05, "loss": 0.4443, "mean_token_accuracy": 0.8715696334838867, "num_tokens": 894206443.0, "step": 23553 }, { "epoch": 2.996310901920875, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.7932896614074707, "learning_rate": 1e-05, "loss": 0.4177, "mean_token_accuracy": 0.8834350109100342, "num_tokens": 894240897.0, "step": 23554 }, { "epoch": 2.9964381121994657, "ewc_loss": 0.08056640625, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.267692565917969e-05, "grad_norm": 2.655186891555786, "learning_rate": 1e-05, "loss": 0.3922, "mean_token_accuracy": 0.8898710608482361, "num_tokens": 894277475.0, "step": 23555 }, { "epoch": 2.996565322478056, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.7935400009155273, "learning_rate": 1e-05, "loss": 0.458, "mean_token_accuracy": 0.8691087961196899, "num_tokens": 894320231.0, "step": 23556 }, { "epoch": 2.9966925327566467, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6584267616271973, "learning_rate": 1e-05, "loss": 0.4225, "mean_token_accuracy": 0.8803435564041138, "num_tokens": 894359196.0, "step": 23557 }, { "epoch": 2.9968197430352372, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.795651912689209, "learning_rate": 1e-05, "loss": 0.4453, "mean_token_accuracy": 0.8733226656913757, "num_tokens": 894401362.0, "step": 23558 }, { "epoch": 2.9969469533138278, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6629889011383057, "learning_rate": 1e-05, "loss": 0.3963, "mean_token_accuracy": 0.8875898718833923, "num_tokens": 894435445.0, "step": 23559 }, { "epoch": 2.9970741635924183, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.795833110809326, "learning_rate": 1e-05, "loss": 0.4392, "mean_token_accuracy": 0.8745874762535095, "num_tokens": 894475968.0, "step": 23560 }, { "epoch": 2.997201373871009, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.665562868118286, "learning_rate": 1e-05, "loss": 0.4039, "mean_token_accuracy": 0.8848318457603455, "num_tokens": 894511692.0, "step": 23561 }, { "epoch": 2.9973285841495994, "ewc_loss": 0.0830078125, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.5299530029296875e-05, "grad_norm": 2.8051445484161377, "learning_rate": 1e-05, "loss": 0.4546, "mean_token_accuracy": 0.8751964569091797, "num_tokens": 894550886.0, "step": 23562 }, { "epoch": 2.99745579442819, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.651273488998413, "learning_rate": 1e-05, "loss": 0.3903, "mean_token_accuracy": 0.890019953250885, "num_tokens": 894584811.0, "step": 23563 }, { "epoch": 2.9975830047067804, "ewc_loss": 0.08349609375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.553794860839844e-05, "grad_norm": 2.8127241134643555, "learning_rate": 1e-05, "loss": 0.4426, "mean_token_accuracy": 0.8755444288253784, "num_tokens": 894622866.0, "step": 23564 }, { "epoch": 2.997710214985371, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6427388191223145, "learning_rate": 1e-05, "loss": 0.3897, "mean_token_accuracy": 0.8934348821640015, "num_tokens": 894663513.0, "step": 23565 }, { "epoch": 2.9978374252639615, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.819852828979492, "learning_rate": 1e-05, "loss": 0.4425, "mean_token_accuracy": 0.8744731545448303, "num_tokens": 894702873.0, "step": 23566 }, { "epoch": 2.997964635542552, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6553215980529785, "learning_rate": 1e-05, "loss": 0.4117, "mean_token_accuracy": 0.8822922110557556, "num_tokens": 894737990.0, "step": 23567 }, { "epoch": 2.9980918458211425, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.8167905807495117, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.8836696147918701, "num_tokens": 894775408.0, "step": 23568 }, { "epoch": 2.9982190560997326, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6641809940338135, "learning_rate": 1e-05, "loss": 0.437, "mean_token_accuracy": 0.873727023601532, "num_tokens": 894815702.0, "step": 23569 }, { "epoch": 2.9983462663783236, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.57763671875e-05, "grad_norm": 2.82338547706604, "learning_rate": 1e-05, "loss": 0.4244, "mean_token_accuracy": 0.8809996247291565, "num_tokens": 894849116.0, "step": 23570 }, { "epoch": 2.9984734766569137, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6550753116607666, "learning_rate": 1e-05, "loss": 0.3894, "mean_token_accuracy": 0.8888450264930725, "num_tokens": 894882020.0, "step": 23571 }, { "epoch": 2.9986006869355046, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.6253204345703125e-05, "grad_norm": 2.824279546737671, "learning_rate": 1e-05, "loss": 0.4234, "mean_token_accuracy": 0.8814740777015686, "num_tokens": 894916994.0, "step": 23572 }, { "epoch": 2.9987278972140947, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.6560330390930176, "learning_rate": 1e-05, "loss": 0.4472, "mean_token_accuracy": 0.8725864291191101, "num_tokens": 894952763.0, "step": 23573 }, { "epoch": 2.9988551074926852, "ewc_loss": 0.083984375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.649162292480469e-05, "grad_norm": 2.840223550796509, "learning_rate": 1e-05, "loss": 0.4201, "mean_token_accuracy": 0.881406843662262, "num_tokens": 894983427.0, "step": 23574 }, { "epoch": 2.9989823177712758, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6573455333709717, "learning_rate": 1e-05, "loss": 0.4749, "mean_token_accuracy": 0.8674162030220032, "num_tokens": 895019082.0, "step": 23575 }, { "epoch": 2.9991095280498663, "ewc_loss": 0.08447265625, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.673004150390625e-05, "grad_norm": 2.819915533065796, "learning_rate": 1e-05, "loss": 0.3934, "mean_token_accuracy": 0.8908478021621704, "num_tokens": 895054051.0, "step": 23576 }, { "epoch": 2.999236738328457, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.6585800647735596, "learning_rate": 1e-05, "loss": 0.4179, "mean_token_accuracy": 0.880405604839325, "num_tokens": 895090185.0, "step": 23577 }, { "epoch": 2.9993639486070474, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.826338052749634, "learning_rate": 1e-05, "loss": 0.4397, "mean_token_accuracy": 0.8778752684593201, "num_tokens": 895130159.0, "step": 23578 }, { "epoch": 2.999491158885638, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.315376281738281e-05, "grad_norm": 2.6564831733703613, "learning_rate": 1e-05, "loss": 0.4247, "mean_token_accuracy": 0.8791521787643433, "num_tokens": 895169639.0, "step": 23579 }, { "epoch": 2.9996183691642284, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.7206878662109375e-05, "grad_norm": 2.8154454231262207, "learning_rate": 1e-05, "loss": 0.381, "mean_token_accuracy": 0.8951367139816284, "num_tokens": 895209939.0, "step": 23580 }, { "epoch": 2.999745579442819, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.291534423828125e-05, "grad_norm": 2.660421371459961, "learning_rate": 1e-05, "loss": 0.4422, "mean_token_accuracy": 0.8721700310707092, "num_tokens": 895253073.0, "step": 23581 }, { "epoch": 2.9998727897214095, "ewc_loss": 0.0849609375, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.696846008300781e-05, "grad_norm": 2.842064380645752, "learning_rate": 1e-05, "loss": 0.4371, "mean_token_accuracy": 0.8772273063659668, "num_tokens": 895295058.0, "step": 23582 }, { "epoch": 3.0, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.3392181396484375e-05, "grad_norm": 2.647559404373169, "learning_rate": 1e-05, "loss": 0.406, "mean_token_accuracy": 0.8858416080474854, "num_tokens": 895335272.0, "step": 23583 }, { "epoch": 3.0, "ewc_loss": 0.0810546875, "ewc_loss_diag": 0.0003795623779296875, "ewc_loss_parallel": 4.3392181396484375e-05, "step": 23583, "total_flos": 5.156325133756924e+19, "train_loss": 0.44972702040424817, "train_runtime": 50365.0092, "train_samples_per_second": 7.492, "train_steps_per_second": 0.468 } ], "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": 5.156325133756924e+19, "train_batch_size": 4, "trial_name": null, "trial_params": null }